Peter Sommer

Mistakes Before,
During and After
Heat Treatment
of Steel
Mistakes Before, During and After Heat
Treatment of Steel
Peter Sommer

Mistakes Before, During and

After Heat Treatment of
Steel
Peter Sommer
Rhine-Waal University of Applied Science
Issum, Germany

ISBN 978-3-658-44159-3 ISBN 978-3-658-44160-9 (eBook)

https://doi.org/10.1007/978-3-658-44160-9

This book is a translation of the original German edition “Fehler vor, während und nach der Wärmebehandlung
von Stahl” by Sommer, Peter, published by Springer Fachmedien Wiesbaden GmbH in 2022. The translation
was done with the help of an artificial intelligence machine translation tool. A subsequent human revision was
done primarily in terms of content, so that the book will read stylistically differently from a conventional trans-
lation. Springer Nature works continuously to further the development of tools for the production of books and
on the related technologies to support the authors.

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Fachmedien Wiesbaden
GmbH, part of Springer Nature 2024

This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or
dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does
not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective
laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors
or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Fachmedien Wiesbaden GmbH, part of
Springer Nature.
The registered company address is: Abraham-Lincoln-Str. 46, 65189 Wiesbaden, Germany

Paper in this product is recyclable.

Preface

Numerous components and the majority of tools acquire their usage properties through
targeted heat treatment. This requires material-dependent temperature-time-atmosphere
cycles, which can lead to very large stresses in the workpieces. Deviations from the opti-
mal selection of heat treatment parameters can prevent further use of these workpieces.
Excessive distortion, crack formation, and unwanted changes in surface composition
are just some of the possible heat treatment errors. However, it is often overlooked that
even with proper heat treatment, errors can occur that are latent in the workpiece and are
brought to light by the specific conditions of the heat treatment. For example, residual
stresses from mechanical processing are released by exceeding the recrystallization tem-
perature. If these residual stresses are unevenly present in the workpiece, there will also
be different results in the release of these residual stresses. The steels used themselves
can have various deficits or errors, which also affect the heat treatment. The subsequent
processing in the hard state after heat treatment requires high precision to avoid crack
formation or property changes. Finally, improper operating conditions can cause damage
when using the workpieces.
This brief list makes it clear that errors can occur before, during, and after heat treat-
ment. The possible causes are almost unlimited, especially since several influencing fac-
tors can overlap.
This book describes sources of error for all the aforementioned manufacturing stages.
All error examples have been examined in our own testing laboratory. The selection of
examples does not claim to be complete, rather, examples with a traceable source of
error have been predominantly selected.
This book may be of assistance to mechanical engineering students in understanding
the complexity of heat treatment through error descriptions. Heat treatment companies
will realize that a faulty result does not necessarily prove a faulty heat treatment. And all
readers will be amazed at the different possible errors that exist.
The reader may realize that the compilation of error examples with illustrations was
time-consuming and would hardly have been possible without support. The execution
of the damage case investigations with all necessary measurement and testing tasks

V
VI Preface

required high technical knowledge, constant readiness, and modern testing devices.
Many employees of Dr. Sommer Werkstofftechnik GmbH have made a reliable contribu-
tion to this, for which I would like to express my sincere thanks at this point.

Issum, Germany Peter Sommer

Contents

1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2 Design Influence Area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.1 Heat Treatment Specifications in Drawings. . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1.1 Specifications for Annealing Processes. . . . . . . . . . . . . . . . . . . . . . 9
2.1.2 Specifications for Hardening and Tempering. . . . . . . . . . . . . . . . . . 11
2.1.3 Specifications for Surface Hardening . . . . . . . . . . . . . . . . . . . . . . . 13
2.1.4 Specifications for Case Hardening. . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.1.5 Specifications for Nitriding and Nitrocarburizing. . . . . . . . . . . . . . 21
2.1.6 Specifications for Boriding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.2 Thermal Stresses and Notch Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.3 Damage Examples—Construction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3 Material Purchasing Influence Area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.1 Case-hardening Steels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.2 Quenched and Tempered Steels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.3 Nitriding Steels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.4 Martensitic Stainless Steels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.5 Precipitation Hardening Steels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.6 Tool Steels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.7 Damage Examples—Material Purchase . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4 Material Influence Area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.1 Trace Elements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
4.2 Segregation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
4.3 Non-metallic Inclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.4 Hardenability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

VII
VIII Contents

4.5 Damage Examples—Material. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
5 Manufacturing Influence Area Before Heat Treatment. . . . . . . . . . . . . . . . . . 125
5.1 Uneven Material Removal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
5.2 Residual Stresses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
5.3 Surface Hardening. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
5.4 Non-removed Functional Layers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
5.5 Passive Layers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
5.6 Damage Examples—Manufacturing Before Heat Treatment . . . . . . . . . . . 133
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
6 Heat Treatment Influence Area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
6.1 Process Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
6.1.1 The Heating Phase. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
6.1.2 The Holding Phase. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
6.1.3 The Quenching Phase. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
6.1.4 Furnace Atmospheres. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
6.2 Thermochemical Heat Treatments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
6.2.1 Damage Examples—Case Hardening. . . . . . . . . . . . . . . . . . . . . . . 231
6.2.2 Damage Examples—Nitriding/Nitrocarburizing. . . . . . . . . . . . . . . 245
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
7 Manufacturing Influence Area After Heat Treatment. . . . . . . . . . . . . . . . . . . 253
7.1 Damage Examples—Manufacturing After Heat Treatment. . . . . . . . . . . . . 255
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
8 Operating Conditions Influence Area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
8.1 Damage Examples—Operating Conditions. . . . . . . . . . . . . . . . . . . . . . . . . 282
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Introduction
1

Inhaltsverzeichnis

References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

The almost unlimited variety of applications for the steels available today is closely
linked to the targeted property modification through heat treatment. Both the description
of the material steel and the process technology of heat treatment have been described in
countless publications. So why this book? As can be inferred from the title, it is about the
question of when and by what a defect or damage can occur to heat-treated components
or tools. This is also the perennial question when damage has already occurred. Which
material and heat treatment conditions are normal and where does an abnormality begin?
And is an abnormality or a deviation from usual process parameters in heat treatment
actually also the cause of the damage?
Modern quality assurance strives to avoid errors through technical and organizational
measures at all stages of production. And yet, heat treatment occupies a very special
place. While in most manufacturing processes the performed machining can be recog-
nized with the naked eye, this is only possible after heat treatment through appropriate
testing procedures and examination methods. A tool hardened and tempered in a vacuum
furnace does not show this externally, unless the treatment has led to crack formation.
Nevertheless, the treatment carried out may have been faulty and this error only becomes
apparent at a later point in time. The tool could exhibit insufficient wear resistance or
become inoperative due to volume growth during use.
But a completely different error can also occur. After a properly conducted heat treat-
ment, the tool is cracked, distored, or shows deficits in wear resistance. The heat treat-
ment process is often prematurely named as the cause of the damage.

© The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, 1

part of Springer Nature 2024
P. Sommer, Mistakes Before, During and After Heat Treatment of Steel,
https://doi.org/10.1007/978-3-658-44160-9_1
2 1 Introduction

Finally, a tool with a flawless material and heat treatment condition can still fail if the
operating conditions do not correspond to the intended load.
It should be expressly reminded that a faulty material and heat treatment condition
does not automatically represent the cause of a damage that has occurred. The under-
shooting, for example, of an alloy element, a hardness outside the target range, a single
large non-metallic inclusion, or a coarsened microstructure are definitely material or heat
treatment errors. However, often completely different factors are responsible for the fail-
ure of a component and the aforementioned shortcomings have been identified as a result
of a detailed condition description in the course of the damage investigation. Conversely,
it must therefore be stated that all material and heat treatment conditions presented as
faulty can be primarily responsible, but do not necessarily lead to failure.
These few remarks alone make clear the complexity that can lie behind a damage case
of a heat-treated component. This is all the more true when not just one influencing fac-
tor has been identified as the primary cause of damage. In many cases, the accumulation
of several influencing factors leads to failure or total failure. Finally, it should be noted
that the material and heat treatment conditions and damage cases to be described do not
claim to be complete. The possible number of variants is almost unlimited, as indicated
in Fig. 1.1. These areas of influence will be discussed in detail in the following chapters.

Troubleshooting through damage analyses

In the preceding introduction and overview, it was evident that the cause of a defective
heat-treated component can be very complex. The following explains the basic approach
that was applied to all the defect characteristics to be described.

A damage analysis can ideally determine the primary and secondary causes of damage.
For easily identifiable defect characteristics, e.g. untreated heat treatment or material
confusion, a quick finding of results with simple testing methods is possible. A damage
analysis is ultimately also subject to economic conditions. The same normative require-
ments are placed on the testing methods that are used in complex and extensive damage
analyses according to VDI 3822.

Construction Material Fabrication according to WB

Shape Hardenability Eroding defects
Radii Macro inclusions Grinding burn
Material selection ... Segregations ... Electroplating ...

Defective component
Residual stresses Temperature Corrosion
Laser hardening Atmosphere Wear
Friction martensite ... Quenching ... Overload ...
Production before WB Heat treatment Operating conditions

Fig. 1.1 Areas of influence on the component and its properties

1 Introduction 3

Carrying out a damage analysis

Damage analysis according to VDI guideline 3822 Simple and obvious cases of damage
Damage description Component and material investigations
Inventory Assessment of the cause of damage
Damage hypothesis(es) Investigation report
Instrumental analyses
Investigation results
Cause(s) of damage
Damage remedy
Report
Knowledge management

Fig. 1.2 Variants for conducting a damage analysis

The latter is recommended for complex and serious damage cases [1]. The required
investigation sections of the two possible variants are listed in Fig. 1.2.
When a damage analysis is to be conducted, the head of the damage analysis must
decide, according to VDI 3822, whether a neutral, appropriate, and qualitatively unam-
biguous execution of the damage analysis is possible under his leadership and in his lab-
oratory. This not only requires an evaluation of the necessary technical competencies, but
also the available investigation capacity and the verification of whether external expertise
is necessary. In most cases, a detailed investigation offer must be created, although at the
time of the damage report, a complete investigation effort cannot be estimated.
For the investigations and tests to be carried out, accreditation according to DIN EN
ISO 17025 is usually required.
Initial information for the damage description is already transmitted with the damage
report. If this information is insufficient, a discussion for inventory taking with the par-
ties involved takes place. In these discussions, it often turns out that the answers to the
expert’s questions knowingly or unknowingly do not correspond to the actual circum-
stances. There are two different reasons for this. If an employee who was present at the
time of the damage does not have sufficient knowledge about the function of the dam-
aged parts, incorrect or missing information may be unintentionally transmitted. Much
more often, however, the situation is encountered that information is withheld in order
not to be prematurely classified as the cause of the damage.
Design drawings, factory test certificates, delivery specifications or process descriptions
are very helpful in the inventory taking. For self-manufactured products, these documents
are usually available. For externally manufactured products, this is not always possible.
A decision must also be made as to whether an inspection of the damaged system or
the destroyed component on site is necessary. This can not only be helpful for evaluating
the plant environment, but is sometimes necessary for large components when transport
or handling of a several-ton damaged part is not readily possible. It also happens that
4 1 Introduction

recognized cracks have not yet led to a complete failure of a functional part and this
must continue to be operated for some time. Overall, a high degree of flexibility and
great expertise are therefore required, especially in the inventory taking.
In the next phase of the damage analysis, damage hypotheses must be created. At this
point, the causal relationships of individual or multiple causes of damage are still com-
pletely open. Nevertheless, the damage hypotheses must be very well analytically devel-
oped, as this results in the necessary investigation and testing methods. If several damage
hypotheses exist, the order of the tests should also be determined. For example, the veri-
fication of the used material can reveal whether the drawing requirement is met or not.
Although a deviation from the specification does not automatically provide the cause of
the damage, a further investigation may be unnecessary in the case of serious deviations.
The decision on this is to be made by the damage expert.
The necessary instrumental analyses are derived from the damage hypotheses. Both
non-destructive and destructive methods are used. The laboratory must determine and
document the order of all investigation procedures to be carried out.
A magnetic particle inspection makes crack progressions clearly visible and possibly
allows an assignment to geometric features. Internal defects can be located by ultrasonic
tests, ensuring a precise removal of test pieces. Exposing crack surfaces can provide very
important information, which is why this investigation step is of great importance. The
necessary cut-offs for this must be carefully documented. This also applies to all other
cut-offs required for metallographic microsections. It must always be traceable at which
component position a corresponding investigation was carried out.
Positions at the origin of the damage are particularly interesting, provided this has
already been clearly localized. A second position far away from the origin of the dam-
age provides information about the material and heat treatment condition. An undamaged
reference part would be desirable for comparison, but this is not given in the vast major-
ity of all cases.
In the investigation planning, it must also be taken into account that the removal of
test parts must be done without any changes. Sufficient cooling is necessary when cut-
ting and any residues to be examined must be removed before cutting.
The examination of fracture surfaces is initially carried out with the naked eye or
at low magnification in the stereo microscope. Only then is a scanning electron micro-
scopic examination useful, in which energy-dispersive X-ray analyses can also be used.
This investigation method is very helpful, especially in the identification of non-metallic
inclusions.
For the evaluation of a hydrogen-induced brittle fracture, a comparison with a labora-
tory-induced fracture has proven very successful. It is also possible to evaluate the frac-
ture surface after a new hydrogen loading.
An important core piece of the damage analysis is the metallographic examination
and evaluation. The microstructure formation must not only be identified, but also evalu-
ated in terms of possible abnormalities or anomalies. The verification of hardness and, in
the case of surface-hardened components, also the hardness progression are usual stand-
ard investigation methods.
1 Introduction 5

The determination of mechanical properties in the tensile test and notch impact bend-
ing test is rather rare in damage analyses, as sufficient sample material is required for
this. This is naturally not available for small parts.
For case-hardened components or high-carbon tool steels, only X-ray examination
methods provide objective results on the residual austenite content.
The list of examination methods could be significantly expanded. However, the spe-
cific testing methods used will only be explained in detail in the damage examples.
After all necessary investigations have been completed, the determination of the cause
of damage is then carried out. Sometimes, several cumulatively acting causes have to be
evaluated. The previously formulated damage hypotheses must now be assessed. In the
best case, a clear cause of error can be identified. However, it also happens that not only
several causes of damage were present, but also no clear cause of damage can be deter-
mined. This can be the case, for example, when the position of the origin of the dam-
age was no longer available or secondary damages made an evaluation impossible. This
result is unsatisfactory, but it is also a possible statement.
In the VDI guideline, the damage assessment is followed by the damage remedy. In
order to make reliable statements about the repetition of similar damages, a consider-
able amount of experience is required. The stages of a component from development to
production to use are so diverse that after describing the cause of the error, only the spe-
cialist departments can initiate appropriate measures. Therefore, it is of great importance
to make the report of a damage analysis available to all specialist departments. The VDI
guideline describes this as knowledge management.
For the report preparation, it should be noted that the results from all tests and inves-
tigations must be presented factually and without evaluation. The evaluation of the
results then takes place in a clearly marked separate section. This procedure is neces-
sary because only the testing procedures can be accredited. The evaluation of a damage
case, on the other hand, is the interpretation of the responsible damage expert and is not
accreditable. It is also possible that the interpretation can vary among different experts.
Both for the damage expert and for the client of a damage analysis, electronic archiv-
ing with meaningful keywords is of great benefit. This allows insights from previous
investigations to be easily found and taken into account for a current investigation.
All the errors and damage cases described in this book have been investigated in the
author’s accredited laboratory. Therefore, this book does not claim to provide a complete
overview of all possible cases of heat-treated components. This is probably not possi-
ble due to the variety of different causes and effects. The assignment to the respective
areas of influence was made based on scientific evaluation and available information.
However, experience-based subjective influences cannot be avoided. If several influenc-
ing factors were responsible for a damage, the primary cause was always sought and
assigned to this area of influence.
In the damage examples listed in this book, a description of all complete investigation
steps and results was deliberately omitted in most cases. Instead, only the results and
6 1 Introduction

boundary conditions that led to the primary cause of damage are presented. However, in
each individual case, a complete damage report was prepared in accordance with VDI
guideline 3822.

References

1. VDI-Richtlinie 3822:2011-11: Grundlagen und Durchführung einer Schadensanalyse.

Design Influence Area
2

Contents

2.1 Heat Treatment Specifications in Drawings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2.2 Thermal Stresses and Notch Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.3 Damage Examples—Construction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

Design and development staff must develop functional solutions and face an enormously
large catalog of requirements. The component to be designed must fulfill the intended
function and have the necessary, load-specific properties. Even this task description con-
tains considerable risks. Often, the specific loads are not exactly known and assumptions
must be made. Intelligent FEM tools are available, but the boundary conditions for these
simulation calculations are often inaccurate or unknown.
In addition to the functional solution, properties such as weight, availability, machina-
bility, or environmental influences must also be considered and evaluated. If the required
properties require heat treatment, e.g., quenching and tempering, the suitability of the
chosen material for this is also an important influencing factor. The term “heat treatment-
compatible” design encompasses the entire complexity of this term [1] and possible
influencing factors are listed in Fig. 2.1.
Any asymmetric mass distribution causes an asymmetric temperature field during heat
treatment, and the poorer the hardenability of a steel, the more prone to warping and
cracking are “difficult” geometries. From [1] two simple design examples are shown in
Fig. 2.2. Further examples can be seen from DIN 17022-3:1989-04 [2] in Fig. 2.3.
In addition to all technically oriented influencing factors, a value-analytical considera-
tion also decides on the feasibility of a component. A brilliant and technically optimal
solution can certainly fail due to manufacturing costs.

© The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, 7

part of Springer Nature 2024
P. Sommer, Mistakes Before, During and After Heat Treatment of Steel,
https://doi.org/10.1007/978-3-658-44160-9_2
8 2 Design Influence Area

Design and development sphere of influence

Design and dimensioning

Cross-section transitions
Definition of properties and tolerances
Notch effect due to insufficient radii
Material selection Insufficient spacing between
openings
Hardenability, availability,
own material data sheet Mass distribution
Heat treatment specifications ...
Definition of test criteria, Value analysis
Test methods, test positions
... Material costs
Production costs
...

Fig. 2.1 Listing of influencing factors in the area of design

unfavorable favorable unfavorable favorable

Fig. 2.2 Simple design examples with unfavorable and favorable shapes [1]

Risk of cracking
during hardening

High risk
of cracking

Low risk
of cracking

Fig. 2.3 Shape optimization to avoid hardening cracks [2]

2.1 Heat Treatment Specifications in Drawings 9

The following only considers and presents damage cases from the area of design that
included the heat treatment manufacturing step, were implemented, and failed or were
destroyed during heat treatment or later in use.

2.1 Heat Treatment Specifications in Drawings

The result of a design solution is a manufacturing drawing with all the information nec-
essary for the production of the component. For the heat treatment specifications, DIN
ISO 15787:2018-08 [3] is a valuable source of information. For all the heat treatment
processes still to be described, the condition after the heat treatment, e.g., “hardened”,
“hardened and tempered”, “case hardened”, “surface hardened”, “nitrided” or “borided”
etc., is specified.
The correct description of the abbreviations avoids misunderstandings if they are
properly named and interpreted by all parties involved. The abbreviations originally
derived from German usage have now been replaced by internationally valid terms,
Table 2.1.
The use of these standardized abbreviations is, however, only the first and easily
understandable step. Much more difficult is the appropriate determination of the prop-
erty value and the tolerance for the target size. While in the predecessor version of DIN
ISO 15787:2010-04 [4] recommendations for limit deviations were given for numerous
surface hardness methods, these recommendations are no longer included in the current
DIN ISO 15787:2018-08. From a standardization point of view, this omission is a con-
sistent step, as it is a drawing standard that cannot evaluate specific process variables of
heat treatment.
The appropriate determination of the permissible tolerance ranges is not freely
selectable by the design. Rather, the analysis scatter ranges of the standardized steels,
the component geometry, the process variables of the heat treatment, and the possible
scatterings of the test methods must be taken into account in the evaluation. Experience
shows that this requires close coordination between the design and the hardening shop.
Even a properly determined tolerance range accepted by all parties can lead to distorted
results and erroneous conclusions due to faulty preparations and measurement methods.

2.1.1 Specifications for Annealing Processes

Annealing treatments are usually carried out on pre-material products to improve

machinability, reduce residual stresses, eliminate hardening from cold forming, or to
achieve a finer grain size. Accordingly, the following terms should be used:

• “stress relieving”,
• “soft annealed”,
10 2 Design Influence Area

Table 2.1  
Abbreviations for specifications in technical drawings according to DIN ISO
15787:2018-09
Abbreviation Naming Abbreviation Naming Naming
According to DIN 6773:2001-04 [5] According to ISO 15787:2001 According to DIN
ISO 15787:2018-08
Eht Case hardening CHD Case hardening Case hardening
depth depth hardness depth
At Carburization depth CD Carburization depth Carburization depth
VS Compound layer CLT Compound layer Compound layer
thickness thickness thickness
Sht Melting hardness FHD Fusion hardness Melting hardening
depth depth hardness depth
Nht Nitriding hardness NHD Nitriding hardness Nitriding hardness
depth depth depth
Rht Hardening depth SHD Surface hardening Hardening depth
depth after surface hard-
ening
SHP Melting hardening FTS Fusion treatment Melting hardness
treatment plan specification plan
SHA Melting hardening – – –
instruction
WBA Heat treatment HTO Heat-treatment Heat treatment
instruction order instruction
WBP Heat treatment plan HTS Heat-treatment Heat treatment plan
specification
– – IOD Internal oxidation Edge oxidation
depth depth
– – OLT Oxide layer thick- Oxide layer thick-
ness ness

• “annealed to spheroidal carbides”,

• “recrystallizing”,
• “normalized”.

In the respective material standards, the property values for hardness, the mechanical
technological properties, and some structural features can be found. For stress-relief
annealing, no sensibly measurable property specifications can be determined. The deter-
mination of residual stresses by means of X-ray methods is very complex and also lim-
ited to small measuring areas.
On the other hand, specifications for annealing to spheroidal carbides are quite rec-
ommendable. This is especially true for products that are later produced by cold forming.
2.1 Heat Treatment Specifications in Drawings 11

In these cases, the proportion of non-formed lamellar pearlite (PA) should be low and the
carbide formation can be described and determined according to Steel-Iron Test Sheet
1520:1998-09 [6]. However, these are always company-specific determinations or deliv-
ery specifications, e.g. PA max. 3.2; CG: 2.1 to 2.3 according to the picture series table
in SEP 1520:1998-09. These structural features are not described in the standardization
works of the steels.

2.1.2 Specifications for Hardening and Tempering

The target specification for a hardness value or a tensile strength with indication of the
material and the heat treatment condition is one of the standard specifications of a heat
treatment to be carried out. This statement seems banal at first, but on closer inspection,
there are some points to be clarified. With the specification, e.g. X37CrMoV5-1, 1.2343,
44 ±2 HRC “hardened and tempered”, the hardness method, the Rockwell-C test accord-
ing to DIN EN ISO 6508-1:2016-12 [7], and the heat treatment condition are defined at
the same time. If the measuring point is also indicated in a component drawing and the
component can be tested using HRC, this target specification is clear and feasible, Fig.
2.4.
However, clarification is needed when the component cannot be tested using HRC
due to its geometry, the tolerance range is unrealistically small, and the heat treatment
process is only indicated as “hardened”, or the material cannot even reach the required
hardness. In this case, the specification is not clear, which can lead to the possible omis-
sion of tempering. This, in turn, can represent a latent cause of damage.
If the hardness measurement cannot be carried out using HRC, it must be agreed
which test should be carried out as a substitute, e.g. Vickers hardness test according
to DIN EN 6507:2018-07 [8], and if desired, which conversion table in DIN EN ISO
18265:2014-02 [9] should be used for the measured hardness to enable a comparison
with the drawing specification.

Fig. 2.4 Marking of a part in the hardened and tempered state [3]
12 2 Design Influence Area

A directly measurable specification for the hardness or the indication of the hardness
method in the component drawing is always the best solution, but by no means always
encountered.
Even more serious is the specification of a tensile strength value that cannot be meas-
ured on the component. The following example for the heat-treatable steel 42CrMo4,
1.7225, shows how significant the agreement of the conversion table is.
Drawing specification: 1000 +150 MPa “quenched and tempered”.
For this material and the heat treatment condition “quenched and tempered”, the con-
version table B2 according to DIN EN ISO 18265 is to be used. However, the speci-
fication of this or another agreed conversion table is very rarely listed in the drawing
specifications or order regulations, which is why the more universal conversion table A.1
(unalloyed steels) is often used.
Using the conversion table A.1, the tensile strength specification of 1000 +150 MPa
results in a hardness specification of 296 +40 HB. On the other hand, if the correct con-
version table B.2 (quenched and tempered steels in the quenched and tempered condi-
tion) is used, the result is 315 +47 HB.
If the average tolerance range is used for the required tempering temperature in both
cases, a difference in the tempering temperature of 640 °C (A.1) to 615 °C (B.2) results.
This would lead to noticeable differences, especially in toughness.
Attention should also be drawn to the following standard requirement of DIN EN
18265:2014-02:

“A direct application of the values determined by a hardness conversion according to this

International Standard is only possible on exactly the same materials. For all other materi-
als, these values serve only as a general indicator. In all cases, the results of the conversion
are not intended to replace the values determined according to the applicable standardized
methods. The least reliable are the tensile strength values estimated according to this Inter-
national Standard.” The steels are listed in Table B.1 of the mentioned standard. It should
also be noted that there are no conversion tables available for a number of material groups,
as was explained in more detail in [10].

The tolerance ranges specified in drawings for the various hardness testing methods are
not contained in any standard and there are also no normative numerical recommenda-
tions. DIN ISO 15787:2018-08 does refer to the fact that each hardness value must be
assigned a tolerance. However, it only states that this tolerance should be as large as
functionally possible. It is often overlooked what tolerances already exist in the hardness
testing devices themselves. The permissible measurement deviations are given in the test
standards [11–13].
For the Rockwell C hardness scale, the permissible measurement deviation during
regular inspection with calibrated hardness comparison plates is ±1.5 HRC. This means
that a drawing specification with a tolerance range of 2 HRC is a stricter requirement
than the standard permissible measurement deviation of the calibrated hardness compari-
son plate.
2.1 Heat Treatment Specifications in Drawings 13

The permissible limit deviations for Brinell and Vickers hardness tests depend both
on the test load and the hardness level. Even at the highest test load and the highest hard-
ness, the permissible limit deviation is ±2 % of the hardness value of the calibrated
hardness comparison plate.
The calculation of long-term measurement uncertainties does allow for some limita-
tion of the tolerance range, but this only applies to the respective hardness testing device.
A comparative measurement on another hardness testing device can lead to different
results again.
It would be desirable if the permissible limit deviations of the hardness test plates
were not undercut when determining the tolerance range of hardness values.

2.1.3 Specifications for Surface Hardening

Surface hardening is characterized by heating only the near-surface area followed by

quenching. This can be done over the entire surface area, but can also be limited to
certain areas. Induction hardening, flame hardening, laser beam hardening, and elec-
tron beam hardening are used as case hardening methods [14]. However, DIN ISO
15787:2018-08 does not distinguish between these quite different method variants.
Only surface hardening is detailed.
The specification of the material, as with all heat treatment processes, is a fundamen-
tally necessary drawing specification. However, in the case of surface hardening meth-
ods, the delivery condition, e.g. +N or +QT, is also of great importance. Depending on
the delivery condition, different hardness results are achieved [15].
The previously used term for the hardening depth Rht has now been replaced by the
Surface-Hardening-Depth (SHD). This specification value must be given with an indi-
vidual limit hardness, e.g. as shown in Fig. 2.5. The limit hardness is in most cases 80 %
of the minimum surface hardness. The hardening depth is measured in the cross-section
using HV 1 and it is the vertical distance from the surface to the limit hardness.
The surface hardness is usually not given in “HV 1” units, which is why the limit
hardness must be determined by a hardness value conversion depending on the hardness
testing method. Much simpler is the determination of the limit hardness using Table A.4
of the predecessor standard DIN ISO 15787-2010-01 [4]. For the hardness testing meth-

Fig. 2.5 Marking of a surface -hardened part [3]

14 2 Design Influence Area

ods HV, HRC, HRA, HR 15 N, HR 30 N and HR 45 N, the limit hardness values are
set in 25 “HV” steps. In the current version of this standard [3], this table is no longer
included. However, this determination is highly recommended and avoids misunder-
standings.
Induction hardening in particular allows many process variants. If surfaces, e.g.
on a cam disc, are hardened all around, a slip point is inevitable. The hardening must
be stopped at a small distance from the initial position to avoid crack formation when
reheating an already hardened zone. Consequently, a small area remains soft. The design
should then mark the best possible position for this slip point in the drawing, see Fig.
2.6.
For gear components, full tooth hardening, tooth flank case hardening, and tooth root
case hardening are possible depending on the module, as detailed in [3].
In all surface hardening methods, the component geometry must be taken into
account in a special way. In the hardness run-out zone, there is always a mixed structure,
which causes a “metallurgical notch”. A cross-sectional transition causes a jump in stiff-
ness and thus a “mechanical notch”. If there is a local overlay of both notch factors, there
is an increased risk of crack initiation in use. Some examples from [14] show positive
and negative hardness images, Fig. 2.7.

2.1.4 Specifications for Case Hardening

The determination of the case hardening depth CHD is standardized in DIN EN ISO
2639:2003-04 [16]. In this standard, the “case hardening depth after carburizing and
hardening is defined as the vertical distance from the surface to the layer that has a Vick-
ers hardness of 550 HV 1 according to ISO 6507-1, Fig. 2.8, [8] or an equivalent Knoop
hardness according to ISO 4545[17]”.

Fig. 2.6 Marking of a surface

-hardened part with a slip point
7 ±2

[3]

Material 42CrMo4 +QT, 1.7225

surface hardened, see HTO
(60 ±2) HRC, SHD 400 = 1.8 ±0.5 mm
2.1 Heat Treatment Specifications in Drawings 15

good bad good bad

a) Groove, notch

d) Journal, flange

b) Drilling

e) Shaft, bearing
c) Career

Fig. 2.7 Positive and negative formation of the hardness layer in case hardenings [14]

900

800
CDH550 = 0.71 mm
700

600
Hardness [HV 1]

500

400

300

200

100

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Edge distance [mm]

Fig. 2.8 Hardness profile for determining the case hardening depth

In this definition, the test after case hardening is specified with “carburizing and hard-
ening”, which is why strict application of this standard would lead to a determination
before tempering. However, since the design specifies the target CHD for the usage prop-
erties and thus for the completely carried out heat treatment process, the determination of
the case hardening depth typically takes place after case hardening and tempering.
16 2 Design Influence Area

The design must not only specify the target value and the permissible tolerance, it
must also determine the test location. Especially for individual parts or smaller series,
the test location often cannot be measured directly, as the component would have to be
destroyed for this. In these cases, a suitable coupon sample must be used. For gearing
components, the requirements for two different coupon samples are defined in ISO 6336-
5:2016-08 [18]:

a) For process control, any sample can be used. This sample only describes the heat
treatment process and the repeatability when using the same type of sample. State-
ments about the microstructure formation of the gearing component are not possible
based on this sample, however, a correlation for the case hardening depth between
component and sample can be worked out.
b) As a representative coupon sample, a module-dependent and material-identical
sample with a diameter of 3 x module and a length of 6 x module is required. This
achieves a comparable cooling speed to the gearing, which allows a representative
microstructure evaluation.

With frequently changing modules, variant b) can only be realized with increased effort,
as different sample formats would have to be produced. Therefore, often a material-iden-
tical sample is used, but with constant dimensions. The correlation between the result of
the sample and the gearing component must then, however, be determined as described
under a).
The differences in the case hardening depth between different test positions are both
geometry- and material-dependent. The tooth head has a large surface and a small vol-
ume, which leads to a higher quenching speed and thus also to a greater case hardening
depth. In the tooth base and at the 30° tangent, the ratio of surface to volume reverses,
which is why the smallest case hardening depth is present at these positions.
The necessity to determine the exact test position is evident from Fig. 2.9. Two gears
with identical module and same weight were case hardened together in one batch. One
gear was made from the case hardening steel 16MnCr5 +HH and the other gear from the
higher alloyed case hardening steel 18CrNiMo7-6 +HH. In the tooth profile made from
the case hardening steel 18CrNiMo7-6 +HH (right part of the picture), the test position
is hardly noticeable. In contrast, the less hardenable case hardening steel 16MnCr5 +HH
shows clear differences at the various positions.
The different results in case hardening depth can also be quantified, Fig. 2.10
and 2.11. If a gear, for example, with module 8 is case hardened from a case harden-
ing steel with low hardenability, differences of 100 % exist within a tooth between the
30° tangent and the flank. The difference to the tooth head is even greater. When using
the higher alloyed case hardening steel 18CrNiMo7-6, the difference between 30° tan-
gent and flank is reduced to 20 %. However, with both materials, it is also clear that the
results on the flank correspond quite well with the results of the material-like reference
sample with a diameter of 30 mm.
2.1 Heat Treatment Specifications in Drawings 17

Fig. 2.9 Formation of the case hardening layer in steels with different hardenability (left
16MnCr5 +HH; right 18CrNiMo7-6 +HH)

1.7131 CHD 1.5-2.0 mm | Module-8 Inner-Ø 212 mm

2.50
Gear wheel CHD Min/Max Reference sample
2.30

2.10
Max: CHD 2.00
1.90
Reference pobe: 1.66 1.74
1.70
CHD [mm]

Min: CHD 1.50 1.62

1.50

1.30

1.10
1.09
0.90
0.89
0.70

0.50
Foot 30° Flank Head
Measuring position

Fig. 2.10 Location-dependent results of the case hardening depth on a gear module 8, inner diam-
eter 212 mm, material 16MnCr5 +HL [19]

When specifying the case hardening depth according to the drawing at a certain posi-
tion, it must also be taken into account whether and possibly which depth is removed
by hard machining after case hardening and tempering. While tooth flanks are almost
always ground, the tooth base usually remains unprocessed.
18 2 Design Influence Area

1.6587 CHD 1.5-2.0 mm | Module-8 Inner-Ø 212 mm

2.50
Gearwheel CHD Min/Max Reference sample
2.30

2.10
Max: CHD 2.00
1.90
Reference pobe: 1.64 1.69 1.73
1.70
CHD [mm]

1.55
1.50
Min: CHD 1.50 1.42
1.30

1.10

0.90

0.70

0.50
Foot 30° Flank Head
Measuring position

Fig. 2.11 Location-dependent results of the case hardening depth on a gear module 8, inner diam-
eter 212 mm, material 18CrNiMo7-6 [19]

+5
10 0

Case-hardened and tempered as a whole part

(680+1400 ) HV 10
CHD = 0.3+0.20

Fig. 2.12 Marking of a part in the hardened and tempered state [4]

If carburizing at certain positions of the component is to be prevented, only the area

to be carburized should be marked in the drawing. Here too, DIN ISO 15787:2018-08
indicates the correct form of marking by a dash-dot line. An example from this stand-
ard is shown in Fig. 2.12. Further details on locally limited case hardening can be taken
directly from this standard.
In addition to the case hardening depth, the specification of a surface hardness is also
necessary. In addition to the information from Sect. 2.1.2, the permissible test load must
2.1 Heat Treatment Specifications in Drawings 19

be taken into account depending on the case hardening depth and the expected surface
hardness. The information contained in [4] is listed in Tables 2.2, 2.3and 2.4. However,
these tables are no longer included in the subsequent edition of DIN ISO 15787:2018-8.
The case hardening depth and surface hardness are determined both by carburizing and
by the subsequent quenching and tempering, as well as by the hardenability of the steel

Table 2.2  Choice of the test method for specifying the hardness details according to the mini-
mum hardening depth and the minimum surface hardness in HV (Table A.1 in [3])
Minimum Minimum surface hardness in HV
hardening 200 to 300 Over 300 Over 400 Over 500 Over 600 Over 700 Over 800
depth HV to 400 HV to 500 HV to 600 HV to 700 HV to 800 HV
SHD,
CHD,
NHD,
FHD
mm
0.05 – – – HV 0.5 HV 0.5 HV 0.5 HV 0.5
0.07 – HV 0.5 HV 0.5 HV 0.5 HV 0.5 HV 1 HV 1
0.08 HV 0.5 HV 0.5 HV 0.5 HV 0.5 HV 1 HV 1 HV 1
0.09 HV 0.5 HV 0.5 HV 0.5 HV 1 HV 1 HV 1 HV 1
0.1 HV 0.5 HV 1 HV 1 HV 1 HV 1 HV 1 HV 3
0.15 HV 1 HV 1 HV 3 HV 3 HV 3 HV 3 HV 5
0.2 HV 1 HV 3 HV 5 HV 5 HV 5 HV 5 HV 5
0.25 HV 3 HV 5 HV 5 HV 5 HV 10 HV 10 HV 10
0.3 HV 3 HV 5 HV 10 HV 10 HV 10 HV 10 HV 10
0.4 HV 5 HV 10 HV 10 HV 10 HV 10 HV 30 HV 30
0.45 HV 5 HV 10 HV 10 HV 10 HV 30 HV 30 HV 30
0.5 HV 10 HV 10 HV 10 HV 30 HV 30 HV 30 HV 30
0.55 HV 10 HV 10 HV 30 HV 30 HV 30 HV 50 HV 50
0.6 HV 10 HV 10 HV 30 HV 30 HV 50 HV 50 HV 50
0.65 HV 10 HV 30 HV 30 HV 50 HV 50 HV 50 HV 50
0.7 HV 10 HV 30 HV 50 HV 50 HV 50 HV 50 HV 50
0.75 HV 30 HV 30 HV 50 HV 50 HV 50 HV 100 HV 100
0.8 HV 30 HV 30 HV 50 HV 50 HV 100 HV 100 HV 100
0.9 HV 30 HV 30 HV 50 HV 100 HV 100 HV 100 HV 100
1 HV 30 HV 50 HV 100 HV 100 HV 100 HV 100 HV 100
1.5a HV 30 HV 50 HV 100 HV 100 HV 100 HV 100 HV 100
2a HV 30 HV 50 HV 100 HV 100 HV 100 HV 100 HV 100
2.5a HV 30 HV 50 HV 100 HV 100 HV 100 HV 100 HV 100
20 2 Design Influence Area

Table 2.3  Selection of the test method for determining the hardness specifications according to
the minimum hardening depth and the surface minimum hardness in HR 15 N, HR 30 N, HR 45 N
(Table A.2 in [3])
Mini- Minimum Surface Hardness in HRN
mum 82 to Over Over 60 to Over Over Over 44 to Over Over Over
Hard- 85 HR 85 to 88 68 HR 68 to 73 to 78 54 54 to 61 to 67
ness 15 N 88 HR 30 N 73 78 HR HR 61 67 HR
Depth HR 15 15 N HR 30 HR 30 30 N 45 N HR 45 HR 45 45 N
SHD, N N N N N
CHD
mm
0.1 – – HR – – – – – – – –
15 N
0.15 – HR 15 HR – – – – – – – –
N 15 N
0.2 HR 15 HR 15 HR – – – HR – – – –
N N 15 N 30 N
0.25 HR 15 HR 15 HR – – HR 30 HR – – – –
N N 15 N N 30 N
0.35 HR 15 HR 15 HR – HR 30 HR 30 HR – – – HR
N N 15 N N N 30 N 45 N
0.4 HR 15 HR 15 HR HR 30 HR 30 HR 30 HR – – HR 45 HR
N N 15 N N N N 30 N N 45 N
0.5 HR 15 HR 15 HR HR 30 HR 30 HR 30 HR – HR 45 HR 45 HR
N N 15 N N N N 30 N N N 45 N
≥ 0.55 HR 15 HR 15 HR HR 30 HR 30 HR 30 HR HR HR 45 HR 45 HR
N N 15 N N N N 30 N 45 N N N 45 N

used. Independent of these process variables are the parameters carburizing depth CD
(carbon depth) and intergranular oxidation depth IGO. While the carburizing depth with
a limit carbon content of e.g. “CD0.35” can certainly be set, the integranularr oxidation
depth depends on the carburizing duration and temperature. Intergranularoxidation is not
entirely avoidable in all gas carburizing processes using CO-containing gases, and appro-
priate specifications are contained in [18]. Only in vacuum carburizing processes, where
acetylene is used as a carbon carrier, is intergranular oxidation completely avoidable.
For case hardening, a process specification is also possible, but this can only be found
in technical drawings in special cases. Fig. 2.13 schematically shows the possible vari-
ants. If the term “case hardened and tempered” is used without a process description,
direct hardening is usually carried out. The three other process variants are carried out,
for example, with very large case hardening depths, where grain growth is possible. Sin-
gle hardening is also used when mechanical processing is required at certain component
positions after carburizing.
2.1 Heat Treatment Specifications in Drawings 21

Table 2.4  Selection of the test method for determining hardness specifications according to the
minimum hardening depth and the minimum surface hardness in HRA or HRC (Table A.3 [3])
Mini- Minimum surface hardness in HRA or HRC
mum 70 to 75 Over 75 Over 78 Over 81 40 to 49 Over 49 Over 55 Over 60
harden- HRA to 78 to 81 HRA HRC to 55 to 60 HRC
ing depth HRA HRA HRC HRC
SHD,
CHD
mm
0.4 – – – HRA – – – –
0.45 – – HRA HRA – – – –
0.5 – HRA HRA HRA – – – –
0.6 HRA HRA HRA HRA – – – –
0.8 HRA HRA HRA HRA – – – HRC
0.9 HRA HRA HRA HRA – – HRC HRC
1 HRA HRA HRA HRA – HRC HRC HRC
1.2 HRA HRA HRA HRA HRC HRC HRC HRC

AC3 Core
AC3 Core
Temperature
Temperature

Hardening AC1Margin
Hardening AC1Margin
Tempering Tempering
Carburizing Carburizing
Time Time

AC3 Core AC3 Core

Temperature

Temperature

AC1Margin AC1Margin
Convert Hardening Hardening
Tempering Tempering
Carburizing Carburizing
Time Time

Fig. 2.13 Process variants of case hardening according to [2] Direct hardening (top left); Single
hardening (top right); Hardening after isothermal transformation (bottom left); Double hardening
(bottom right)

2.1.5 Specifications for Nitriding and Nitrocarburizing

Although the nitriding hardness depth is also determined by a hardness progres-

sion measurement according to DIN 50190-3:1979-03 [20], the determination of the
limit hardness value differs from the CHD measurement. First, the core hardness must
be determined. The limit hardness then results from the core hardness +50 HV, where
rounding must be done to 10 HV each time. As a rule, the limit hardness is given as
22 2 Design Influence Area

Fig. 2.14 Marking of a

plasma nitrided part on all
sides [3]

plasma-nitrided NHD = 0.30 ±0.05 mm

Surface hardness ≥ 950 HV10

Fig. 2.15 Marking of a 40 ±0.5

locally nitrided part [3]

nitrided NHD400 = 0.4 ±0.1 mm

Surface hardness ≥ 900 HV10
…………… not nitrated

Vickers hardness HV 0.5. However, the aforementioned test standard DIN 50190-3
allows for different test forces from HV 0.3 to HV 2 by agreement. This must then also
be specified in the drawing.
The short form of the nitriding hardness depth Nht given in this standard has now
been updated to NHD (Nitriding Hardness Depth) in DIN ISO 15787:2018-06 (see also
Table 2.1). An example from this standard is shown in Fig. 2.14. It should be noted that
this value cannot be specifically set, but depends on the amount of available nitride form-
ers and the delivery condition.
Similar to case hardening, localized nitriding is also possible. However, unlike case
hardening, the area not to be nitrided is marked in nitriding, as exemplified in Fig. 2.15.
Fig. 2.16 shows a typical result of determining the nitriding hardness depth of a gas
nitrided component made of material 34CrAlNi7-10, material no. 1.8550. It is also
important to note the standard specification that in case of disputes, hardness progres-
sions must be created at at least two points.
The listed examples of nitriding refer to steels that have been quenched and tempered
before nitriding. In these cases, the quenched and tempered heat treatment condition pro-
vides sufficient support hardness below the nitriding layer, which is why higher surface
pressures are permissible. On the other hand, nitrocarburizing preferably defines the for-
mation of a certain compound layer thickness as the target size, and this heat treatment
2.1 Heat Treatment Specifications in Drawings 23

1200

1000

NHD370 = 0.39 mm
800
Hardness [HV 1]

600

400

200

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Edge distance [mm]

Fig. 2.16 Determination of nitriding hardness depth

Fig. 2.17 Marking of a

nitrocarburized part on all
sides [3]

nitrocarburiert
CLT = (15 ±5) µm

method is particularly suitable for reducing adhesive wear. In the drawing, the abbrevia-
tion CLT (Compound Layer Thickness) should be used (Fig. 2.17).
If nitriding is to be carried out in a specific medium, the term nitrocarburized should
be expanded to include the process variants salt bath nitrocarburizing, plasma nitrocar-
burizing, or gas nitrocarburizing.
Determining the compound layer thickness requires special care, which is why the
implementation regulations of DIN 30902:2016-12 [21] should be made mandatory. This
also applies if an oxide layer is to be formed after nitrocarburizing to improve corrosion
resistance. The valid abbreviation for this is OLT (oxide layer thickness). The pore seam,
which occurs in almost all nitriding and nitrocarburizing processes, is not described in
24 2 Design Influence Area

DIN ISO 15787:2018-08. However, information is often found on drawings such as:
“Pore seam max. 40 % of the CLT”.
Fig. 2.18 shows a microstructure image of a nitrocarburized and oxidized compo-
nent made from the unalloyed heat-treatable steel C45E, material no. 1.1191. A ferrite-
pearlite microstructure, the subsequent pore-free compound layer, the pore seam, and the
oxide layer can be seen. To avoid chipping of the compound and oxide layer during the
microsection preparation, the use of a support film is recommended [22]. In this exam-
ple, a thin copper foil was used.

2.1.6 Specifications for Boriding

Boriding is one of the rarer thermochemical heat treatment processes in which a com-
pound layer (CLT) is also produced. The correct drawing specification can be found in
Fig. 2.19.

Fig. 2.18 Compound layer and oxide layer on a nitrocarburized component made of C45E
2.2 Thermal Stresses and Notch Effects 25

Fig. 2.19 Marking of a

borided part [3]

boriert CLT = (30 ±10) µm

2.2 Thermal Stresses and Notch Effects

Numerous publications describe recommendations to reduce the risks of crack formation

or tendency to distore already in the design phase [1, 2, 23, 24]. Of great importance is
the fact that an inhomogeneous mass distribution causes an inhomogeneous temperature
field both during heating and during quenching. The temperature differences become
larger the higher the heating or cooling speed and the larger the cross-section.
The magnitude of possible thermal stresses is often underestimated. When quenching
a cylindrical body with a diameter of 100 mm from 850 °C to 20 °C in water, max. tem-
perature differences between edge and core of approx. 500 °C occur, Fig. 2.20. Under
linear-elastic behavior, tensile stresses of max. 620 MPa and equal compressive stresses
in the core would occur, Fig. 2.21. After temperature equalization, the stresses under this
assumption would be completely relieved. However, this rather unlikely case under real
conditions makes it clear that the occurring thermal stresses are directly proportional to
the temperature difference.
The resulting thermal stresses can lead to distortion or, in extreme cases, to crack
formation. When the hot yield limit is exceeded, plastic flow occurs, leading to a stress
relief, Fig. 2.22. After the heat treatment is completed, this can be identified as distor-
tion. If plastic flow is no longer possible, cracking can also occur.
In components with very large cross-sectional differences, the risk of distortion
increases due to the formation of large thermal stresses. For very complicated component
geometries, it is therefore recommended to use steels with higher hardenability, which
can be quenched in milder quenching media, e.g. with gas pressure cooling. Further
information can be found in [25].
Another factor in minimizing the risk of a component to be heat-treated is a suffi-
ciently dimensioned radius at cross-sectional transitions that cause jumps in stiffness.
Insufficiently sized radii always have a large notch effect, which leads to stress increases.
Numerous recommendations can also be found in the standards and in the literature for
this, [23, 24]. A typical example is shown in Fig. 2.23.
26 2 Design Influence Area

1000
Temperatures
Core
Edge
Difference
800
Temperature [°C]

600

400

200

0
0.001 0.01 0.1 1 10 100
Time [min]

Fig. 2.20 Cooling curves during quenching of a cylinder with 100 mm Ø in water

800

600

400

200
Tensile stresses
Tension [MPa]

in the edge
0
Compressive stresses
in the core
-200

-400

-600

-800
0.001 0.01 0.1 1 10
Time [sec]

Fig. 2.21 Thermal stresses during quenching of a cylinder with 100 mm Ø in water without plas-
tic deformation
2.3 Damage Examples—Construction 27

800

600

400

FS
200
Tensile stresses
Tension [MPa]

in the edge
0
Compressive stresses lg t
in the core
-200
FS

-400

-600

-800
0.001 0.01 0.1 1 10
Time [sec]

Fig. 2.22 Thermal stresses during quenching of a cylinder with 100 mm Ø in water with plastic
deformation

Risk of cracking

unfavorable favorable

Fig. 2.23 Example of less abrupt cross-sectional transitions [24]

2.3 Damage Examples—Construction

Example 1—Cracks at the cross-section transition of a load-bearing axle

Component: Load-bearing axle with different diameters

Material: 42CrMoS4, Material No. 1.7227
Error pattern: Cracks at the cross-section transition Ø 530 mm to Ø 340 mm
Heat treatment: Tempering with quenching in a polymer solution
28 2 Design Influence Area

A load-bearing axle made of the heat-treatable steel 42CrMoS4, 1.7227, developed

cracks during quenching and tempering at the transition, Fig. 2.24.
The heat treatment plant had austenitized the load-bearing axle in a directly gas-
heated furnace at 860 °C with a total holding time of 4 h and quenched it in a quenching
bath with a 6 % polymer solution. During the tempering treatment at 580 °C for 8 h, the
furnace was slowly heated from 100 °C and after the tempering treatment, cooling took
place in still air.
The load-bearing axle had cracked in the radius of the transition from the Ø 530 mm
to the Ø 340 mm shaft shoulder. The crack extended to the opposite end face and had a
circumferential extent of approx. 220–230°.
The used material 42CrMoS4, Material No. 1.7227, was in accordance with its chem-
ical composition standard compliant according to DIN EN 10083-3:2007-01 [27] and
without complaint. The copper and nickel contents of 0.08 % Cu and 0.10 % Ni indi-
cated an electrosteel production. However, both elements were significantly below the
limit values in DIN 10020:2000 [28]. All other trace elements were inconspicuous.
The hardenability calculated from the chemical composition was slightly below the
middle of the standard scatter band, Fig. 2.25.
The radius in the cracked transition was at r = 0.9 mm. A fine acicular martensitic
quenched and tempered microstructure was present. The surface was scaled, but metal-
lographically only a very slight decarburization could be detected, Fig. 2.26.
Copper-nickel segregations were visible up to a depth of about 31 μm at the near-
surface grain boundaries. The fracture surface showed no decarburization and no copper-
nickel segregation, Fig. 2.27.
The hardness at the surface was 343 HV 10, which corresponds to a tensile strength
of 1078 MPa converted according to DIN EN ISO 18265:2014-02 (Table B2).
The absence of copper-nickel segregation at the crack surface proved that the crack
could not have occurred during heating to the hardening temperature. A heating stress
crack during tempering was also ruled out due to the low furnace start temperature of

Fig. 2.24 Crack formation after quenching and tempering on a load-bearing axle [26]
2.3 Damage Examples—Construction 29

Material 42CrMoS4, 1.7227

70 CrMo-alloyed heat-
treatable steels
65 Chemical
analysis
60 C: 0.39 %
Si: 0.29 %
Mn: 0.67 %
+HH P: 0.010 %
55
S: 0.024 %
Cr: 1.05 %
50 Mo: 0.20 %
Hardness [HRC]

+H Ni: 0.08 %
+HH Al: 0.24 %
45 Cu: 0.10 %
N: 0.0090%
+H
40 [mm] [HRG]
+HL 1 1.0 56.4
+H 2 3.0 55.2
35 3 5.0 54.7
+HL 4 7.0 53.9
5 9.0 53.0
30 6 11.0 51.4
7 13.0 49.1
8 15.0 47.1
25 9 20.0 42.7
10 25.0 39.4
11 30.0 37.5
20 12 35.0 36.4
13 40.0 35.1
0 5 10 15 20 25 30 35 40 45 50

Fig. 2.25 Hardenability calculated according to SEP 1664 and standard scatter band [29]

Fig. 2.26 Results of the metallographic examination; (a) radius measurement; (b) microstructure [26]

100 °C. Rather, the investigations carried out suggested that a quenching stress crack was
present.
The primary cause of damage was the tightly measured radius of r = 0.9 mm, which
caused a significant notch effect. However, this seemingly uncomplicated component
also had secondary reinforcing influences.
Austenitization took place in a directly gas-heated furnace with an excess of air,
which led to copper-nickel segregation on the surface and thus to a weakening of the
near-surface grain boundaries [30].
30 2 Design Influence Area

Fig. 2.27 Results of the metallographic examination; (a) Copper-nickel segregations at near-sur-
face grain boundaries, (b) Fracture surface without copper-nickel segregation [26]

The heat-treatable steel 42CrMoS4 has limited hardenability. Therefore, a polymer

quenching agent with a low concentration and thus increased quenching performance
was used for the present dimension. A more abrupt quenching leads to larger temperature
differences and thus to larger thermal stresses.
The three previously described measurable facts were overlaid by a non-quantifia-
ble but still important factor. Should this increased risk of cracking have already been
noticed during work preparation in the hardening shop? Heat treatment-appropriate
design also includes shape-appropriate heat treatment [1]. A timely hint of an existing
risk of cracking before carrying out the quenching and tempering treatment is an impor-
tant task of work preparation.
However, it is recommended for the design to make the radii larger in order to reduce
the notch effect and thus the risk of cracking.

Example 2—Fatigue fracture on a truck crankshaft after overhaul

Component: Truck crankshaft

Material: 42CrMo4, Material No. 1.7225
Error pattern: Fatigue fracture after engine overhaul
Heat treatment: Nitridingas part of the engine overhaul

After an engine mileage of approx. 1,000,000 km, numerous truck engines were over-
hauled. The engines were not repaired in contract workshops of the various vehicle man-
ufacturers, but in a company that specialized in these repair works. The designers in this
company had learned that the fatigue strength of crankshafts is increased by a nitrid-
ing treatment. Therefore, as part of the overhaul, the crankshafts, which were made of
the material 42CrMo4, were nitrided. A nitriding hardening depth of 0.2 + 0.1 mm was
specified.
2.3 Damage Examples—Construction 31

It is not precisely known how many crankshafts were nitrided. However, it is known
that after installation and a mileage of 20,000–30,000 km, a total of 160 trucks had failed
due to crankshaft damage. In all cases, there was a rapidly propagating fatigue fracture,
Fig. 2.28.
The metallographic examination confirmed a flawless and specification-compliant
formation of the nitriding layer with a compound layer thickness of approx. 6 μm and
a nitriding hardening depth of 0.24 mm, Fig. 2.29. This diagram also schematically
includes a hardness profile after inductive surface hardening. The actual hardness pro-
file before nitriding could no longer be determined, but it should be very similar to the
assumed profile.
All standard examinations (analysis, purity level, surface quality, etc.) were incon-
spicuous and not relevant to the damage. Therefore, a separate documentation was omit-
ted.
From the investigation results, it was evident as the primary cause of damage that it
was not considered or recognized that the crankshafts in their original condition were
inductively hardened in the crankpins. Nitriding did create a nitriding layer with higher
hardness in the near-surface area, but the hardening profile of the inductive hardening
was completely equalized.
The maximum surface pressure is not caused at the immediate surface, but at a certain
depth. The inductive surface layer had a constant hardness up to a depth of 2 mm and
could withstand the occurring stress tensions. Nitriding acted on the inductive surface
hardening like a high tempering and at a very short distance from the surface, the hard-
ness achieved during inductive hardening was undercut. The rapidly decreasing hardness

Fig. 2.28 Fatigue fracture on a nitrided crankshaft [32]

32 2 Design Influence Area

800

700

600
induction hardened

500
Hardness [HV 1]

400

nitrated
300

200

100

0
0.0 1.0 2.0 3.0 4.0 5.0
Edge distance [mm]

Fig. 2.29 Microstructure and hardness profile after nitriding and schematic hardness profile after
inductive hardening [29]

of the nitriding hardness profile was no longer sufficient to withstand the stress tensions,
which is why a crack occurred and subsequently a rapidly propagating fatigue fracture.

Example 3—Surface defects on a worm drive after test stand trials

Component: Worm drive

Material: 100Cr6, Material No. 1.3505
Error pattern: Surface defects after test stand trials
Heat treatment: Nitrocarburizing in the +AC state

The wear-resistant effect of nitrocarburizing was well known to the designers of this
component. However, the long-standing choice of using a quenched and tempered
steel 42CrMo4 was very cost-intensive. The raw material could be obtained in the pre-
quenched and tempered state, but the complex machining due to the large chip volume
in worm drives offset the cost advantage. A quenching and tempering treatment after soft
machining was also ruled out for cost reasons, especially since these delicate compo-
nents were not without problems in terms of dimensional stability.
The design therefore decided to experimentally use a material with a higher carbide
content and to forego pre-quenching and tempering. The choice fell on the roller bearing
steel 100Cr6 in the annealed state +AC.
2.3 Damage Examples—Construction 33

After 10,000 load cycles on a test stand, an intermediate inspection was carried out
and surface damage was detected, the cause of which was to be determined, Fig. 2.30.
A typical annealed microstructure (GKZ) for this type of rolling bearing steel was
present. The hardness was 225 HV 10. In the area of the macroscopically recogniz-
able surface damage, it was seen that the compound layer was pressed into the soft base
matrix due to high point-like surface pressures, Fig. 2.31. In other places on the worm
surface, the compound layer had chipped off and these broken particles probably caused
the indentation into the soft base material during rolling processes.
The determination of the nitriding hardness depth made it clear that there was only a
very slight increase in hardness behind the compound layer.
The chosen design (material 100Cr6, GKZ-annealed + nitrocarburized) can be
described as unusual. Such a combination can be used without disadvantage only if
low surface pressures can be guaranteed and only adhesive wear (cold welding) is to be
avoided. However, if high surface pressures occur, for example due to particle indenta-
tions, the low hardness of the base material is not sufficient to support the compound
layer. The resulting damage mechanism is colloquially referred to as the eggshell effect.
To complicate matters, high amounts of chromium are already stably bound in the
carbides of the annealed microstructure. This greatly reduces the amount of nitride form-
ers still available, leading to a very slight increase in hardness in the diffusion layer.

Fig. 2.30 Surface damage on a worm drive after 10,000 load cycles
34 2 Design Influence Area

Fig. 2.31 Compound layer pressed into the base material

Example 4—Breakage of a shrink disc after a few load cycles

Component: Shrink disc

Material: 42CrMo4, Material No. 1.7225
Error pattern: Breakage after a few load cycles
Heat treatment: Quenching and tempering

The mechanical properties of the heat-treatable steels are defined in the material stand-
ards for the respective cross-section. However, it must be noted that for all cross-sec-
tions >25 mm Ø, the test position is 12.5 mm below the surface. Therefore, lower core
strengths can be expected for larger cross-sections. The decrease in strength properties in
the core is also determined by the hardenability of the material used. Steels with lower
hardenability show a greater difference between the standardized test position and the
core.
This relationship is clearly illustrated in the example from Fig. 2.32. A shrink disc
was made from a quenched and tempered ring of the heat-treatable steel 42CrMo4
+QT. Initially, a hot-rolled ring with a cross-section of 340 × 300 mm and an outer
diameter of 1650 mm was produced. After proper tempering with liquid quenching, the
standard-compliant testing was carried out 12.5 mm below the surface. The thus treated
and quenched and tempered ring was then mechanically processed and the cross-section
2.3 Damage Examples—Construction 35

Standard-compliant Heat
test position for treatment
testing at the cross-section
supplier's premises

Crack start
position

Finished surface after

tempering

Fig. 2.32 Fracture surface of a shrink disc

was reduced to 260 × 230 mm. In addition, for functional reasons, a surface was worked
out on one end face, as shown in Fig. 2.32.
The frictional connection to the shaft was made and released several times. After a
few loads and unloads, a spontaneous break occurred during the establishment of the
frictional connection.
The significant heat treatment diameter calculated from the ring geometry resulted in
a value of Ø 375 mm. For this dimension, a 0.2 % yield limit of ≥460 MPa and a tensile
strength of Rm = 690–840 MPa are specified in SEW 550 [33]. These values were even
significantly exceeded at the marked and typical test site after quenching and tempering
with Rp0,2 = 1010 MPa and Rm = 1110 MPa with a high impact strength of Av = 72 J.
At the crack start position, however, the 0.2 % yield limit was only 690 MPa and the
tensile strength 895 MPa. The impact strength was significantly below the test value with
Av = 21 J.
However, this drop in strength was not expected in the design of the load capacity.
Rather, it was assumed that the properties from the test certificate are present throughout
the cross-section. The undershooting of the yield limit by 30 % compared to the design
value in combination with the low impact toughness has used up the safety factor based
on 1010 MPa and resulted in a violent break.
The sharp drop in impact toughness was caused by a lamellar mixed microstructure of
ferrite-pearlite-bainite and tempered martensite. To quantify the microstructural inhomo-
geneity, an “HV 0.2” hardness scan was carried out over a distance of 2 mm perpendicu-
36 2 Design Influence Area

lar to the segregation-related line structure, analogous to ASTM A534 [31], Fig. 2.33.
There was a very large hardness difference of 172 HV 0.2 between the maximum and
minimum hardness.
In SEW 550, the material 42CrMo4 is specified with a nickel content of max. 0.60 %,
while the used steel melt only contained 0.19 % nickel. The utilization of the maximum
nickel content would have resulted in significantly improved hardenability.

Example 5—Insufficient surface hardness on a spindle

Component: Spindle ∅ 63 mm × 1280 mm

Material: 34CrAlNi7-10, Material No. 1.8550
Error pattern: Insufficient surface hardness after nitriding
Heat treatment: Purchase quenched and tempered and nitrideafter manufacturing

For the subsequent damage case to be described, the classification into a clearly recog-
nizable area of responsibility is difficult, as new insights were gained for all involved
departments with the damage case. However, since a drawing change was required, the
damage case was described in the area of responsibility of design.
Pre-quenched and tempered rods made of nitriding steel 34CrAlNi7-10, 1.8550, were
used to manufacture spindles. The order was placed based on the drawing specification
with a tensile strength of 850–1050 MPa. This range of tensile strength corresponded to

800

700
HVmax = 174 HV 0.2
600

500
Hardness [HV 0.2]

400

300

200

100

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Distance [mm]

Fig. 2.33 Inhomogeneity assessment using HV 0.2 hardness scan

2.3 Damage Examples—Construction 37

the delivery condition according to DIN EN 10085:2001-07 for the product thickness of
∅ 40–100 mm. The 3.1 test certificate showed the results of the tensile test with a tensile
strength of 857 MPa. The tensile strength was therefore at the lower limit, but within the
target specification.
After mechanical processing, the spindles were gas nitrided at 500 °C for 84 hours,
and a surface hardness of at least 900 HV 10 was to be achieved. However, only a sur-
face hardness of 600 HV 10 was measured. The compound layer was mechanically
removed before the measurement. Since a process error during nitriding was suspected,
a repeat treatment was carried out with the same parameters. The surface hardness could
not be increased, only the nitriding hardness depth was significantly increased.
When determining the nitriding hardness depth, a maximum hardness of 648 HV 0.5
was measured in the diffusion layer and the core hardness was 257 HV 0.5, Fig. 2.34.
Due to the very low core hardness, a tensile test sample was taken from the spindle and
tested at room temperature. The tensile strength of 825 MPa was below the required min-
imum of 850 MPa, although the deviation from the 3.1 test certificate was still within
the range of experimental scatter. Nevertheless, there was clearly a tensile strength at the
lowest specified limit, which indicated a high tempering temperature.
With increasing tempering temperature, the nitride-forming elements are increas-
ingly bound in the carbides. As a result, there are not enough nitride-forming alloy ele-
ments available for the subsequent nitriding process, and the maximum hardness is not
achieved.
In the next step of the investigation, several samples were taken from the spindle,
which had already been nitrided twice, hardened, and tempered at 620 °C or 650 °C.
The newly quenched and tempered samples were then nitrided together with a produc-
tion batch at 500 °C for 84 hours. The hardness profile curve of the sample tempered and
nitrided at 620 °C showed a maximum edge hardness of 1091 HV 0.5 and a core hard-
ness of 320 HV 0.5, Fig. 2.35. At the higher tempering temperature of 650 °C, the same
edge hardness was measured and the core hardness was 300 HV 0.5.
The resulting damage was not insignificant, as the finished spindles could not be re-
quenched and tempered for dimensional reasons. These spindles had been made from
this steel with the same order specification for years without any comparable behavior
being observed. It was therefore recommended to adjust the order specification with a
higher minimum tensile strength.

Example 6—Crack formation on a hollow shaft after oil quenching

Component: Hollow shaft∅ 625 mm × 2430 mm

Material: 42CrMo4, Material No. 1.7225
Error pattern: Crack formation during oil quenching
Heat treatment: Quenching and tempering Tempering
38 2 Design Influence Area

900

800
NHD310 = 0.87 mm
700

600
Hardness [HV 0.5]

500

400

300

200

100

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Edge distance [mm]

Fig. 2.34 Nitriding hardness profile after nitriding twice at 500 °C for 84 hours

1200

1000

NHD370 = 0.39 mm
800
Hardness [HV 0.5]

600

400

200

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Edge distance [mm]

Fig. 2.35 Nitriding hardness profile after re-quenching and tempering a sample tempered at 620 °C
2.3 Damage Examples—Construction 39

The stress-increasing effect of notches is one of the most common design-related

causes of failure when cracking occurs during quenching. This is particularly true for
components that must be abruptly quenched due to moderate hardenability. This cause
of failure was also responsible for the cracking in the hollow shaft depicted in Fig. 2.36.
The dye penetrant inspection localizes the crack position. On the sawn-off segment, the
crack origin was located at the sharp-edged end of the borehole, Fig. 2.37. The hollow
shaft was austenitized at 860 °C and quenched in oil.
For the used heat-treatable steel 42CrMo4, 1.7725, the supplier’s certificate with the
chemical composition was available, Table 2.5. From this, the hardenability was calcu-
lated according to the equations from SEP 1664:2004-06, and the values were plotted
into the standard scatter band, Fig. 2.38. Accordingly, this analysis corresponded to the
central position of the hardenability class +HL.
For the present dimension, this was a very low hardenability, which led to a mixed
microstructure, Fig. 2.39. In the positively segregated areas, a mixed microstructure of
bainite and pearlite was present, while ferrite-pearlite was formed in the negatively seg-
regated areas.
The mechanical properties were determined at the inner diameter in the tensile test
and notch impact bending test, each at room temperature, and compared with the infor-
mation in the supplier’s certificate, Table 2.6. The notch impact bending samples were
taken tangentially. The certificate values showed a tensile strength at the lower end of the
normative target specification. The notch impact toughness values were much higher in
the forging test than in the determination in the context of the damage case investigation.
This was not only due to the lower tensile strength. It should be reminded again that the
testing for the supplier certificates according to standard always takes place 12.5 mm
below the surface and a clear gradient of toughness with increasing surface distance is to
be expected.

Saw cut

Fig. 2.36 Dye penetrant inspection on the cracked hollow shaft and marking for the saw cut
40 2 Design Influence Area

Fig. 2.37 Crack origin at the sharp-edged end of the borehole

Material 42CrMo4, 1.7225 | hardening temperature 850 °C

70

65

60
+HH
55

+HH
50
Hardness [HRC]

+H
45
+H
40
+HL
35 +H
+HL
30

25

20
0 5 10 15 20 25 30 35 40 45 50
Face distance [mm]

Fig. 2.38 Hardenability scatter band of the heat-treatable steel 42CrMo4 and calculated harden-
ability of the used steel melt
2.3 Damage Examples—Construction 41

Table 2.5  Information on the chemical composition in the supplier’s certificate

Target analysis C Si Mn P S Cr Ni Mo
42CrMo4, 1.7225 % % % % % % % %
acc. to DIN EN 0.38 Max. 0.60 Max. Max. 0.90 – 0.15
10250-3:1999-12
0.45 0.40 0.90 0.035 0.035 1.20 – 0.30
Certificate information 0.38 0.30 0.82 0.012 0.003 0.96 0.16 0.18

Fig. 2.39 Structural formations in the area of the bore surface; (a) bainite-pearlite microstructure
in positively segregated areas, (b) ferrite-pearlite microstructure in negatively segregated areas

Table 2.6  Results of the tensile tests and notch impact bending tests
Tensile tests Notch impact bending tests
Rp0,2 Rm A Z 1st value 2nd value 3rd value Average
[MPa] [MPa] [ %] [ %] [J] [J] [J] value
[J]
Target Min. 500 750–900 Min. 14 – 38
values
Certificate 540 750 15 55 60 68 98 75
values
Inside 551 876 16 37 27 27 26 27

Example 7—Cracking of shift rails during straightening

Component: Shift rails

Material: C35, Material No. 1.0501
Error pattern: Cracking during straightening
Heat treatment: Quenching and tempering to a hardness >70 HRA
42 2 Design Influence Area

Fig. 2.40 Cracking of quenched and tempered shift rails made of C35

Shift rails were made from the unalloyed heat-treatable steel C35, 1.0501, in the
annealed state by punching and cold forming. Subsequently, these components were
quenched and tempered in an external hardening shop with the target specification
>70 HRA (~367 HV). Due to the complex shape and the manufacturing history, a certain
distortion was unavoidable, which is why the shift rails were calibrated by a straighten-
ing process at the manufacturer after quenching and tempering. Numerous batches were
straightened completely without problems and without cracking in the past. A sample
from a previous heat treatment batch was also provided as a reference. However, the cur-
rent batch showed almost 100 % cracked parts, Fig. 2.40, as a crack test was only carried
out after all parts had been calibrated.
Comparative investigations were carried out on the damaged part and the reference
sample. The chemical composition and the purity of the steels used showed no abnor-
malities. The heat treatment was carried out according to the specifications, but the
results were very noticeable.
The reference part was already significantly above the minimum target hardness of
70.0 HRA (~367 HV 10) with 76.0 HRA (~510 HV 10). The hardness of the damaged
part was even higher at 78.1 HRA (~577 HV 10). Tempering attempts were carried out
on segments of this damaged part to determine the tempering temperature.
The evaluation of the results after re-tempering at 150 °C, 200 °C, 250 °C and 300 °C
showed that a significant drop in hardness was only noticeable at 300 °C. From this it
could be concluded that the shift rails, which formed cracks during straightening, were
tempered at 250 °C, Fig. 2.41.
This damage example shows that the specification of only a minimum hardness can
lead to scrap. A properly specified hardness range with minimum and maximum values
would have prevented the scrap of this batch. The hardening shop carried out the quench-
ing and tempering process according to the specifications, although it would have been
advisable to ask the manufacturer and previous batches were tempered at a higher
References 43

800
Hardness of the
Delivery status
700

600
Hardness [HV 1]

500

400

300

200
0 100 200 300
Restart temperature [°C]

Fig. 2.41 Hardness results after various re-tempering treatments

t­emperature. The employee during the straightening process finally relied on the fact that
all shift rails of the previously carried out heat treatments could be straightened without
cracking. A check after a few parts would also have been desirable here, especially since
the cracks were visible to the naked eye. A further tempering at a higher temperature
would then have been possible.
This example also shows a chain of several influencing factors that ultimately led to
scrap parts.

References

1. Prietzel, Karl-Otto und Horst Lichtenberg: Wärmebehandlungsgerechtes Gestalten und

gestaltgerechtes Wärmebehandeln von Bauteilen des Maschinen-, Geräte und Anlagenbaus.
DER WÄRMEBEHANDLUNGSMARKT (2) 1995, S. 6–9
2. DIN 17022-3:1989-04: Wärmebehandlung von Eisenwerkstoffen; Verfahren der Wärmebehan-
dlung; Einsatzhärten
3. DIN ISO 15787:2018-08: Technische Produktinformation—Wärmebehandelte Teile aus
Eisenwerkstoffen—Darstellung und Angaben
4. DIN ISO 15787:2010-01: Technische Produktinformation—Wärmebehandelte Teile aus
Eisenwerkstoffen—Darstellung und Angaben
5. DIN 6773:2001-04: Wärmebehandlung von Eisenwerkstoffen—Darstellung und Angaben
wärmebehandelter Teile in Zeichnungen
44 2 Design Influence Area

6. Stahl-Eisen-Prüfblatt 1520:1998-09: Mikroskopische Prüfung der Carbidausbildung in Stählen

mit Bildreihen
7. DIN EN ISO 6508-1:2016-12: Metallische Werkstoffe—Härteprüfung nach Rockwell—Teil 1:
Prüfverfahren
8. DIN EN ISO 6507-1:2018-07: Metallische Werkstoffe—Härteprüfung nach Vickers—Teil 1:
Prüfverfahren
9. DIN EN ISO 18265:2014-02: Metallische Werkstoffe—Umwertung von Härtewerten
10. Schruff, I.: Weitere Erkenntnisse zu den Problemen des Umwertens von Härte und Zugfestig-
keit bei Werkzeugstählen nach DIN EN ISO 18265—Bericht aus der Arbeit des Unterauss-
chusses für Werkzeugstähle im Stahlinstitut VDEh HTM J. Heat Treatm. Mat. 63 (2008) 4, S.
201–206
11. DIN EN ISO 6506-2:2019-03: Metallische Werkstoffe—Härteprüfung nach Brinell—Teil 2:
Überprüfung und Kalibrierung der Prüfmaschinen
12. DIN EN ISO 6507-2:2018-07: Metallische Werkstoffe—Härteprüfung nach Vickers—Teil 2:
Überprüfung und Kalibrierung der Prüfmaschinen
13. DIN EN ISO 6508-2:2015-06: Metallische Werkstoffe—Härteprüfung nach Rockwell—Teil 2:
Überprüfung und Kalibrierung der Prüfmaschinen und Eindringkörper
14. Wärmebehandlung von Stahl—Randschichthärten. Merkblatt 236. Stahl-Informations-Zen-
trum, Düsseldorf
15. Fiedler, Andreas: Einfluss des Werkstoffzustands auf das Wärmebehandlungsergebnis beim
induktiven Randschichthärten. Dr.-Ing. Diss. 2012, Darmstadt
16. DIN EN ISO 2639:2003-04: Bestimmung und Prüfung der Einsatzhärtungstiefe
17. ISO 4545-1:2017-12: Metallische Werkstoffe—Härteprüfung nach Knoop—Teil 1: Prüfver-
fahren
18. ISO 6336-5:2016-08: Tragfähigkeitsberechnung von gerad- und schrägverzahnten Stirn-
rädern—Teil 5: Festigkeit und Werkstoffqualitäten
19. Sommer, P. und K. Rahman: Untersuchung der modul- und werkstoffabhängigen Einsatzhär-
tungstiefen sowie deren Korrelation zu Mitfahrproben. Abschlussbericht Technologie Forum
Werkstoff & Wärme 2020
20. DIN 50190-3:1979-03: Härtetiefe wärmebehandelter Teile; Ermittlung der Nitrierhärtetiefe
21. DIN 30902:2016-12: Wärmebehandlung von Eisenwerkstoffen—Lichtmikroskopische
Bestimmung der Dicke und Porosität der Verbindungsschichten nitrierter und nitrocarburierter
Werkstücke
22. Riesbeck, Sabine und Ulrich Baudis: Universelle Präparationsmethode für randscharfe Schliff-
proben zur Ermittlung der Verbindungsschichten an nitrierten und nitrocarburierten Werk-
stücken. Der Wärmebehandlungsmarkt (14) 2007-02. Seite 5–8
23. DIN 17022-1:1994-10: Wärmebehandlung von Eisenwerkstoffen. Verfahren der Wärmebehan-
dlung. Teil 1: Härten, Bainitisieren, Anlassen und Vergüten von Bauteilen
24. DIN 17022-2:1986-06: Wärmebehandlung von Eisenwerkstoffen. Verfahren der Wärmebehan-
dlung. Härten und Anlassen von Werkzeugen
25. Liscic, B., H.M. Tensi und W. Luty: Theory and Technology of Quenching. Springer Verlag
Berlin, Heidelberg, New York 1992
26. Sommer, P.: Schadensfälle an wärmebehandelten Bauteilen—nicht immer ein Wärmebehand-
lungsfehler. Praktische Metallographie 56 (2019) S. 188–206
27. DIN EN 10083-3:2007-01; Vergütungsstähle—Teil 3: Technische Lieferbedingungen für legi-
erte Stähle (zurückgezogen)
28. DIN EN 10020:2000-07: Begriffsbestimmung für die Einteilung der Stähle
References 45

29. Stahl-Eisen-Prüfblatt 1664:2004-06: Ermittlung von Formeln durch multiple Regression zur
Berechnung der Härtbarkeit im Stirnabschreckversuch aus der chemischen Zusammensetzung
von Stählen
30. Sommer, P; Kisters, P. und Brenke, B.: Untersuchung des Einflusses von Kupfer-Nickel-
Segregation auf die Betriebsfestigkeit vergüteter Stähle. DER WÄRMEBEHANDLUNGS-
MARKT 21 (2014) Seite 5–9
31. ASTM A534:2017: Standard Specification for Carburizing Steels for Anti-Friction Bearings
32. Sommer, P.: Wärmebehandlung—Fehler, Schäden und Ursachen. HTM J. Heat Treatm. Mat.
70 (2015) 2, S. 97–107
33. SEW 550:1976-08: Stähle für größere Schmiedestücke; Gütevorschriften
Material Purchasing Influence Area
3

Contents

3.1 Case-hardening Steels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

3.2 Quenched and Tempered Steels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.3 Nitriding Steels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.4 Martensitic Stainless Steels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.5 Precipitation Hardening Steels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.6 Tool Steels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.7 Damage Examples—Material Purchase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

The steel specified by the designer in the drawing must be purchased by the purchas-
ing department in the appropriate dimensions, quantity, and delivery time. If there is no
specific order specification for the steel to be procured, the purchaser will typically refer
to the valid material and product standards and indicate these in the order as a refer-
ence, Fig. 3.1. In addition, many delivery specifications have been created, which in most
cases define the property specifications more narrowly than the national and international
material standards.
It is very common in the purchasing department to assume that the mechanical
properties and quality characteristics required for their own application are already
sufficiently described by reference to a valid and current material standard. Since this
assumption is incorrect in most cases, the information in the material standards will be
described and evaluated in more detail below.

© The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, 47

part of Springer Nature 2024
P. Sommer, Mistakes Before, During and After Heat Treatment of Steel,
https://doi.org/10.1007/978-3-658-44160-9_3
48 3 Material Purchasing Influence Area

Material standards

Superordinate steel groups

Product-specific standards Case-hardening steels

Cold band Quenched and tempered steels

Wire rod Tool steels

Bright steels Stainless steels

Cold heading steels Nitriding steels

Rolling bearing steels ...

Free-cutting steels Company-specific delivery variants

Steel tubes ZF, VW, Siemens...

... ...

Fig. 3.1 Material and product standards as well as delivery specifications

3.1 Case-hardening Steels

The European DIN EN 10084 [1] has been replaced by the international DIN EN ISO
683-3 [2]. The content differences between both standards are marginal. However, it
should be noted that the material numbers often used in German language are no longer
included in the international standard. This standard describes a total of 34 different
unalloyed and alloyed steel grades. Characteristic for this group of steels are the low car-
bon contents in the delivery condition from 0.07% to 0.31%.
When ordering according to this standard, the following order details are mandatory:

• the quantity to be delivered

• the designation of the product form
• dimensional specifications
• reference to DIN EN ISO 683-3
• designation of the steel grade
• specification of a factory certificate

If no further requirements are specified, the standard states that delivery must be made
according to the basic requirements listed above. In the following Sect. 5.2 of this stand-
ard, options or special requirements are then named. From this list, the options that are
relevant for heat treatment and the resulting properties are described in more detail.

Heat treatment condition

The basic requirements include an untreated heat treatment condition (+U), which is pos-
sible for all product forms and case-hardening steels. After hot forming, no additional
heat treatment is carried out on these steels. Any other desired or required heat treatment
3.1 Case-hardening Steels 49

Fig. 3.2 Fine-grained and homogeneous ferrite-pearlite microstructure in the heat treatment con-
dition +N

condition, e.g. treated to ferrite and pearlite (+FP) or normalized (+N), must be explicitly
stated. A complete list of possible heat treatment conditions can be found in DIN EN
ISO 683-3 in Table 1.

Fig. 3.2 shows an optimally formed ferrite-pearlite microstructure in the heat treatment
condition +N for the unalloyed case-hardening steel C15E.
However, it should be noted that not all heat treatment conditions listed in Table 1 of
DIN EN ISO 683-3 can be realized for all case-hardening steels contained in this stand-
ard. The standard expressly points this out.
For the very commonly used case-hardening steel 18CrNiMo7-6, the +N condition
is formally feasible, but the properties achieved are unsuitable for mechanical process-
ing. This case-hardening steel has such high hardenability that after austenitization with
subsequent air cooling, there is not a pure ferrite-pearlite microstructure, but very high
bainite proportions with significantly higher hardness are to be expected. The continu-
ous CCT diagram shows a calculated cooling curve for a rod dimension of ∅ 60 mm, Fig.
3.3 [3]. Economical machining is therefore not possible. In addition, mixed microstruc-
tures in the initial state lead to a higher tendency to distore during the final case harden-
ing.
To achieve a ferrite-pearlite microstructure, the heat treatment condition +FP (treated
to ferrite and pearlite) is therefore recommended. In this heat treatment, after austenitiza-
tion, cooling is as rapid as possible to a temperature of 630–650 °C with subsequent long
holding at this temperature, Fig. 3.4. The rapid cooling allows a high number of nuclei to
form, resulting in a fine-grained ferrite-pearlite microstructure. The holding time at the
isothermal transformation temperature must be chosen so that a transformation is com-
pletely finished.
However, this heat treatment process also has limits. For large dimensions, rapid cool-
ing to the isothermal transformation temperature is no longer possible for thermophysi-
cal reasons. This was explicitly pointed out for the first time in the current standard. It is
50 3 Material Purchasing Influence Area

1200

1100
18CrNiM7-6, 1.6587, bar 60 mm, austenitization 870 °C 20 min, air cooling
1000

900
Ac3 (825 °C)
800
Temperature [°C]

700 75 Ac1 (745 °C)

5 Ferrite 70
40 Perlite 25
600 Austenite 2

500
Bainite
Ms (395 °C) 3 90
400
90
300 Martensite
55
25
200

100
431 362 324 271 208 158
0
0.1 1 10 100 1000 10000 1E5 1E6
Time [sec]

Fig. 3.3 Continuous CCT diagram for the case-hardening steel 18CrNiMo7-6 with calculated
temperature profile during air cooling of a rod with 60 mm Ø

1000

800

600
Temperature [°C]

400

200

0
Time

Fig. 3.4 Schematic representation of a +FP heat treatment

3.1 Case-hardening Steels 51

explicitly stated: “For certain types of steel, a 100% ferritic-pearlitic structure cannot be
achieved despite adjusted heat treatment, e.g.: in the isothermal annealing of rods of the
steel grade 18CrNiMo7-6, an F+P+B structure is always obtained”
This addition was also the main difference for a standard revision after only a few
months.
For large dimensions, therefore, soft annealing (+A) or pre-hardening (+QT) is rec-
ommended after appropriate agreement.
In the case of the heat treatment condition +QT, the experience of the supplying plant
should be used to determine from which dimension a +QT treatment is recommended, as
this type of treatment is the most cost-intensive pretreatment. The heat treatment condi-
tion +QT is not explicitly mentioned in the standard, but it is stated that other heat treat-
ment conditions than those listed in Table 3.1 can be agreed upon at the time of inquiry
and order. In pre-hardening, a high tempering temperature is usually chosen to be able to
carry out economical mechanical processing.
The following images show microstructures of the case-hardening steel 18CrNiMo7-6
in different delivery conditions. In Fig. 3.5, two microstructural formations of this steel
in the heat treatment condition +A are shown. The microstructural condition in the left
part of the image is typical for larger dimensions and the condition “forging + air cooling
+ soft annealing”. There is a coarsened and completely globularly shaped soft-annealed

Table 3.1  Requirements for the degree of purity according to ISO 6336-5:20106-08 [6] Measure-
ment method: ISO 4967 Method A [7]
A B C D DS
Fine Thick Fine Thick Fine Thick Fine Thick –
MQ 3.0 3.0 2.5 1.5 2.5 1.5 2.0 1.5
ME 2.5 1.5 2.0 1.0 0.5 0.5 1.0 1.0 2.0

Fig. 3.5 Different microstructural formations in the heat treatment condition +A

52 3 Material Purchasing Influence Area

Fig. 3.6 Microstructure formation in the heat treatment condition +FP bar material 90 mm Ø

Fig. 3.7 Microstructure formation in the heat treatment condition +QT ring with 160 mm wall
thickness

microstructure. In the right part of the image, the carbides are very evenly and finely dis-
tributed.
Figure 3.6 shows the strongly lamellar ferrite-pearlite microstructure of a rolled bar
material with a 90 mm Ø in the heat treatment condition +FP. Due to a slightly extended
transformation duration, a certain globular formation of the carbides has already
occurred.
Figure 3.7 clearly shows that the microstructure formation after pre-hardening is very
homogeneous. The rolled ring with a wall thickness of 160 mm has globular carbides in
the ferritic matrix. This homogeneous microstructure condition across the entire cross-
section provides the best conditions for low-distortion case hardening.

Hardenability
For the unalloyed case-hardening steels, no hardenability specifications are defined in
the standard. In contrast, the alloyed case-hardening steels have the three hardenabilty
3.1 Case-hardening Steels 53

classes hardness classes+H (normal hardness), +HH (restricted hardness in the upper
scatter band) and +HL (restricted hardness in the lower scatter band). An order without
specifying the hardenability class is automatically assumed as +H. It is therefore advis-
able to consciously specify the hardenabilty class.

The graphical representation of the hardness ranges for the three hardenability classes
can be seen in the example of the case-hardening steel 16MnCr5 in Fig. 3.8.
The experimental determination of hardenabilty is carried out by the end quench test
(Jominy test) according to DIN EN ISO 642:2000-01 [4]. Since the effort for such a
proof of hardenabilty is very high, regression equations for determining the hardenabilty
from the chemical composition have been developed. The first publication was already
made in 1990 by the Association of German Steelmakers (VDEh) in a report volume and
later in the Steel-Iron Test Sheet 1664:2004-06 [5] (SEP). The possibility of calculating
the hardenabilty from a melt analysis was then also included in the material standardiza-
tion.
The steel producer is given the opportunity in DIN EN ISO 683-3 to determine
the hardenabilty of a specific melt analysis from available calculation formulas. The
calculation method is not bound to the SEP 1664, but is left to the manufacturer. This
means that their own regression formulas can also be used. From this it can be concluded

Material 16MnCr5, 1.7131

70

65

60

55

50
Hardness [HRC]

45

40

35 +HH
+HH
30
+H
+H +H
25
+HL +HL
20
0 5 10 15 20 25 30 35 40 45 50
Face distance [mm]

Fig. 3.8 Graphical representation of the three hardenability classes +H, +HH and +HL using the
example of the case-hardening steel 16MnCr5
54 3 Material Purchasing Influence Area

that information on hardenabilty in factory test certificates can come from different cal-
culation methods, which only allows comparability and verifiability to a limited extent.
The order should therefore clearly express whether an experimental determination
of hardenabilty is required or a calculation method may be used. If a calculation is per-
missible, the calculation formulas should be specified in order to compare and verify
results from different suppliers.
The two restricted hardenabilty classes +HL and +HH are not narrowly defined
enough for numerous company-specific applications, which is why narrower limits are
demanded in many delivery agreements.

Cleanliness
The content of non-metallic inclusions is an important quality feature. Here, micro-
scopic and macroscopic inclusions are to be distinguished. DIN EN ISO 683-3 points
out that high grade steels must have a certain microscopic purity. However, proof of the
content of non-metallic inclusions must be agreed separately. The test method and the
permissible limit values must also be agreed.

In the predecessor standard DIN EN 10084, it was still stipulated that a review
according to EN 10247 should be applied unless otherwise agreed. This determination
method with a limit value of KaR ≤ 40 μm2/mm2 for all types of inclusions has not yet
been established in practice. The accuracy of the specified limit value was also ques-
tioned.
A concrete statement can be found in ISO 6336 [6]. For the measurement method
according to ISO 4967 [7], the maximum inclusion sizes permissible for evaluation
according to method A are given in the following Table 3.1. A correlation between the
purity grade standard DIN 50602 [8], method K, which is still frequently used in Ger-
many but has been withdrawn for many years, and ISO 4967 method A is not possible.
Therefore, a separate evaluation of a once documented data set of the inclusion image
analysis is required.
In the numerous damage case investigations and material sampling, the microscopic
purity grade was only relatively rarely a primary cause of damage, although the purity
grade does indeed have an influence on the service life of dynamically stressed compo-
nents.
On the other hand, macro inclusions at functionally highly stressed points almost
always lead to a failure of the component. Corresponding damage examples are
described in Sect. 3.7. Macro inclusions exert a strong notch effect and are often the
starting point for a fatigue crack.
DIN EN ISO 683-3 also refers here to an inspection method and determination of
limit values to be agreed between both parties. Ultrasonic testing is practically the only
inspection method that can be used. When determining the maximum permissible equiv-
alent defect sizes, the component condition, the testing facilities and the tester qualifi-
cation must be taken into account. For example, a forged part with forging scale is to
3.1 Case-hardening Steels 55

be assessed completely differently in terms of testing technology than a mechanically

machined part after pre-hardening. A very complex, but technically best possible, ultra-
sonic test is the immersion technique test according to SEP 1927 [9]. Detailed investiga-
tion results on case-hardening steels have been published, for example, in [10].

Grain size
Due to the sometimes long carburizing times at high temperatures, the grain size in case-
hardening steels is of particular importance. If coarse grain formation occurs during
carburizing, this leads to a deterioration of both the static and dynamic mechanical prop-
erties. An example of numerous published investigation results is the fatigue strength
under fluctuating torsional stress from [11], where a reduction in fatigue strength of 40%
was measured, Fig. 3.9.

In DIN EN ISO 683-3, Sect. 7.4 describes that all case-hardening steels must have
a fine grain structure with an austenite grain size of 5 or finer when tested according
to ISO 643 [12]. As an alternative to metallographic determination, it is stated that fine
grain is usually achieved with a total aluminum content of at least 0.018%. In this case, a
metallographic examination is not necessary and the proportion of aluminum must then
be indicated in the test certificate.
The specification of a minimum aluminum content aims at the formation of fine
grain stabilizing aluminum nitrides, although the also required element nitrogen is nei-

1300
1200
Fine grain
1100

1000
Coarse grain
Torosion stress [MPa]

900

800

700

40 %
600
Material 20MnCr5+B (ZF7B)
EHT610 = 0.4 - 0.7 mm
Tumescent torsion Pu = 50
500
104 105 106
Number of oscillation cycles

Fig. 3.9 Influence of grain size on fatigue strength under fluctuating torsional stress [10]
56 3 Material Purchasing Influence Area

ther mentioned nor specified in this standard. An aluminum/nitrogen ratio of at least 2:1
is required. The fine grain stabilizing effect of aluminum nitrides has been extensively
investigated and published, e.g., [13]. The influence of the aluminum and nitrogen con-
tent on the grain size and other properties was described in [14] with a very informative
graphic, Fig. 3.10.
Therefore, too low aluminum or nitrogen contents lead to coarse grain formations.
However, both alloy elements are also limited by maximum contents. Too high alu-
minum contents can lead to a deterioration of the purity level, as excess aluminum can
form aluminum oxides due to its high affinity for oxygen. Excessive nitrogen contents,
on the other hand, lead to a significant reduction in toughness.
The minimum aluminum content of 0.018% described in DIN EN ISO 683-3 is not
yet within the optimal alloy window. Rather, the optimal aluminum content was given in
[14] as 0.025% to 0.035%. Taking into account the nitrogen content also listed of 0.0075
to 0.012%, this results in aluminum/nitrogen ratios of 3–4, which means a certain excess
of aluminum.
The method for determining grain size according to the McQuaid-Ehn method
described in DIN EN ISO 643 involves carburizing at 925 ± 10 °C for 6 hours in carbur-
izing powder followed by slow cooling. In this process, carbides precipitate at the aus-
tenite grain boundaries, which then allows for a metallographic determination of grain
size. However, the holding time of 6 hours at a temperature of 925 °C already indicates

0.020

Zero toughness
0.015
Nitrogen content [%]

Degree of purity
Coarse grain
not in order
0.010
Optimum
Al:N ratio

0.005

Coarse grain

0.000
0.000 0.010 0.020 0.030 0.040 0.050 0.060
Aluminum content [%] Source: Klenke et al, HTM 2005

Fig. 3.10 Graphical representation of the optimal alloy range for aluminum and nitrogen for fine
grain stabilization [14]
3.1 Case-hardening Steels 57

that relatively small case hardening depths are detected. The result of this test is consid-
ered satisfactory when 70% of the area is within the specified limits. An upper limit for
individual grains and a determination method are not specified.
For gearing parts, ISO 6336-2016-08 [6] allows a max. grain size of individual grains
of G = 3. The area to be tested is specified as 3 mm2.
The fine grain stabilizing effect is not exclusively readable from the chemical com-
position. The presence of sufficient aluminum and nitrogen contents is merely a basic
requirement. However, effectiveness is only given if the aluminum nitrides have a size of
30–50 nm and are evenly distributed [15]. This means that the fine grain stability is influ-
enced by the entire production chain [16].
A metallographic examination of such fine precipitate particles is not possible in
routine operation, as high-resolution electron microscopic examinations are required.
Instead, blind hardness tests with long annealing times at typical carburizing tempera-
tures of 920–950 °C are carried out to prove fine grain stability. The specific treatment
specifications and evaluation methods are defined in company-specific delivery regula-
tions and are based on the case hardening depths to be set.
In the case of local coarse grain formation or inhomogeneous grain size distribution,
a different result is achieved depending on the evaluation method, as exemplified in Fig.
3.11a. The line cut method results in an average value from the number of intersection

14 14

12 12
G5 or finer = 96.86 % G5 or finer = 56.79 %
G6 or finer = 94.42 % G6 or finer = 46.69 %
G7 or finer = 86.88 % G7 or finer = 32.02 %
10 10
Percentage of area [%]
Frequency [%]

8 8

6 6

4 4

2 2

0 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
c Grain size class according to DIN EN ISO 643 d Grain size class according to DIN EN ISO 643

Fig. 3.11 Different evaluation methods for determining grain size (a) Line cut method; (b)
Traced grain boundaries; (c) Grain class distribution; (d) Grain area distribution
58 3 Material Purchasing Influence Area

points and the total measurement length. In the example shown, an average grain diam-
eter of 37.2 μm was obtained, which corresponds to a grain size of G = 6.
For an evaluation of the grain area distribution, a clear grain boundary structure is
required, which is why the grain boundaries were traced, Fig. 3.11b. With an image
analysis software, all grain areas are then measured and sorted by their grain size class
and their area. In the histogram for the grain number distribution, the local coarse grain
formation is not visible, Fig. 3.11c. Only the histogram with the grain area distribution
shows the extent of the inhomogeneity, Fig. 3.11d. However, individual coarse grains can
significantly reduce the fatigue strength, see Fig. 3.9.
For case hardening depths of 3 to 5 mm and carburizing times of over 100 hours, reli-
able results cannot be achieved with the McQuaid-Ehn method specified in the standard.
In the recent past, there have therefore been extensive investigations to shift grain growth
to higher temperatures through metallurgical measures and suitable heat treatments
before the carburizing treatment [17, 18]. The focus was on the formation of additional
hard precipitates, which only dissolve at high temperatures and thus hinder grain growth
at usual carburizing temperatures and long carburizing times. In addition to aluminum
nitride (AlN), the precipitates titanium nitride (TiN), niobium carbonitride (NbCN), and
vanadium nitride (VN) are of particular importance.

Steel selection and Treatment recommendations

Table D.1 in the case-hardening steel standard DIN EN ISO 683-3 provides very useful
information, especially for designers. The “Classification of steel grades according to the
minimum tensile strength depending on the diameter after case hardening and tempering
at 200 °C” listed there is also applicable for any heat treatment company to assess the
feasibility of set requirements. For a minimum tensile strength of e.g. 1000 MPa, the fol-
lowing steels are listed, Table 3.2.

Another recommendation for the required heat treatment can also be found in the
same standard. Case-hardening steels are carburized, hardened and tempered after
mechanical manufacturing in a near-final dimension state to achieve a case-hardening
depth. The temperatures recommended for these treatment processes are listed in DIN
EN ISO 683-3 in Table 8. Also included in this table is the austenitizing temperature for

Table 3.2  Classification of steel grades according to the minimum tensile strength according to
DIN EN ISO 683-3
Rm d ≤ 16 mm 16 mm < d ≤ 40 mm 40 mm < d ≤ 100 mm
MPa
min
1000 15NiCr13, 16MnCr5, 17NiCrMo6-4
16MnCrS5, 16MnCrB5,
16NiCr4, 16NiCrS4
3.2 Quenched and Tempered Steels 59

the end quench test. For the frequently used case-hardening steel 16MnCr5 and variants,
Table 3.3 shows an example of implementation.
For gas carburizing in atmosphere furnaces, the temperature of 950 °C is only
exceeded in exceptional cases. On the other hand, carburizing treatments in vacuum fur-
naces can also encounter higher temperatures up to 1050 °C.
The DIN EN ISO 683-3 is an important international standard for the binding descrip-
tion of properties in the delivery state, as well as recommendations for heat treatments
and information on expected properties after case hardening. Beyond these binding
delivery properties, specifications are defined in large numbers by company-specific
delivery conditions. However, it should be reminded that in each individual case, a
requirement deviating from the standard must be individually requested or agreed upon
at the delivery plant. It also needs to be clarified which form of the test certificate is
required. Specific tests cause additional costs, which can lead to higher material prices.

3.2 Quenched and Tempered Steels

The European standards for quenched and tempered steels DIN EN 10083-1 to 3 [19–21]
have also been replaced by the international DIN EN ISO standards 683-1 and -2 [22,
23]. Part 1 of this standard describes the unalloyed and part 2 the alloyed quenched and
tempered steels. The carbon contents of this steel group range from 0.22 to 0.65%, fol-
lowing on from the case-hardening steels described earlier.
In both parts of this standard, the binding order specifications differ only by an exten-
sion for the alloyed quenched and tempered steels:

• the quantity to be delivered

• the designation of the product form
• dimensional specifications
• reference to DIN EN ISO 683-1 or -2
• designation of the steel grade

Table 3.3  Conditions for the heat treatment of test bars and the treatment of steels according to
DIN EN ISO 683-3
Short name End quench Carburizing Direct or Double hardening Tempering
test aus- temperature single hard- Core Edge °C
tenitizing °C ening tem- hardening hardening
temperature perature temperature temperature
°C °C °C °C
16MnCr5 900 ± 5 880 to 980 820 to 860 860 to 900 780 to 820 150 to 200
16MncrS5
16MnCrB5
60 3 Material Purchasing Influence Area

• specification of a factory certificate

• indication of the desired heat treatment condition (only for alloyed steels).

Similar to the case-hardening steels, the numerous possible options must be consciously
specified by the purchaser. Without further option specifications, delivery must also be
made here according to the basic requirements of DIN EN ISO 683-1 and 683-2. It is
therefore advisable to describe some options in more detail with regard to heat treatment.

Heat treatment condition

Without specifying a particular heat treatment condition, the untreated condition +U is
delivered. The following heat treatment conditions are possible:

• treated for shearability +S (semi-finished products, bars)

• soft annealed +A (all product forms)
• normalized +N (bars, flat products, forgings)
• quenched and tempered +QT (bars, flat products, forgings)
• annealed to spherical carbides +AC (for cold extrusion grades).

If a heat treatment condition other than +U (untreated) is to be delivered, the correspond-

ing symbol must be specified.
For the heat treatment condition +QT, the location of the sample must be taken into
account in particular, as the mechanical properties may not be present throughout the
cross-section depending on the dimension and hardenability. For bars, rings, and wires
with a diameter of d > 25 mm, testing is carried out at a surface distance of 12.5 mm,
Fig. 3.12. For smaller dimensions, the sample is in the center of the product.
The sample positions and orientations for flat products can also be found in the men-
tioned standard. It should be pointed out again that the test values in the supplier certifi-
cates were measured at the positions shown and do not necessarily have to be present at
every other position of the cross-section.
Under Sect. 3.1 of this standard, a term is defined that is primarily important for
heat treatment and does not directly affect material procurement. A “relevant heat treat-
ment cross-section” is defined. For each product form, an “equivalent diameter” is thus
defined, which has the same cooling rate compared to the sample position of the respec-
tive product form.
For some simple geometries, such as square, rectangular, hexagonal or octagonal
cross-sections, this standard offers in the normative appendix A.1 the determination of
the relevant heat treatment cross-section. For rectangular cross-sections, an implemen-
tation example is marked in Fig. 3.13. The relevant heat treatment diameter for a flat
dimension of 65 × 160 mm is 100 mm. This determination method offers heat treatment
companies a simple aid in determining heating times for all heat treatment processes.
For the description of the heat treatment condition, note 2, referred to as an editorial
change in the standard, is also important:
3.2 Quenched and Tempered Steels 61

Fig. 3.12 Location of samples in bars, seamless rolled rings, and wire rod according to DIN EN
ISO 683-1
62 3 Material Purchasing Influence Area

Thickness

Width

Fig. 3.13 Determination of the relevant heat treatment cross-section for flat dimensions according
to DIN EN ISO 683-1

For the sake of simplification, the term “quenched and tempered”, unless otherwise
stated, is also used for the austempered condition.
This determination was already once in an earlier edition of the standard, the DIN
EN 10083-1:1996-10 [24], and was removed in the subsequent version DIN EN 10083-
3:2007-01 [24]. Now this note is again fixed in the current ISO standard. This determina-
tion is relevant for mechanical processing, as the example in Sect. 3.7 shows.

Austempering
The term austempering is listed for the first time as a heat treatment condition in DIN
EN ISO 683-1 and -2. This refers to a heat treatment for the production of a bainitic or
ausferritic microstructure. While the bainitic microstructure is well known in the heat
treatment of quench and temper steels, the term austempering has so far been used pri-
marily in the heat treatment of alloyed cast materials. This is also described in DIN EN
ISO 4885:2018-07 [25], which is given as a source in the heat-treatable steel standard.

Ausferrite is understood to be a microstructure of ferrite and stabilized austenite with

high hardness and at the same time high ductility. This microstructure formation is char-
acteristic for ADI cast iron after a targeted isothermal transformation.
3.2 Quenched and Tempered Steels 63

Ordering options
Similar to the case hardening steels described earlier, the quench and temper steels can
also be delivered with special requirements. However, these additional options must also
be explicitly defined in the inquiry or order and confirmed by the supplying plant.

From the list of ordering options listed under Sect. 5.2, the requirements that are of par-
ticular relevance for the usage properties of a heat-treatable steel are described in more
detail below.

Requirements for the Hardenability

From the term for the steel group quench and temper steels, the hardenability is already
recognizable as a central and important property. For high grade steels, the standard
DIN EN ISO 682-2:2018-09 sets the hardenability classes hardness classes +H, + HH
and +HL. This is exemplified in Fig. 3.14 using the example of the heat-treatable steel
42CrMo4. For the hardenability class +H, the entire scatter band applies. In the harden-
ability class +HH, the scatter band is reduced by the lower third and in the hardenability
class +HL, the upper third is faded out.

In martensitic transformation, the carbon content alone determines the achievable mar-
tensite hardness. This relationship was already presented in 1938 by Burns, Moore and
Archer and is still cited in the current literature [26].

Material 42CrMo4, 1.7225 | hardening temperature 850 °C

70

65

60
+HH
55

+HH
Hardness [HRC]

50
+H
45
+H
40
+HL
35 +H
+HL
30

25

20
0 5 10 15 20 25 30 35 40 45 50
Face distance [mm]

Fig. 3.14 Hardenabilty scatter band of the heat-treatable steel 42CrMo4

64 3 Material Purchasing Influence Area

In the face distance of 1.5 mm, therefore, only the permissible scatter width in the
carbon content is noticeable, as a martensitic transformation takes place here. The mar-
tensite reaches hardness values of min. 53 HRC and max. 61 HRC.
The biggest difference between the minimum and maximum hardness lies in the face-
to-face distance of 20 mm. The minimum value is 34 HRC and the maximum value is
56 HRC. These large hardness differences of up to 22 HRC become noticeably apparent
with larger dimensions. Although the DIN EN ISO 683-2 provides dimension-depend-
ent information for the achievable mechanical properties, it should be remembered that
the testing for these specifications always takes place 12.5 mm below the surface. State-
ments about the properties throughout the entire cross-section cannot be derived from
this information.
For a round dimension of Ø 75 mm, the mechanical properties were exemplarily rep-
resented for three quenched and tempered steels 42CrMo4, which differ in hardenability,
Table 3.4 and Fig. 3.15. The simulation program “StahlRegression” was used to deter-
mine the mechanical properties [27].
With identical heat treatment, there were significant differences in tensile strength.
The analysis with the low hardenability does not reach the tensile strength range speci-
fied in the standard at the tempering temperature of 550 °C. By lowering the tempering
temperature below 500 °C, the minimum value of 900 MPa can be achieved, but this is
then associated with significantly reduced toughness.

1300
Hardenability: high medium low
1200
Calculated tensile strength [MPa]

1100

1000

900

800

700
Dimension: 75 mm Ø
850 °C oil + 550 °C 2h
600
Edge Center Core Center Edge

Fig. 3.15 Calculated hardenability for different analyses of the steel 42CrMo4
3.2 Quenched and Tempered Steels 65

Table 3.4  Analysis specifications for calculating hardenability

Target analysis C Si Mn P S Cr Ni Mo Cu
42CrMo4 % % % % % % % % %
acc. to DIN EN ISO 0.38 0.10 0.60 max. max. 0.90 – 0.15 max.
683-2:2018-09
0.45 0.40 0.90 0.025 0.035 1.20 – 0.30 0.40
High hardenability 0.45 0.25 0.85 0.012 0.009 1.15 0.15 0.25 0.1
Medium hardenability 0.42 0.25 0.75 0.012 0.009 1.05 0.05 0.20 0.1
Low hardenability 0.37 0.25 0.65 0.012 0.009 0.95 0.05 0.15 0.1

The material with high hardenability would have to be tempered at 625 °C to achieve
1000 MPa, which would result in significantly better notch toughness.
The mechanical properties of quenched and tempered steels up to a round dimension
of Ø 250 mm and a thickness for flat products up to 160 mm are specified in DIN EN
ISO 683-2.
For larger dimensions of open-die forgings, property specifications are listed in DIN
EN 10250-3:1999-12 [28] up to Ø 500 mm. The specified minimum value for the tensile
strength of 600 MPa can be achieved by all three analysis variants listed in Table 3.4.
The heat-treatable steel 42CrMo4 is described in the Steel-Iron Material Sheet SEW
550 [29] up to a dimension range of Ø 750 mm. In contrast to the previously explained
DIN EN 10250-3, this standard provides for a nickel content of up to 0.6%. This
increases the possible tensile strength range for dimensions up to Ø 500 mm to 690 to
890 MPa.
For all unalloyed quenched and tempered steels, consideration of hardenability is of
particular importance. The hardenability-increasing alloying elements chromium, molyb-
denum, and nickel are specified with maximum values. The sum of these three alloying
elements is also limited to 0.63%. This wide range can lead to significant hardness varia-
tions in surface hardening.

Non-metallic inclusions
DIN EN ISO 683-2 states that the high grade steels must have a certain microscopic
purity level. A dimension-dependent limit value table, as was still informatively given
in Annex E in DIN EN 10083-1:2006, no longer exists in the ISO standard. Instead, it is
pointed out that both the evaluation method and the limit values must be agreed between
the supplier and the customer. It must also be determined whether a check must be car-
ried out.

For many decades, the purity level determination according to DIN 50602:1985 [8] was
a widely used determination method. This standard was already replaced in 2007 by DIN
EN 10247:2007-07 [30]. Reference values for this determination method were not nor-
matively published for quenched and tempered steels. A transferability of the results of
66 3 Material Purchasing Influence Area

both mentioned standards is also not possible. Therefore, many factory standards still
contain a limit value table for oxidic non-metallic inclusion types from DIN 50602.
For heat-treatable steels with regulated sulfur content, e.g. C45R or 42CrMoS4, DIN
EN ISO 683-1 or 683-2 recommends excluding the evaluation of sulfides. For heat-treat-
able steels from the group of free-cutting steels, DIN EN 10087:1999-01 [31], with sig-
nificantly increased sulfur content, e.g. 44SMn28, the material standard does not contain
any information about the manganese sulfide size and distribution. However, the Steel
Iron Test Sheet 1572 [32] allows an evaluation using image panels, which is why an
agreement between the customer and the supplier is also possible for this steel group.
A comparable regulation for macroscopic non-metallic inclusions is described in DIN
EN ISO 683-2. If proof is required, the method and limit values must be agreed upon at
the time of inquiry and order. In most cases, an agreement is reached on an ultrasonic
test with a replacement error size to be agreed upon.
The determination of the microscopic purity level according to specified methods
rarely revealed a primary cause of damage in damage investigations on heat-treated com-
ponents. In contrast, macroscopic non-metallic inclusions or inclusion nests were very
frequent causes of errors.

3.3 Nitriding Steels

Both case-hardening steels and quenched and tempered steels can be nitrided or nitro-
carburized. Nevertheless, there is a separate standard for describing nitriding steels, DIN
EN ISO 683-5:2021-08 [33]. Nitriding steels are alloyed high grade steels with special
alloy additives for nitride formation. These include the alloying elements aluminum,
chromium, molybdenum, and vanadium. The materials 31CrMoV9 and 34CrAlNi7-10
are the best-known nitriding steels. A typical analysis is compared with the target specifi-
cations of the standard, Table 3.5 and 3.6.
Most of the binding order specifications are identical to the general quenched and
tempered steels and are therefore not described again here. The order options lack state-
ments on hardenability.
The usual and appropriate heat treatment condition before the final nitriding is the
tempering condition +QT. This is also a frequently ordered heat treatment condition. The
nitriding temperature must be 30–50 °C below the tempering temperature to avoid sub-
sequent reduction in hardness in the base material. Therefore, if the nitriding temperature
is 550 °C, the tempering temperature should be at least 600 °C. However, a too high tem-
pering temperature can also lead to insufficient nitriding results, as the following exam-
ple shows.
Bar steel of dimension Ø 75 mm was ordered from the nitriding steel 34CrAlNi7-10
+QT with the tensile strength range of 850 to 1050 MPa specified according to DIN EN
10085:2001-07 [34]. The 3.1 mill test certificate confirmed the measured tensile strength
3.3 Nitriding Steels 67

Table 3.5  Analyses for the nitriding steel 31CrMoV9

Target analysis C Si Mn P S Cr Ni Mo V
31CrMoV9 % % % % % % % % %
acc. to DIN 0.27 max. 0.40 max. max. 2.30 – 0.15 0.10
EN ISO
683-5:2017-07 0.34 0.40 0.70 0.025 0.035 2.70 – 0.25 0.20

Permissible ± 0.02 + 0.03 ± 0.04 + 0.005 + 0.005 ± 0.10 – + 0.03 + 0.02

limit deviation
Typical analy- 0.32 0.33 0.59 0.013 0.011 2.34 0.13 0.19 0.11
sis

Table 3.6  Analyses for the nitriding steel 34CrAlNi7-10

Target analysis C Si Mn P S Cr Ni Mo Al
34CrAlNi7-10 % % % % % % % % %
acc. to DIN EN 0.30 max. 0.40 max. max. 1.50 0.85 0.15 0.80
ISO
683-5:2017-07 0.37 0.40 0.70 0.025 0.035 1.80 1.15 0.25 1.20

Permissible ± 0.02 + 0.03 ± 0.04 + 0.005 + 0.005 ± 0.05 ± 0.05 ± 0.03 ± 0.10
Limit deviation
Typical analysis 0.35 0.26 0.71 0.001 0.012 1.62 0.92 0.15 0.95

with 857 MPa. The chemical composition corresponded to the specifications of the men-
tioned standard, see Table 3.6.
Spindles were made from the quenched and tempered bars, which were nitrided at
500 °C for 84 hours after mechanical processing. The surface hardness achieved was 680
HV 10, significantly below the target specifications and otherwise achieved results. A
second nitriding treatment also did not lead to the required and otherwise usual surface
hardness values of at least 900 HV 10. The nitriding hardness depth determined after two
nitriding treatments was 0.87 mm, Fig. 3.16.
As part of the investigation, a tensile test sample was taken from the nitrided spindle
at the standard position and tested. The tensile strength achieved was 825 MPa, below
the target requirement of at least 850 MPa and also below the certificate indication of
857 MPa. The difference between the two tensile strength values was small and could
well have been caused by usual measurement scatter. More significant than the differ-
ence between the two tensile test results, however, was the fact that a very high tem-
pering temperature was chosen during tempering at the mill. To check the achievable
surface hardness after nitriding, parts of the spindle were therefore re-hardened and tem-
pered at 650 °C.
The usual nitriding process was carried out on these newly quenched and tempered
samples at 500 °C for 84 hours. The surface hardness now reached values of over 1000
HV 10. The nitriding hardness profile corresponded to the expected and usual result with
this steel and the performed nitriding treatment, Fig. 3.17.
68 3 Material Purchasing Influence Area

900

800
NHD310 = 0.87 mm
700

600
Hardness [HV 0.5]

500

400

300

200

100

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Edge distance [mm]

Fig. 3.16 Determination of the nitriding hardness depth after two nitriding treatments at 500 °C
for 84 hours

1200

1000

NHD370 = 0.39 mm
800
Hardness [HV 0.5]

600

400

200

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Edge distance [mm]

Fig. 3.17 Determination of the nitriding hardness depth after further tempering at a tempering
temperature of 650 °C and nitriding treatment at 500 °C for 84 hours
3.4 Martensitic Stainless Steels 69

This shows that the lower limit of 850 MPa is not suitable for achieving the max.
surface hardness. The tempering temperature was indicated in the supplier's certificate
as 680 °C. At this high tempering temperature, significant amounts of the nitride form-
ers are already stably bound in the carbides and are therefore not available in sufficient
quantities for nitride formation. This correlation with concrete measurement results was
also described for the steel 42CrMo4 in [35].
To achieve a max. surface hardness after nitriding, a tensile strength of at least 900–
950 MPa is required. This also applies to components that are purchased in the annealed
condition, machined and only quenched and tempered before the final nitriding.

3.4 Martensitic Stainless Steels

From the extensive collection of standards for stainless steels, DIN EN 10088-3:2014-12
[36] with the description for semi-finished products, bars, rolled wire, drawn wire, pro-
files and bright steel products is most suitable for examining and evaluating properties
after mechanical manufacturing and heat treatment. From this standard, only the marten-
sitic stainless steels are considered.
With these steels, conventional hardening and tempering can be carried out. The steels
of this steel group have a minimum content of 10.5% chromium and carbon contents up
to max. 1.2%. They can be obtained in the annealed condition +A or in the quenched
and tempered condition +QT. The chemical composition of the two frequently used
steels X20Cr13 and X90CrMoV18 are shown in Table 3.7 and 3.8. The designation
+QT is followed by a number that describes the lower limit of the tensile strength, e.g.,
X20Cr13 QT700 stands for the tensile strength range of 700 to 850 MPa.

Table 3.7  Analyses for the steel X20Cr13

Target C Si Mn P S Cr Ni Mo% Cu
analysis % % % % % % % %
X20Cr13 0.16 max. max. max. max. 12.0 – – –
acc. DIN
EN 10088- 0.25 1.00 1.50 0.040 0.015 14.0 – – –
3: 2014-12
Permissible ± 0.02 + 0.05 ± 0.04 + 0.005 + 0.005 ± 0.15 – – –
Limit
deviation
Typical 0.202 0.31 0.40 0.017 0.003 13.3 0.22 0.07 0.07
analysis
70 3 Material Purchasing Influence Area

Table 3.8  Analyses for the nitriding steel X90CrMoV18

Target C Si Mn P S Cr Mo V Ni
analysis % % % % % % % % %
X90Cr- 0.85 max. max. max. max. 17.0 0.90 0.07 –
MoV18
acc. 0.95 1.00 1.00 0.040 0.015 19.0 1.30 0.12 –
DIN EN
10088-3:
2014-12
Permis- ± 0.03 + 0.05 ± 0.03 + 0.005 + 0.005 ± 0.20 ± 0.05 ± 0.03 –
sible
Limit
devia-
tion
Typical 0.87 0.16 0.48 0.029 0.009 19.1 0.95 0.08 0.37
analysis

The standard also stipulates the following necessary order specifications:

• the required quantity;

• the form of the product (e.g., round bar, square bar, or rolled wire);
• if a suitable dimensional standard is available (see Table 7 and Appendix C) the num-
ber of the standard and the selected requirements; if no dimensional standard is avail-
able: the nominal dimensions and the required limit dimensions;
• the type of material (steel)
• the number of this document;
• short name or material number;
• if the table for the mechanical properties contains more than one treatment condition
for the relevant steel, the abbreviation for the required heat treatment or the required
cold hardening condition;
• the required type of execution (see abbreviations in Table 7);
• if an inspection of the internal condition is required, the products must be inspected
according to EN 10306 or EN 10308;
• any additional optional test (see 7.2.3d);
• designation for a factory certificate 2.2 or if required another test certificate according
to EN 10204 (see 7.2.1).

Furtherorder options are only defined for the surface condition and the type of execution.
For this group of steels, the term hardenability commonly used for quenched and tem-
pered steels is not used, as no usable results can be achieved in the hardenability test
3.5 Precipitation Hardening Steels 71

according to DIN EN ISO 642 due to the high chromium content of at least 10.5%. How-
ever, DIN EN 10088-3 defines dimension ranges for the specified mechanical properties.

3.5 Precipitation Hardening Steels

The precipitation hardening steels are also listed in DIN EN 10088-3 [36]. The analysis
specifications and a typical analysis measured in the laboratory for the very frequently
used steel X5CrNiCuNb16-4 are listed in Table 3.9.
In contrast to the hardenable martensitic steels, these steels undergo a solution treat-
ment (+AT) followed by precipitation hardening (+P). The solution annealing takes place
in the temperature range of 1030–1050 °C. To avoid pre-precipitation, rapid cooling in
oil or water is required. After solution annealing and quenching, a nickel martensitic
microstructure is present.
Mechanical processing takes place in the solution-treated state, followed by harden-
ing. Depending on the material, different precipitation states can be set. The treatment
abbreviation P is followed by a number, e.g. P930, which describes the minimum tensile
strength in the hardened state.
The heat treatment to be carried out in the +AT state for hardening is carried out at
different temperatures and times depending on the required tensile strength. The follow-
ing treatment steps are required for the target states P800 and P930:

P800 (+AT) + 760 °C 2 h → Air + 620 °C 4 h → Air

P930 (+AT) + 620 °C 4 h → Air

In individual cases, the aforementioned hardening did not achieve the expected tensile
strength. The cause was always insufficiently rapid cooling after solution annealing. A
further solution annealing with sufficient cooling speed and subsequent hardening led to
the required target values in these cases.

Table 3.9  Analyses for the steel X5CrNiCuNb16-4

Target analysis C Si Mn P S Cr Ni Mo Cu Nb
X5CrNi- % % % % % % % % % %
CuNb16-4 max. max. max. max. max. 15.0 3.00 max. 3.00 5xC
acc. DIN EN max.
10088-3: 0.07 0.70 1.50 0.040 0.030 17.0 5.00 0.60 5.00
0.45
2014-12
Permissible ± 0.02 + 0.05 ± 0.04 + 0.005 + 0.005 ± 0.15 – – –
Limit devia-
tion
Typical analy- 0.03 0.36 0.43 0.015 0.001 15.42 4.32 0.13 3.03 0.29
sis
72 3 Material Purchasing Influence Area

3.6 Tool Steels

Tool steels were already used by blacksmiths in antiquity to make tools. It is therefore
not surprising that this group of materials has a long tradition with numerous develop-
ments that were regionally and industry-specific. The variety of standardized tool steels
is very large and is constantly being expanded by numerous further developments by
steel producers. In the current DIN EN ISO 4957:2018-11 [37] a distinction is made
between unalloyed cold work steels, alloyed cold work steels, alloyed hot work steelsand
high-speed steels.
The binding order specifications are almost identical with all other steel groups. The
most common heat treatment condition is the soft annealed state +A. Only when order-
ing tool steels in the quenched and tempered state is it necessary to specify the required
hardness in addition to the heat treatment condition. If no order options have been
agreed, delivery must be made according to the basic requirements of the mentioned
standard.
The order options indicate that other heat treatment conditions and special surface
conditions can be agreed.
Additional requirements are normatively described in Annex B.

Piece analysis
It is stated that a piece analysis per melt is to be carried out for elements that are speci-
fied in the target analysis. This means that additional element specifications must be
agreed separately.

Proof of minimum hardnessin the hardening test

The very important minimum hardness for tool steels can be proven in a hardening test
by agreement. For this purpose, a sample is to be taken, the size and removal position of
which are detailed in Chap. 11 of DIN EN ISO 4957. For products with a diameter or
thickness of ≥15 mm, a sample is to be taken, the min. dimension of which is 15 × 15
mm, Fig. 3.18.

For products with dimensions of <15 mm, a sample length of 20 mm ±5 mm is to be

taken while maintaining the cross-sectional dimension. For the hardening test, the mate-
rial-specific information

• on the austenitizing temperature,

• on the quenching medium, and
• on the tempering temperature
is specified. The minimum hardness to be achieved is also indicated. For the different
tool steel groups, an example from the standard is listed below:
3.6 Tool Steels 73

w ≥ 15

w ≥ 15

w ≥ 15
3

w ≥ 15 3

Fig. 3.18 Removal position for samples to prove the minimum hardness in the hardening test
according to DIN EN ISO 4957

C70U: 800 °C ±10 °C → Water + 180 °C ±10 °C: Minimum hardness

57 HRC
90MnCrV8: 790 °C ±10 °C → Oil + 180 °C ±10 °C: Minimum hardness 60 HRC
X153CrMoV12: 1020 °C ±10 °C → Air + 180 °C ±10 °C: Minimum hardness
61 HRC
X37CrMoV5-1 1020 °C ±10 °C → Oil + 550 °C ±10 °C: Minimum hardness
48 HRC
HS6-5-2 1220 °C ±10 °C → Oil + 560 °C ±10 °C: Minimum hardness
64 HRC
In the informative part of DIN EN ISO 4957, hardness-tempering temperature curves are
depicted. These diagrams are only suitable for concrete specifications of the tempering
temperatures to a limited extent. The standard itself states that the curves currently only
provide a rough guide for the tempering behavior.
The hardness-tempering temperature curves were determined on samples of the
dimension from Fig. 3.18 and a salt bath hardening. The high-alloy cold work steels, the
hot work steels, and the high-speed steels are now predominantly hardened and tempered
in a vacuum furnace. The conditions present in these heat treatment processes differ
significantly from the salt bath heat treatment parameters. There are also no details for
larger cross-sections, and secondary hardening processes are also not described.
However, for the steels that can be hardened in a vacuum furnace, detailed informa-
tion on the required heat treatment including deep freezing treatments has been published
online by the steel producers. The necessary information is therefore freely accessible.
The tool steel standard contains the promised properties in the soft or quenched
and tempered delivery condition as well as information on the minimum hardness after
a specified hardness test. Information on hardenability, purity level, toughness in hot
work steels is not included. The example of the toughness properties important for hot
work steels shows that further developed primary and secondary metallurgical meas-
74 3 Material Purchasing Influence Area

ures have enabled significant improvements. With the electroslag remelting process as
a tertiary metallurgical process, tool steels of the highest purity and homogeneity can
be produced. Compared to pure electric steel production without secondary metallurgy
and special measures, improvements in the impact bending work of the hot work steel
X37CrMoV5-1, 1.2343, in the quenched and tempered condition (44 HRC) from 90 J to
380 J were achieved [38]. The following measures were necessary for this:

• lowest accompanying elements

• calcium treatment
• electro-slag remelting
• diffusion annealing
• special heat treatment.

Since these special measures are also not included as an option in DIN EN ISO 4957,
individual agreements with the steel supplier are necessary when placing an order.
A reference exclusively to the standard can lead to very different toughness properties
depending on the manufacturing process.
The group of melt metallurgical materials has been expanded by the group of pow-
der metallurgical tool steels. Numerous tool steels are available from both manufactur-
ing processes. A standardization of the powder metallurgical tool steels is still pending,
which is why only the manufacturer’s information about the specific properties and heat
treatments informs here.

3.7 Damage Examples—Material Purchase

Example 1—Poor Machinability of Turned Parts from Rods

Component: Turned parts from rods∅ 22 mm

Material: 25CrMoS4, Material No. 1.7213
Error pattern: Poor machinability in one of two deliveries
Heat treatment: Tempering

For turned parts, a purchaser ordered rods Ø 22 mm made of the material 25CrMoS4,
1.7213 from two different steel suppliers. The order text specified the delivery condi-
tion and the tensile strength with the indication: “quenched and tempered to Rm = 790–
910 MPa”.
During the turning process, it was noticed that one delivery was significantly harder to
machine. A hardness test and conversion into a tensile strength showed that both deliver-
ies had almost identical values and were within the target specification.
3.7 Damage Examples—Material Purchase 75

Table 3.10  Results of the analysis determinations

Target analysis C Si Mn P S Cr Ni Mo Cu
25CrMoS4 % % % % % % % % %
acc. to DIN EN ISO 0.22 0.10 0.60 max. 0.020 0.90 – 0.15 max.
683-2:2018-09
0.29 0.40 0.90 0.025 0.040 1.20 – 0.30 0.40
Delivery A 0.25 0.17 0.71 0.007 0.022 1.06 0.17 0.17 0.15
Delivery B 0.26 0.20 0.77 0.006 0.021 1.05 0.11 0.15 0.13

The determination of the chemical composition of both deliveries also revealed no

significant differences, Table 3.10.
Only the metallographic microstructure examination provided an explanation for
the different machining behavior, Fig. 3.19. The easily machinable batch showed a
“classic” quenched and tempered microstructure after hardening and tempering. In
contrast, the poorly machinable batch showed a bainite microstructure of the upper
transformation stage. The form of the carbides differed significantly. While after hard-
ening and tempering small carbides are evenly distributed in the ferrite matrix, in the
bainite microstructure the carbides are arranged in a line one after the other. Although
the individual carbides in the bainite are very fine, the directed arrangement of many car-
bides acts as an obstacle in mechanical machining.
The electron-optical documentation made this difference clear, Fig. 3.20. The car-
bides in the martensitic tempering microstructure were randomly distributed, while
the individual carbides in the bainite, although of comparable size, were arranged in
a directed manner. This arrangement caused the observed differences in mechanical
machining.
The steel supplier of the bainitically transformed rods delivered in accordance with
the standard, as note 2 from DIN EN ISO 683-2 “For the sake of simplification, the term
“tempered”, unless otherwise stated, is also used for the quenched and tempered condi-
tion” was not contradicted. Therefore, a complaint or free post-treatment could not be
demanded. However, it is questionable whether even a technical buyer can recognize the
consequences for mechanical machining from this note.
To avoid such a situation, the order for the +QT condition should therefore addition-
ally specify “hardening and tempering” or “liquid quenching ”.

Example 2—Soft spots on crane wheels after flame hardening

Component: Crane wheels∅ 520 mm

Material: C45E, Material No. 1.1191
Error pattern: Soft spots after flame hardening
Heat treatment: Flame hardening
76 3 Material Purchasing Influence Area

Fig. 3.19 Microstructure of the two different batches (left: easily machinable, right: poorly
machinable)

Fig. 3.20 Carbide distribution in tempered martensite (left) and in bainite (right)

For many years, the heat-treatable steel C45E was ordered for the production of crane
wheels without a hardenability requirement. After mechanical manufacturing, the run-
ning surface was flame-hardened. The required surface hardness of 55 + 3 HRC and the
hardening depth of 4 + 2 mm were always reliably achieved. The supplying plant was
aware of this application and heat treatment and always delivered the C45E steel with
analyses similar to batch 1 in Table 3.11.
At a later point in time, an order with the same order text was placed with another
supply plant, referencing DIN EN 10083-2. After flame hardening, soft spots were
detected on the entire running surface and the hardness values varied from 32 to 45
HRC. The determination of the chemical composition is shown in Table 3.11 as batch 2.
For both deliveries, the hardenability was calculated according to the formulas of SEP
1664 [5] and the HRC hardness values are compared in Table 3.12. The relatively small
difference in the total content of Cr+Ni+Mo led to mixed microstructures of martensite,
3.7 Damage Examples—Material Purchase 77

Table 3.11  Analyses for the heat-treatable steel C45E

Target analysis C Si Mn P S Cr Ni Mo Cu
C45E % % % % % % % % %
acc. to DIN EN ISO 0.42 0.10 0.50 max. max. max. max. max. max.
683-1:2018-09
0.50 0.40 0.80 0.025 0.035 0.40 0.40 0.10 0.30
Batch 1 0.48 0.21 0.75 0.015 0.008 0.25 0.11 0.05 0.12
Batch 2 0.41 0.22 0.55 0.009 0.012 0.03 0.03 0.01 0.08

Table 3.12  Hardenability values calculated according to SEP 1664 for the two deliveries from
Table 3.11
[mm] 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0
Ch 1 58.8 58.6 57.9 55.0 49.8 42.9 37.2 35.0 33.2 32.2 31.5
Ch 2 55.8 54.0 44.7 29.8 23.9 21.8 21.6 22.4 23.2 23.3 22.8

bainite, pearlite, and ferrite in the wheels from delivery 2. In contrast, the microstructure
was purely martensitic in the analysis from the original batch.
This example shows that for achieving a constant hardness result, it is absolutely nec-
essary to specify sufficient hardenability in the order.

Example 3—Insufficient hardness after oil quenching in stabilizers

Component: Stabilizers
Material: 38MnB5, Material No. 1.5532
Error pattern: Insufficient hardness after oil quenching
Heat treatment: Tempering

For the production of stabilizers, tubes of the dimension Ø 60 mm × 8 mm made of the

boron-alloyed heat-treatable steel 38MnB5 were supplied. The tubes were sealed on the
face with flat bars made of the material 42CrMo4 by welding. The hardening tempera-
ture was 880 °C and due to the geometry, the quenching treatment was carried out in oil
transversely to the flow direction and only over the outer surface of the tubes, Fig. 3.21,
right part of the image. The required hardness of 500 HB as a guarantee for a martensi-
tic quenched and tempered microstructure was also achieved with this large tube wall
thickness and unfavorable quenching situation, Fig. 3.22, left part of the image.
The requirement of minimum hardness was specified in the order specification. The
geometry and the quenching conditions, on the other hand, were not documented, but
were known to the original supplier.
A new supply plant checked the hardness before delivery according to the require-
ment from the order specification. Unaware of the special quenching situation, the
78 3 Material Purchasing Influence Area

a b

Fig. 3.21 Schematic representation of the quenching (a) open vertical quenching; (b) face-sealed
horizontal quenching

Fig. 3.22 Martensitic quenched and tempered microstructure (a) after vertical quenching (Sup-
plier 1) (b) with a high proportion of bainite after horizontal quenching (Supplier 2)

test part was quenched vertically, allowing cooling from inside and outside, Fig. 3.21,
left part of the image. With this arrangement, the hardness of 500 HB could be easily
achieved even with lower hardenability. On the other hand, a high proportion of upper
bainite was achieved when hardening the face-sealed stabilizer, Fig. 3.22, right part of
the image.
The original supplier was involved in the development of the product and delivered a
hardenability at the upper limit due to the peculiarity. The delivery from the second sup-
ply plant was also within the permissible range, but closer to the lower limit.
3.7 Damage Examples—Material Purchase 79

The purchaser should therefore pay special attention to the material property “hard-
enability” for the aforementioned reasons and clearly state the requirements taking into
account the component geometry and the heat treatment requirements.
In analogy to the case-hardening steels, it is also up to the purchaser to determine
whether a check of the hardenability is required and whether the hardenability is to be
measured experimentally or can be calculated via the analysis. Since the calculation
method is left to the supply plant, the calculation formula should also be documented.
Only if the calculation formulas from different suppliers are known, a check and com-
parison are possible.

Example 4—Crack formation in cast mold parts

Component: Cast mold parts

Material: EN-GJS-600-3, Material No. 5.3201
Error pattern: Crack formation
Heat treatment: Inductive surface hardening

In order to be able to surface harden spheroidal graphite cast iron, the chemical composi-
tion must be adjusted so that a sufficient amount of pearlite is formed. Cast mold parts
from two different foundries of the cast material EN-GJS-600-3 (GGG 60) according to
DIN EN 1563:2019-04 [39] were inductively hardened, Fig. 3.23. After the inductive
hardening, a high failure rate occurred due to crack formations in one of the two delivery
series. The cracks were discovered during the mechanical milling process.
Since it involved two different foundries, the chemical composition of one casting
part from each was initially determined. For spectroscopic analysis, the sample mate-
rial must be remelted and solidified white through rapid cooling in the case of graphite-

Fig. 3.23 Cast mold part and marking of the surfaces to be hardened
80 3 Material Purchasing Influence Area

Table 3.13  Analysis results of both casting deliveries

C Si Mn P S Cr Ni Al Cu Mg
% % % % % % % % % %
Casting part A 3.31 2.88 0.43 0.047 0.003 0.03 0.03 0.015 0.02 0.0210
(n. i. O.)
Casting part B 3.32 2.71 0.61 0.037 0.008 0.03 0.05 0.021 0.63 0.0260
(i. O.)
All information in mass-%

containing materials. After that, a measurement with an optical emission spectrometer is

easily possible. Alternatively, a carbon determination can also be carried out via a com-
bustion analysis (Table 3.13).
The analysis results were very similar except for the copper content. However, there
was a significant difference in the copper content. The element copper is a strong pearlite
former or reduces the amount of ferrite, which was already visually clearly recognizable
metallographically, Fig. 3.24.
Through image analysis, the different microstructural phases were also quantifiable
(Table 3.14).
In the case of cast part A, there was crack formation perpendicular to the inductively
hardened sidewall. To visualize the crack, a flat microsection was cut and prepared perpen-
dicular to the hardened side surface, Fig. 3.25. The image documentation in the unetched
state not only shows the main crack with transverse secondary cracks, but also a widely
extended graphite-free area. Only at the immediate surface was there an approximately
250 μm wide seam with spheroidal graphite. The amount of graphite increased again in
the further course and the crack ended in an area with a full amount of spheroidal graphite.
Another microsection was cut and prepared behind the crack, which made the entire
hardened sidewall visible. Initially, the graphite-free edge area was also recognizable
here in the unetched state, Fig. 3.26a. At the lower end of the hardening layer, there was
no graphite left on the surface. Slightly above, the thin seam with spheroidal graphite
was visible again.
The overview magnification in Fig. 3.26b and the higher magnification in Fig. 3.26c
show a white solidified cast iron with martensite and primary cementite. Occasionally,
small remnants of unmelted graphite balls can still be seen. At the immediate surface, the
graphite balls are present in a martensitic matrix.
The area of white solidification is followed by an area with increasing graphite con-
tent. The microstructure consists of martensite with a very high proportion of residual
austenite, Fig. 3.26d. At this microstructural position, only a few primary carbide pre-
cipitations are present.
In the further course, a proper martensitic quenched and tempered microstructure
with spheroidal graphite is present, Fig. 3.26e. Finally, the hardening layer ends and the
base microstructure becomes visible again, Fig. 3.26f.
3.7 Damage Examples—Material Purchase 81

Cast A Cast B

Fig. 3.24 Casting microstructure in the unhardened area of both casting deliveries and quantita-
tive determination of the ferrite content by image analysis

Table 3.14  Results of the quantitative image analysis

Ferrite content Perlite content Graphite quantity
Cast A (n. i. O.) 26% 61% 13%
Cast B (i. O.) 8% 82% 10%

The microstructure formation was documented at a comparable position on a i.-O.-

cast part B, Fig. 3.27. This component had a flawless martensitic surface layer with a low
residual austenite content. A depletion of graphite was not noticeable.
The microstructure documentation clearly shows that during the induction hardening
of the cast part A, melting occurred, followed by white solidification and crack forma-
tion.
However, the question still needs to be clarified as to why such severe overheating
occurred. Here, the microstructure condition of the cast microstructure must be included
in the evaluation. The i.-O.-cast batch B, due to its higher copper content, had a sig-
nificantly larger amount of carbide. To achieve a proper quenched and tempered micro-
structure, a usual hardening temperature of 880–900 °C was sufficient. With the same
82 3 Material Purchasing Influence Area

Fig. 3.25 Crack origin and crack end in a flat microsection perpendicular to the inductively hard-
ened side surface (Cast A)

induction hardening parameters, this hardening result was not achieved with the n.-i.-O.-
cast batch A due to the lower amount of pearlite.
Up to this result, the responsibility for the insufficient hardness lay with the material
purchasing department, which was not aware of the connection and had not made any
specifications to the foundry. A cast alloy was produced without a sufficient amount of
pearlite, and full quench hardness could not be achieved. The determination of the chem-
ical composition or a microstructure examination could have revealed the cause.
However, it could be inferred from the microstructures that the hardening shop must
have made a consequential decision with the best of intentions. Since the procedure was
not communicated, the following explanations are assumptions derived from the micro-
structures.
In the mistaken assessment that the previously chosen hardening temperature does
not dissolve enough carbides from the pearlite, the hardening temperature was increased.
Whether this was done in small steps or one large step was not discernible. However, the
hardening temperature eventually reached the melting temperature, the spheroidal graph-
ite was dissolved, and primary carbides were formed in the white solidified cast iron dur-
ing quenching.
The question is certainly justified as to whether this is a heat treatment error in the
true sense. The purchased cast material could not achieve the required hardness after
induction hardening. However, it is a frequently made experience that a treatment stop
for cause finding involving the manufacturer is unrealistic due to the great time and cost
pressure and therefore very rarely occurs. However, after the significant change in the
hardening parameters, a review should have been carried out to avoid subsequent dam-
age.
3.7 Damage Examples—Material Purchase 83

Fig. 3.26 Microstructure images from the surface to the base material (n.-i.-O.-cast part A); (a)
graphite-free edge area; (b) white solidified cast iron with martensite and primary cementite; (c)
unmelted graphite balls in a martensitic matrix; (d) Martensite with a high proportion of resid-
ual austenite; (e) proper quenched and tempered microstructure with spheroidal graphite; (f) base
microstructure behind the hardening layer
84 3 Material Purchasing Influence Area

Fig. 3.27 Microstructure images in the hardening layer of the i.-O.-cast part B

Whether the test laboratory or the quality department was aware of the changed set-
ting data was not known. The cause of the damage was therefore both with the material
purchasing and the induction hardening. Such a sequence or chain of several causes of
error is, however, not an isolated case.

Example 5—Seizing on the gripper pin of an excavator bucket

Component: Gripper pin of an excavator bucket

Material: 42CrMo4, Material No. 1.7225 and C45, Material No. 1.0503
Error pattern: Seizing of the pin
Heat treatment: Inductive surface hardening

The gripper pins of an excavator bucket are subject to wear and need to be replaced at
regular intervals. For this purpose, pins made of the heat-treatable steel 42CrMo4,
1.7225, were quenched and tempered and subsequently nitrocarburized. Since the orig-
inal spare part caused high costs, purchasing was supposed to find and test a cheaper
solution. Pins were made from the unalloyed heat-treatable steel C45, 1.0503, and induc-
tively hardened. Compared to the original spare part, the material costs were lower and
the quenching and tempering followed by nitrocarburizing was no longer necessary. The
hardness profile after induction hardening even showed a greater hardening depth. Thus,
the ideal solution seemed to have been found and the first pins were installed.
The disillusionment came after a short time. After a few days of standstill, it was
found upon restarting that the bolts had seized, Fig. 3.28a. Deep corrosion scars had
formed. The nitrocarburizing of the original replacement part not only led to a reduction
in wear, but also to improved corrosion resistance due to the compound layer on the sur-
face, Fig. 3.28b.
References 85

Fig. 3.28 Surface of the replica and the original component (a) Inductively hardened replica with
corrosion scars (b) Original component quenched and tempered and nitrocarburized

References

1. DIN EN 10084:2008-06: Einsatzstähle – Technische Lieferbedingungen

2. ISO 683-3:2019-04: Für eine Wärmebehandlung bestimmte Stähle, legierte Stähle und Auto-
matenstähle – Teil 3: Einsatzstähle
3. Datenbank Stahlwissen®: Dr. Sommer Werkstofftechnik
4. DIN EN ISO 642:2000-01: Stahl – Stirnabschreckversuch (Jominy-Versuch)
5. SEP 1664:2004-06: Ermittlung von Formeln durch multiple Regression zur Berechnung der
Härtbarkeit im Stirnabschreckversuch aus der chemischen Zusammensetzung von Stählen.
Beiblatt zum SEP 1664: Ermittlung von Formeln durch multiple Regression zur Berechnung
der Härtbarkeit im Stirnabschreckversuch aus der chemischen Zusammensetzung von Stählen
6. ISO 6336-5:2016-08: Tragfähigkeitsberechnung von gerad- und schrägverzahnten Stirnrädern
– Teil 5: Festigkeit und Werkstoffqualitäten
7. ISO 4967:1998-10-1: Determination of content of nonmetallic inclusions – Micrographic
method using standard diagrams
8. DIN 50602:1985-09: Mikroskopische Prüfung von Edelstählen auf nichtmetallische Ein-
schlüsse mit Bildreihen
9. SEP 1927:2001-08: Ultraschall – Tauchtechnik – Prüfung des makroskopischen Reinheits-
grades von gewalzten oder geschmiedeten Stäben
10. Hock, St. Und D. Wiedmann: Längere Lebensdauer mit reineren Einsatzstählen – was lohnt
sich und wie rein ist rein? HTM J. Heat Treatm. Mat. 67 (2012) 1, S. 58–64
11. Hock, St.; Kleff, J.; Schulz, M.; Sollich, A.; Wiedmann, D.: Einfluß von Umform- und
Wärmebehandlungsfolgen auf Korngröße und Schwingfestigkeit von einsatzgehärteten
Bauteilen. HTM Härterei-Techn. Mitt. 54 (1999) 1, S. 45–52
12. DIN EN ISO 643-2013-05: Stahl – Mikrophotographische Bestimmung der erkennbaren
Korngröße
13. Huchtemann, B.; Schüler, V. Beitrag zur Beeinflussung der Austenitkorngröße von Edel-
baustählen. HTM Härterei-Techn. Mitt. 48 (1993) 2, S. 124–132
14. Klenke, K.; Kohlmann, R. Einsatzstähle in ihrer Feinkornbeständigkeit, heute und morgen.
HTM Z. Werkst. Wärmebeh. Fertigung 60 (2005) 5, S. 260–270
86 3 Material Purchasing Influence Area

15. Bleck, W.; Pariser, G.; Trute, S.: Herstellung und Verarbeitung moderner Stahlwerkstoffe.
HTM Z. Werkst. Wärmebeh. Fertigung 58 (2003) 4, S. 181–188
16. Clausen, B.; Konovalov, S.; Hoffmann, F.; Prahl, U.; Zoch, H.-W.; Bleck, W. Feinkornbestän-
digkeit von Bauteilen aus dem mikrolegierten Werkstoff 18CrNiMo7-6 in Abhängigkeit der
Prozesskette. HTM J. Heat Treatm. Mat. 65 (2010) 5, S. 257–268
17. Trute, S.: Einfluss der Prozesskette auf die Feinkornbeständigkeit von mikrolegierten Ein-
satzstählen. Dr.-Ing. Diss. RWTH Aachen, 2008
18. Hippenstiel, F.: „Einsatzstähle für die Hochtemperaturaufkohlung in der Praxis“, 4. Interna-
tionales Getriebestahlsymposium, 17.–18. Februar 2005. Wetzlar
19. DIN EN 10083-1:2006-10: Vergütungsstähle – Teil 1: Allgemeine technische Lieferbedingun-
gen
20. DIN EN 10083-2:2006-10: Vergütungsstähle – Teil 2: Technische Lieferbedingungen für unl-
egierte Stähle
21. DIN EN 10083-3:2009-01: Vergütungsstähle – Teil 3: Technische Lieferbedingungen für legi-
erte Stähle
22. DIN EN ISO 683-1:2018-09: Für eine Wärmebehandlung bestimmte Stähle, legierte Stähle
und Automatenstähle – Teil 1: Unlegierte Vergütungsstähle
23. DIN EN ISO 683-2:2018-09: Für eine Wärmebehandlung bestimmte Stähle, legierte Stähle
und Automatenstähle – Teil 2: Legierte Vergütungsstähle
25. DIN EN 10083-3:2007-01 Vergütungsstähle – Teil 3: Technische Lieferbedingungen für legi-
erte Stähle
26. DIN EN ISO 4885:2018-07: Eisenwerkstoffe – Wärmebehandlung – Begriffe
27. Burns, J.L., T.L. Moore, R.S. Archer: Trans. Amer. Soc. Met. 26 (1938), S. 1–36
28. Grimm, W., Sommer, P.: Pre-computation of Material Characteristics of Large Forgings Made
From Tempering Steels. Proceedings 14th International Forgemasters Meeting, Wiesbaden
September 3–8, 2020, S. 492–496
29. DIN EN 10250:1999-12: Freiformschmiedestücke aus Stahl für allgemeine Verwendung – Teil
3: Legierte Edelstähle
30. SEW 550:1976-08: Stähle für größere Schmiedestücke; Gütevorschriften
31. DIN EN 10247:2007-07: Metallographische Prüfung des Gehaltes nichtmetallischer Ein-
schlüsse in Stählen mit Bildreihen
32. DIN EN 10087-1999-01: Automatenstähle. Technische Lieferbedingungen für Halbzeug,
warmgewalzte Stäbe und Walzdraht
33. SEP 1572:2019-03: Mikroskopische Prüfung von Automatenstählen auf sulfidische nicht-
metallische Einschlüsse mit Bildreihen
34. DIN EN ISO 683-5:2021-08: Heat treatable steels, alloy steels and free-cutting steels- Part 5:
Nitriding steels
35. DIN EN 10085:2001-07: Nitrierstähle – Technische Lieferbedingungen
36. Huber-Gomman, U.: „Gasnitrieren-Anwendungsbeispiele und Einflussgrößen“ Stahl 1992
Seite 92/96
37. DIN EN 10088-3:2014-12: Nichtrostende Stähle – Teil 3: Technische Lieferbedingungen für
Halbzeug, Stäbe, Walzdraht, gezogenen Draht, Profile und Blankstahlerzeugnisse aus korro-
sionsbeständigen Stählen für allgemeine Verwendung
38. DIN EN ISO 4957:2018-11: Werkzeugstähle
39. Ernst, C.: Stand der Technik bei den Werkzeugstählen. Seminarfolien des Führungskräfte-
Seminars der Dr. Sommer Werkstofftechnik 2017
40. DIN EN 1563:2019-04: Gießereiwesen – Gusseisen mit Kugelgraphit
Material Influence Area
4

Contents

4.1 Trace Elements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

4.2 Segregation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
4.3 Non-metallic Inclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.4 Hardenability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
4.5 Damage Examples—Material. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

The material steel is still the most frequently used construction material in mechanical
engineering. This is not only due to the enormous number of different steel composi-
tions, but also to the experience with this material. In countless studies, scientific prin-
ciples and application-related properties have been developed and published. The first
edition of the journal Stahl und Eisen dates back to July 1, 1881 and is still published
monthly. Numerous other proven and new specialist publications appear at regular
intervals. The Stahl-Eisen-Liste is the official current compilation of all registered steel
grades contained in European standards with their material numbers. The last printed
edition was the 11th edition [1]. After that, the Stahl-Eisen-Liste was made available as
the StahlDat database [2] based on the internet. In international standardization, steels
are now listed without a material number.
Many steels are described in standards in terms of their chemical composition and
their delivery properties, which was discussed in the previous chapter. The differences
in properties between the basic quality and the optional additions have been sufficiently
described in this chapter. Beyond the standardized scope, steels have been further
developed for the most diverse areas of application, whose specific properties are not

© The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, 87

part of Springer Nature 2024
P. Sommer, Mistakes Before, During and After Heat Treatment of Steel,
https://doi.org/10.1007/978-3-658-44160-9_4
88 4 Material Influence Area

Defects in the material steel

Unintended trace elements

Segregations Copper-nickel segregation

Center risers Hydrogen

Inhomogeneities
Non-metallic influences
Lack of fine grain stability Oxidic inclusions

Aluminum-nitrogen ratio Blowhole

Hardenability
Reconciliation and information errors

Fig. 4.1 List of the most common errors in the material steel

contained in the material standards. The Handbook Stahl provides a comprehensive over-
view of properties and applications [3].
For the production of raw steel, the processes blast furnace+oxygen converter and
scrap+electric arc furnace are used on a large scale. The induction furnace systems used
for special applications are insignificant in terms of quantity. The steel production in Sie-
mens-Martin furnaces, which was once very intensively operated, now only exists spo-
radically. There is no longer a Siemens-Martin furnace in Germany. Worldwide, about
70% of raw steel production is melted in the oxygen top-blowing converter and 30% in
the electric arc furnace. These global production figures also reflect the ratios in German
steel production.
The melted raw steel is subsequently treated by secondary metallurgy and vacuum-
treated to a high percentage to remove the gaseous components as far as possible. Here
too, reference is made to the extensive literature on steel production, e.g. [4, 5]. The pro-
duced raw steel is now cast as continuous casting in over 95% of cases [6]. However,
ingot casting is still used for very large forgings.
All the aforementioned production and processing variants leave a specific “finger-
print” that affects the property profile in all subsequent manufacturing stages.
The most common errors in steel production and further processing are listed in Fig.
4.1 and divided into groups.

4.1 Trace Elements

The determination of unintentional trace elements already provides a clear indication

of the respective steel production technology [7]. Elements such as copper and nickel
cannot be removed during steel production due to their specific oxidation behavior. It is
4.2 Segregation 89

Table 4.1  Contents of trace elements in both steel production technologies

Element Pig Iron Scrap Raw Steel/LD Converter Raw Steel/Electric Arc Furnace
Nitrogen < 0.1 < 0.006 0.003–0.006 0.006–0.009
Nickel < 0.04 0.01 < 0.12 0.08–0.30
Copper < 0.01 0.08 < 0.10 0.10–0.30
Tin < 0.002 0.03 < 0.01 0.008–0.025
Arsenic < 0.002 0.006 0.002–0.012 0.005–0.026
All information in weight-%

therefore obvious that the use of scrap leads to higher concentrations of these elements
in the raw steel, Table 4.1. This is of great importance for heat treatment, which will be
explained in detail in the following error examples.

Gaseous Elements
Through vacuum degassing, the hydrogen content is to be reduced to values of max. 2.0
ppm, which can be taken from the test certificates of the steel producers. In most cases,
however, the actual value in the delivered products is lower. Through heating for hot
forming and subsequent heat treatments under flue gas atmosphere, this value typically
reduces to values below 1.0 ppm. Since hydrogen can be damaging in many ways, this
will also be discussed intensively.

4.2 Segregation

Another “fingerprint” can be taken from the macro-segregation profiles, Fig. 4.2. In the
case of ingot casting block casting, there is a negative segregation index at the foot. In
the middle of the ingot, the element concentrations are lower than at the ingot edges.
Just above the foot, there is an ideally segregation-free area. After that, the segregation
index continuously increases towards the ingot head and reaches maximum values at the
head. The element concentrations are then significantly higher in the middle of the ingot
than at the ingot edges. This means that the components made from a ingot can have dif-
ferent properties.
In the case of continuous casting, two different segregation profiles can occur. With-
out magnetic stirring, a pronounced center segregation occurs over the entire strand
length. The element concentrations are significantly increased compared to the edge
area. The edge areas, on the other hand, show a high uniformity over the strand length.
For products that are centrally drilled during production, this casting variant offers quali-
tative advantages, or the disadvantages of center segregation are eliminated.
A reduction in center segregation occurs when the melt is stirred electromagnetically.
Only minor concentration differences exist across the entire cross-section, with the sur-
90 4 Material Influence Area

Ingot casng Connuous ESU

Head not srred casng srred
Segregaon index

Foot Billet cross-secon

Fig. 4.2 Segregation profiles in billets from ingot casting, continuous casting and ESU remelting
(schematic) [8]

face also retaining the best possible properties here. A further improvement in homoge-
neity can be achieved by slight deformations during continuous casting (soft-reduction).
A very high uniformity both across the entire cross-section and over the block length
is achieved in the electro-slag remelting process. This production variant, referred to as
tertiary metallurgy, is used particularly in hot work steels to achieve the best possible
toughness properties.

4.3 Non-metallic Inclusions

The process sequences in steel production inevitably lead to non-metallic inclusions. A

distinction must be made between the microscopic and macroscopic purity. The deter-
mination methods of the microscopic purity are normatively described and belong to the
standard examination methods [9–12]. However, the current standardization points out
that both the determination method and limit values must be agreed between the supplier
and the customer. If a limit value has been set, the test certificates sometimes only con-
firm the limit value without specifying a measurement result.
An agreement on the determination method and the limit value must also be made for
macroscopic non-metallic inclusions. In addition to conventional ultrasonic testing, the
immersion technique is also used for the detection of macro inclusions [13–15]. How-
ever, numerous damage investigations have shown that macro inclusions can also cause
damage below the detection limit. Rarely does cracking occur directly during heat treat-
ment. If, on the other hand, the macro inclusions offer hydrogen an energetically favora-
ble storage location, cracking almost always occurs.
4.5 Damage Examples—Material 91

4.4 Hardenability

Hardenability is a very important property of steels. It is determined in the end quench

test according to DIN EN ISO 642:2000-01 [16] and is specified for case hardening and
quenched and tempered steels in the respective material standards. For tool steels and
martensitic corrosion-resistant steels, no hardenability values are given in the material
standards.
The possible scatter band of hardenability can cover a very wide range, Fig. 4.3,
which is why a limitation from +H to +HH or +HL can be useful. Company specifica-
tions often prescribe even tighter tolerances.
The possibility of calculating the hardenability by regression equations was already
pointed out in Chap. 3.

4.5 Damage Examples—Material

Example 1—Crack formation on a chain link during size calibration

Component Round steel chain link

Material: 21Mn5, Material No. 1.0495

Material 42CrMo4, 1.7225 | hardening temperature 850 °C

70

65

60 +H
+HH +HH
55

+HH
50
Hardness [HRC]

+H
45
+H
40
+HL
35 +H
+HL
30 +HL

25

20
0 5 10 15 20 25 30 35 40 45 50
Face distance [mm]

Fig. 4.3 Hardenability scatter band of the heat-treatable steel 42CrMo4 [17]
92 4 Material Influence Area

Error pattern Crack formation during size calibration

Heat treatment Normalizing rolling

One of the specific and significant properties of steel is its recyclability, which also
makes scrap a valuable raw material for steel production. In electric steel production,
steel scrap is melted down and the scrap sorting has a major influence on elements that
cannot be removed by metallurgical means. These include the elements copper and
nickel. Typical contents of these elements are listed in Table 4.1. During further process-
ing into semi-finished products, copper-nickel segregation can occur. During annealing
under oxidizing atmosphere, iron oxidizes, but the elements copper and nickel do not.
This leads to an enrichment of these two elements on the surface and creates a potential
gradient to the base material. This results in a diffusion of the two elements along the
grain boundaries. The high copper and nickel contents weaken the grain boundaries and
thereby increase the probability of cracking.
A total of 13 steels for welded round steel chain links are standardized in DIN
17115:2012-07 [18]. In the following example, rods ∅ 22 mm made of the chain steel
21Mn5, material no. 1.0495, were cold bent into round steel chain links. The chain links
were in the +N condition with a tensile strength of 740 MPa.
After cold bending, the chain links were dimensionally calibrated and during this
manufacturing step, several cracks occurred, originating from the inside, Fig. 4.4.
The examination of the chemical composition and the tensile strength showed a match
with the specifications of the standard. The contents of the elements copper of 0.18%
and nickel of 0.08% indicated an electric steel production. Both elements were each

Fig. 4.4 Broken round steel chain link

4.5 Damage Examples—Material 93

Fig. 4.5 Microstructure formation in the base material and on the surface (a) Ferrite-pearlite
microstructure in the delivery condition; (b) Fine-grained ferrite-pearlite microstructure; (c) Cold
forming structures at the crack opening; (d) Copper-nickel-enriched grain boundaries (e) Further
crack formations; (f) Copper-nickel-enriched grain boundaries

below the permissible limit contents of 0.25% for copper and 0.30% for nickel. The fine-
grained ferrite-pearlite structure confirmed the delivery condition +N, Fig. 4.5a, b.
Cold forming features were recognizable at the crack initiation position. After etch
contrasting with 3% nitric acid, the copper- and nickel-enriched grain boundaries are
94 4 Material Influence Area

difficult to recognize, Fig. 4.5c, e. Only the contrasting with the etchant according to
Klemm made the grain boundary segregation very visible, Fig. 4.5d, f.
At the grain boundaries of the crack initiation position, a copper content of 3.4% and
a nickel content of 2.8% were measured by energy-dispersive X-ray analysis. During
size calibration, these weakened grain boundaries cracked and initiated a crack or frac-
ture.
The copper and nickel contents of this steel melt were by no means unusually high.
Only a normal annealing under oxidizing atmosphere caused the damaging grain bound-
ary segregation. DIN 17115:2012-07 states that products with diameters up to 26 mm
must be able to be cold bent 180° without cracks on the tension and compression side.
Proof can be agreed upon at the time of inquiry and order. Since only a small number of
the chain links were broken or cracked, this sample proof does not guarantee the total
quantity. Therefore, a 100% crack inspection may be necessary after all manufacturing
steps have been completed.
An increase in process safety during normal annealing can be achieved by optimizing
the gas-air mixture with the aim of a low oxygen content. This reduces scale formation,
which leads to a lower enrichment of the elements copper and nickel.

Example 2—Chipping on a guide rail after induction hardening

Component: Linear guide rail

Material: 58CrMoV4, Material No. 1.7792
Error pattern: Chipping after induction surface hardening
Heat treatment: Inductive surface hardening

The material 58CrMoV4 is not included in the standard for quenched and tempered
steels, but it is listed under material no. 1.7792 as machine construction steel. The raw
material was cold drawn after hot rolling and was in the +AC state after a GKZ anneal-
ing. From this, linear guide rails were manufactured and the ball guide tracks were
inductively hardened, for which this steel is very suitable. Immediately after the induc-

Fig. 4.6 Material breakouts in the inductively hardened track

4.5 Damage Examples—Material 95

tive hardening, material breakouts were detected at several points in a track, Fig. 4.6. The
breakouts were in a line in the axial direction. Therefore, tempering of this defective rail
was omitted, which explains the high hardness of 850 HV.
The chemical composition showed no abnormalities. A microsection was taken per-
pendicular to a breakout point and documented in the unetched state and after contrast-
ing with 3% nitric acid, Fig. 4.7. Remnants of foreign particles could be seen at the
breakout point. The part of the foreign substance responsible for the breakout had broken
out, so the original size could no longer be determined.
The hardening microstructure in the surface layer was fine acicular and proved a
proper inductive hardening. A fully formed soft annealed microstructure was also present
in the base material.
The foreign particles were analyzed using EDX measurements and high manganese,
silicon, and oxygen contents were measured, Fig. 4.8. Typical slag components such as
calcium, magnesium, or aluminum could not be detected. Therefore, the foreign sub-
stances are likely to be incompletely melted ferromanganese. The steel mill did not issue
a statement. Since several breakouts were present in a line in the axial direction, it was
assumed that an extended foreign substance was present in the longitudinal direction.

Example 3—Excessive hardness in axle journal blanks after quenching and temper-
ing

Component Forged axle journal blanks

Material C53G, Material No. 1.1213
Error pattern after quenching and tempering Excessive hardness after tempering
Heat treatment Quenching and Tempering Tempering

Forged axle journal blanks made of the material C53G, which was standardized as
steel for surface hardening in the now withdrawn DIN 17212:1972-08 [19], were pre-
quenched and tempered in a pusher type furnace. The hardening temperature was slightly

Fig. 4.7 Microsection position perpendicular to the breakout

96 4 Material Influence Area

Spectrum 1
Element contents [ mass%]
Oxygen: 21.7
Aluminium: 0.3
Silicon: 8.6
Calcium: 0.3
Vanadium: 1.4
Chromium: 5.2
Manganese: 19.0
Iron: Rest

Scale range 668 cts Cursor: 0.000

Fig. 4.8 Backscattered electron image, element mappings, and EDX measurements at the break-
out point

above the recommended hardening temperature range of 815–845 °C. After the temper-
ing treatment was completed, the blanks were blast cleaned and finally the hardness was
measured. For this purpose, the surface was ground to avoid the influence of decarburiza-
tion. A hardness of 280 HBW was measured, which was within the target specification of
260–310 HBW. The batch was approved and sent to the automotive supplier.
During the incoming inspection at the automotive supplier, a hardness of 330 HBW
was measured and the batch was rejected because this measurement was above the per-
4.5 Damage Examples—Material 97

Fig. 4.9 Different hardness test positions on the axle journal blank

Fig. 4.10 Macroscopic representation of the center segregation

missible maximum hardness. Upon review, it was found that measurements were taken
at different positions. The heat treatment plant measured the hardness on the surface and
the automotive supplier measured it in the core, Fig. 4.9. The influence of decarburiza-
tion was not present. This result was initially unusual, as a higher quenching speed is
present on the surface than in the core. From this situation, a lower core hardness would
have been expected.
The explanation for this result came from the metallographic examination. Even at
a macroscopic view of the microsection surface, a pronounced center segregation was
noticeable, Fig. 4.10.
In the surface area, a ferrite-pearlite microstructure with net-like ferrite and Widman-
stätten ferrite was present, Fig. 4.11. The core microstructure, on the other hand, showed
large proportions of martensitic quenching and tempering microstructure, Fig. 4.12.
98 4 Material Influence Area

Fig. 4.11 Ferrite-pearlite microstructure at the edge with Widmanstätten ferrite

Fig. 4.12 Large proportions of martensitic tempering microstructure in the core area

To visualize the hardness differences, HV 0.2 hardness progression measurements

were made perpendicular to the segregation lines. In this process, 20 hardness impres-
sions are evaluated at intervals of 0.1 mm and the max. hardness difference ΔHVmax of
the 20 measurements is determined.
In the surface area, an average hardness of 296 HV 0.2 and a max. hardness differ-
ence of ΔHVmax = 41 HV 0.2 was determined, Fig. 4.13.
The core area showed not only a higher hardness of 336 HV0.2, but the hardness dif-
ference was with ΔHVmax = 106 HV02 more than twice as large, Fig. 4.14.
Since the differences between edge and core are unavoidable due to segregation, the
determination of the test position is of particular importance. The heat treatment com-
pany acted correctly with the near-surface hardness test, as no specific test position was
specified. Center segregations are due to the raw material and cannot be eliminated dur-
ing tempering. They are also unpredictable for a heat treatment company. To avoid these
different test results, the client must clearly define the test position.
4.5 Damage Examples—Material 99

800

700
HVmax = 41 HV 0.2
600

500
Hardness [HV 0.2]

400

300

200

100

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Distance [mm]

Fig. 4.13 Homogeneity assessment in the edge area

800

700
HVmax = 106 HV 0.2
600

500
Hardness [HV 0.2]

400

300

200

100

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Distance [mm]

Fig. 4.14 Homogeneity assessment in the core area

100 4 Material Influence Area

Example 4—Strongly fluctuating impact toughness values on a sheet

Component Sheet 27 mm thick

Material P460NH, Material No. 1.8935
Error pattern Strongly fluctuating impact toughness values
Heat treatment Normalizing rolling

The determination of the impact toughness of a sheet with 23.5 mm sheet thickness
made of the material P460NH led to very different results. In the surface area, the impact
work was more than twice as high at 66 J as in the core area, which showed 27 J.
The microstructure examination revealed a very pronounced center segregation in this
sheet. The edge area showed a lamellar ferrite-pearlite microstructure, Fig. 4.15. In the
core area, on the other hand, the pearlite content was considerably larger and there was
lamellar arranged martensite, Fig. 4.16.

Fig. 4.15 Lamellar ferrite-pearlite microstructure at the surface

Fig. 4.16 Higher pearlite contents and martensite in the core area
4.5 Damage Examples—Material 101

600

500

HVmax = 43 HV 0.2

400
Hardness [HV 0.2]

300

200

100

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Distance [mm]

Fig. 4.17 Homogeneity assessment in the edge area

The significantly different microstructure formations led to a large difference in

the homogeneity assessment. Initially, hardness gradients perpendicular to the seg-
regation lines were created with a small test load of HV0.2, Fig. 4.17 and 4.18. The
edge area showed very good homogeneity with ΔHVmax = 43 HV0.2. The core
area was much more inhomogeneous and showed a very large hardness scatter with
ΔHVmax = 192 HV0.2.
The hardness gradient measurements for homogeneity assessment reflect the micro-
structures and their distribution. The cause of these microstructure and hardness differ-
ences are segregations that occur during continuous casting solidification. To quantify
the segregations, segregation coefficients were determined using energy-dispersive X-ray
analysis in the scanning electron microscope. Initially, the chemical composition of the
sheet was determined in the emission spectrometer, Table 4.2. For the determination of
the segregation coefficients, the two elements manganese and nickel are important.
The hardness impressions served as orientation points and the EDX measurements
were carried out at the highest and lowest hardness values respectively. The segregation
coefficient was determined using the following equation.
max. Elementgehalt
Ki = [%]
min. Elementgehalt
102 4 Material Influence Area

600

500

HVmax = 192 HV 0.2

400
Hardness [HV 0.2]

300

200

100

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Distance [mm]

Fig. 4.18 Homogeneity assessment in the core area

Table 4.2  Results of the emission spectroscopic analysis determination

Target analysis according to C Si Mn P S Cr Ni
DIN EN 10028-3:2009 % % % % % % %
max. max. 1.10 max. max. max. max.
0.20 0.60 1.70 0.025 0.010 0.30 0.80
Sheet 27 mm 0.18 0.32 1.54 0.010 0.0004 0.040 0.65
Target analysis according to Mo Al N Nb Cu V Ti
DIN EN 10028-3:2009 % % % % % % %
max. min. max. max. max. max. max.
0.10 0.020 0.025 0.050 0.70 0.20 0.030
Sheet 27 mm 0.010 0.024 0.010 0.001 0.02 0.16 0.004

In the edge area, the segregation coefficients for the elements manganese and nickel were
very close to the ideal value of 100% and the edge area was thus practically free of seg-
regation, Table 4.3.
In contrast, the core area showed very clear differences for both the element man-
ganese, 184.4%, and the element nickel, 155.1%, see Table 4.4. A t8/5-time calculated
from the respective analysis for the upper critical cooling speed resulted in a value of 5 s
4.5 Damage Examples—Material 103

Table 4.3  Segregation coefficients of the elements manganese and nickel in the edge area
Measurement position/measurement Mn Ni
A B A B
Initial state edge area Positively segregated area 1.68 1.44 0.74 0.53
Average value 1.560 0.635
Negatively segregated area 1.62 1.54 0.75 0.53
Average value 1.580 0.640
Segregation coefficient 101.3 100.8

Table 4.4  Segregation coefficients of the elements manganese and nickel in the core area
Measurement position/measurement Mn Ni
A B A B
Initial state core area positively segregated area 2.72 2.70 0.93 0.73
Average value 2.710 0.830
negatively segregated area 1.43 1.51 0.52 0.55
Average value 1.470 0.535
Segregation coefficient 184.4 155.1

for the negatively segregated area and a value of 22 s for the positively segregated area.
These differences explain the respective microstructure formation.
The very different results of the notch impact work between the edge and core area
were therefore clearly caused by the intense segregations. A diffusion annealing could
achieve a certain homogenization, which, however, is unusual for these steel grades.

Example 5—Crack indication on a piston

Component Piston
Material: 42CrMo4, Material No. 1.7225
Error pattern 12 mm long crack indication during the Magnetic particle inspection
Heat treatment Quenching and tempering and nitrocarburizing

To increase wear resistance, a piston made of quenched and heat-treatable steel

42CrMo4, 1.7725, was nitrocarburized in gas. In the process, a uniform compound layer
of approximately 10 μm was formed, Fig. 4.19. The microstructure in the base material
corresponded to a highly tempered quenched and tempered microstructure.
In the final magnetic particle inspection, a 12 mm long defect was indicated. The met-
allographic examination of this defect revealed an oxidic non-metallic inclusion, around
which a compound layer had also formed, Fig. 4.20.
104 4 Material Influence Area

Fig. 4.19 Uniformly formed compound layer

Fig. 4.20 Defect on the piston surface

To identify the oxidic non-metallic inclusion, an EDX examination was conducted.

The result of 23% oxygen, 1.7% magnesium, 21% aluminum, 1% silicon, 24% calcium,
remainder iron confirmed a slag inclusion.

Example 6—Insufficient core strength in forged chain links

Component Forged chain linkswith a wall thickness of 8 mm

Material: 30MnB5, Material No.: 1.5531
Error pattern Insufficient core strength
Heat treatment Quenching and tempering Quenching

The error pattern to be described below is a typical example of an order and delivery that
conform to standards, but with a missing analysis supplement. The material 30MnB5,
1.5531, was standardized in DIN EN 10083-3:2006-10 [20] and the raw material was
ordered with reference to this standard. In Table 4.5, the measurement result is compared
4.5 Damage Examples—Material 105

Table 4.5  Results of the emission spectroscopic analysis determination

Target analysis according to C Si Mn P S Cr B
DIN EN 10083-3:2006 % % % % % % ppm
0.27 max. 1.15 max. max. – 8
0.33 0.40 1.45 0.035 0.040 – 50
Measurement results 0.30 0.25 1.22 0.014 0.021 0.26 22
Trace elements Mo Ni N Nb Cu Al Ti
% % % % % % %
Measurement results 0.010 0.002 0.007 0.001 0.01 0.021 0.004

with the target values of the standard specification. All analysis ranges specified in the
standard were met and therefore this batch corresponded to the standardized material
30MnB5, 1.5531 in terms of analysis.
The forged chain links were tempered with a target specification of 1050–1200
MPa. The verification of the target specification was carried out by a hardness test on
the surface according to Brinell with subsequent conversion according to DIN EN ISO
18265:2004-02, Table B2 [21]. From the average value of the hardness test of 335 HBW,
a tensile strength of 1065 MPa resulted. The result was at the lower limit of the target
specification.
After a chain link failed, a metallographic examination was carried out. Near the sur-
face, there was a predominantly martensitic quenching and tempering microstructure
with minor bainite components, Fig. 4.21a. In the core area, a mixed microstructure of
ferrite, pearlite, bainite, and tempered martensite was present, Fig. 4.21b.
A hardness measurement in the core area resulted in an average value of 292 HBW
or a tensile strength of 928 MPa after conversion. The mixed microstructure not only led
to a lower tensile strength but also reduced the yield strength ratio. A tensile test with a
direct determination of the yield strength ratio was not possible due to the geometry.

Fig. 4.21 Microstructure (a) in the near-surface area and (b) in the core area
106 4 Material Influence Area

Without sufficient titanium content, there is a risk of boron nitride formation, which
causes the alloying element boron to lose its effectiveness as a hardenability-enhancing
alloying element. Titanium forms titanium carbonitrides and the boron remains dissolved
in the lattice. Information about minimum titanium contents is also not included in the
current DIN EN ISO 683-2:2018-09 [17].

Example 7—Edge breakouts on a gear ring during mechanical processing

Component Housing gear ring2nd planetary stage

Material: EN-GJS-800-10, Material No. 5.3400 (ADI 800)
Error pattern Edge breakouts during mechanical processing
Heat treatment Austempering(Microstructure of ferrite and austenite)

Austempered ductile iron grades with spheroidal graphite are standardized in DIN EN
1564:2012-01 [22]. These are alloyed cast iron grades, but their chemical compositions
are not specified in the standard. The standard only states that the required microstruc-
ture is produced by choosing the appropriate composition and by the subsequent process
steps. Only the mechanical properties, which are determined on samples, are specified.
A housing gear ring made of the cast material EN-GJS-800-10, 5.3400, was mechani-
cally processed after the heat treatment to form the austempered microstructure. During
the processing, edge breakouts occurred on the end faces of the gearing of the second
planetary stage, Fig. 4.22. The end faces themselves were not processed and still had the
cast surface.

Fig. 4.22 Documentation of the complained breakout points on the end faces
4.5 Damage Examples—Material 107

Table 4.6  Results of the analysis determination

C Si Mn P S Cr Ni Mo Mg Al Cu
% % % % % % % % % % %
3.22*) 2.00 0.24 0.017 0.010*) 0.06 1.41 0.15 0.023 0.025 0.73
All data in mass-% *) Determination by combustion analysis

Table 4.7  Results of the hardness test

Measured hardness values [HBW Aver- b Cor- UCorr Test result
2.5/187.5] age [HBW] rected [HBW] [HBW]
1st 2nd 3rd 4th 5th value average
value value value value value [HBW] value
[HBW]
292 298 299 300 293 296 -2.7 299 ± 4.4 299 ± 4.4

The verification of the chemical composition was carried out for the carbon and sulfur
content by a combustion analysis, and the alloy elements were determined in the optical
emission spectrometer, Table 4.6. A comparison with target values was not possible, as
the specifications were not available.
The results of the hardness test are listed in Table 4.7. For this cast material, guideline
values for Brinell hardness are informatively given in DIN EN 1564:2012-01. The hard-
ness of 299 HBW was within the guideline range of 250–310 HBW.
In the metallographic examination, a microsection each was taken parallel to the
tooth flank from the tooth with the breakage area and a tooth behind it. The breakage site
showed numerous casting defects and the second tooth flank also showed similar defects,
Fig. 4.23.
In the area of the casting defects, the graphite formation was lamellar and two places
were documented as examples, Fig. 4.24. The observed defect occurs more frequently in

Fig. 4.23 Casting defects on the front side of the damaged tooth flank and a tooth flank behind it
108 4 Material Influence Area

Fig. 4.24 Lamellar graphite in the area of the casting defects

spheroidal graphite cast iron and is referred to as dross in English terminology. Extensive
measures are required to combat this type of defect, which are described in detail in [23].
Outside the area affected by defects, the formation of spheroidal graphite was in
accordance with the specification and the ausferritic microstructure was also flawless,
confirming a proper isothermal transformation, Fig. 4.25.
Given the multitude of casting defects, there was a high probability that breakouts
would occur during mechanical processing. The defects were not limited to the breakout
points, but were likely present on all end faces. Although there were several breakout
points, only one tooth flank without breakouts was examined metallographically.

Fig. 4.25 Spheroidal graphite form VI, size 7-8 according to ISO 945 [24] (left) and ausferrite
(right)
4.5 Damage Examples—Material 109

Example 8—Defective FP microstructure on rods

Component Rods Ø 90 mm
Material: 18CrNiMo7-6, Material No. 1.6587
Error pattern Defective FP structure
Heat treatment Annealing to ferrite+pearlite (+FP)

A classic normal annealing is not possible with the case-hardening steel 18CrNiMo7-6,
1.6587, as a pure ferrite-pearlite microstructure cannot be achieved with air cooling. Air
cooling always results in a proportion of a hardening microstructure, Fig. 4.26. A con-
trolled furnace cooling to avoid the hardening microstructure is not economical and does
not lead to the desired result. This results in a strong segregation of the ferrite and the
pearlite. For this reason, these steels are rapidly cooled to 600–650 °C after austenitiza-
tion, followed by isothermal transformation at this temperature, Fig. 4.27.
The hot-rolled bars were delivered to a heat treatment plant for FP annealing, and the
treatment was carried out as shown in Fig. 4.27. While a line-like ferrite-pearlite micro-
structure without hardening microstructure was present near the surface, Fig. 4.28, there
was already a significant proportion of bainite + martensite 20 mm below the surface,
Fig. 4.29. This microstructure formation was complained about, as tool breakage was
feared during mechanical processing. With the very large hardness differences, this fear

1200

1100
18CrNiM7-6, 1.6587, bar 90 mm, austenitization 870 °C 20 min, air cooling
1000

900
Ac3 (825 °C)
800
Temperature [°C]

700 75 Ac1 (745 °C)

5 Ferrite 70
40 Perlite 25
600 Austenite 2

500
Bainite
Ms (395 °C) 3 90
400
90
300 Martensite
55
25
200

100
431 362 324 271 208 158
0
0.1 1 10 100 1000 10000 1E5 1E6
Time [sec]

Fig. 4.26 Continuous CCT diagram and FEM-calculated air cooling for a round bar of Ø 90 mm
110 4 Material Influence Area

1200

1100
18CrNiM7-6, 1.6587, bar 90 mm, austenitization 900 °C 10 min, isothermal transformation
1000

900
Ac3 (820 °C)
800
Temperature [°C]

700 Ac1 (730 °C)

Ferrite 170
600 Perlite
erlit 179 HV
Austenite
500
Bainite
400 Ms (395 °C)
308

300 Martensite
200

100

0
0.1 1 10 100 1000 10000 1E5 1E6
Time [sec]

Fig. 4.27 Isothermal TTT diagram and FEM-calculated isothermal transformation for a round bar
of Ø 90 mm

Fig. 4.28 Strongly line-like ferrite-pearlite microstructure in the near-surface area of the round
bar Ø 90 mm

was also justified, as could be seen from the hardness scans at various positions, Fig.
4.30.
While at the edge-near position with a hardness difference of max. 62 HV 0.2 units
and a max. hardness of 212 HV0.2, expected results were achieved, already 20 mm
below both much larger hardness differences and significantly higher maximum values
were present. In the core area, hardness values of up to 424 HV0.2 were even measured,
causing a hardness difference of 294 HV0.2, Fig. 4.30.
4.5 Damage Examples—Material 111

Fig. 4.29 Hardening microstructure at a distance of 20 mm below the surface of the round bar Ø
90 mm

To find the cause, a sample was taken from a position 20 mm below the surface and
isothermally transformed in the perlite stage in a quenching dilatometer. The holding
time at 640 °C was extended to 6 hours. After the holding time had elapsed, the sample
was cooled in free air convection. Even after this significantly longer holding time, a
complete transformation to ferrite-pearlite was not possible, Fig. 4.31.
The results prove that due to segregation, no complete isothermal transformation is
to be expected even with a significantly extended holding time. The heat treatment plant
has therefore not carried out any faulty treatment. With the delivered bars, the stated goal
of a +FP annealing could not be achieved.

800

700
Close to the
HVmax = 62 HV 0.2
surface
600

500
Hardness [HV 0.2]

400

300

200

100

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Distance [mm]

Fig. 4.30 Hardness scans at various positions of the round bar Ø 90 mm

112 4 Material Influence Area

800

700
20 mm deeper HVmax = 214 HV 0.2
600

500
Hardness [HV 0.2]

400

300

200

100

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Distance [mm]

800

700
Core area HVmax = 294 HV 0.2

600

500
Hardness [HV 0.2]

400

300

200

100

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Distance [mm]

Fig. 4.30 (continued)

This is now also explicitly mentioned in the current DIN EN ISO 683-3:2019-04 [25].
A pre-hardening with a high tempering temperature, on the other hand, would lead to a
much more uniform hardness and a more homogeneous microstructure. However, this
heat treatment process is more costly.
4.5 Damage Examples—Material 113

Fig. 4.31 Microstructure formation after isothermal transformation with 6 h holding time

Example 9—Tooth Breakages on a Crown Wheel

Component Crown Wheel Ø 465 mm

Material: 24CrMo5, Material No. 1.7258
Error Image Tooth Breakages
Heat Treatment Case Hardening and Tempering

The premature failure of gear components is in many cases caused by macroscopic non-
metallic inclusions. Very often these are inclusions whose size is suitable to start a crack,
but their probability of detection, even with elaborate ultrasonic testing, is very low.
After a short running time, a crown wheel showed breakages on several teeth, two of
which are depicted in Fig. 4.32. In both cases, it was recognizable on the fracture surface
that a non-metallic inclusion was present at the center of the crack initiation, Fig. 4.33a–d.
To identify the type of inclusion, element distribution images were created using
energy-dispersive X-ray analysis, Fig. 4.34a–d. The elements aluminum, oxygen, cal-
cium, and magnesium showed a geometrically identical distribution image. The men-
tioned elements are typical for slag residues. The element distributions of the second
non-metallic inclusion were identical, so a repeating representation is omitted.
All further investigation results were inconspicuous and are therefore not listed.

Example 10—Cracking of a camshaft during installation

Component Camshaft
Material: 18CrNiMo7-6, Material No. 1.6587
Error pattern Cracking during installation
Heat treatment Case hardening and tempering
114 4 Material Influence Area

Fig. 4.32 Breakages on two teeth of a crown wheel

Fig. 4.33 Non-metallic inclusions in the center of the crack lens (a) inclusion size
730 × 520 μm; (b) inclusion size 290 × 55 μm; (c) secondary electron image; (d) secondary elec-
tron image
4.5 Damage Examples—Material 115

Fig. 4.34 Element distribution images of the non-metallic inclusion (a) aluminum; (b) oxygen;
(c) calcium; (d) magnesium

During engine assembly, intensive cracking was discovered on a camshaft, Fig. 4.35a. At
the center of the crack initiation, there was a macro inclusion, Fig. 4.35b. Starting from
this non-metallic inclusion, the crack progression lines were clearly visible.
In Fig. 4.36a, the macro inclusion is shown in the secondary electron image. The
backscattered electron image makes it easier to locate non-metallic inclusions in the
crack surface, Fig. 4.36b.
The EDX spectrum of this inclusion reveals its origin from the slag due to the magne-
sium content, Fig. 4.36c.

Example 11—Secondary damage to planet gears after tooth breakage

Component Planet gearsfrom a wind turbine

Material: 18CrNiMo7-6, Material No. 1.6587
Error pattern Secondary damage after a tooth breakage
Heat treatment Case hardening and tempering

In a planetary gear system of an older wind turbine, all three planet gears were deliv-
ered for examination in a very heavily damaged state, Fig. 4.37a. After a primary dam-
age event, massive secondary damage to all gear components in the gearbox occurred,
116 4 Material Influence Area

Fig. 4.35 Crack in a camshaft and opened crack surface (a) Crack formation; (b) Crack surface
with crack initiation position

Fig. 4.36 Secondary- (a) and backscattered electron image (b) as well as EDX spectrum of the
measurement position marked in (b) (c) of the non-metallic inclusion
4.5 Damage Examples—Material 117

Fig. 4.37 (a) Heavily damaged planet gears; (b) Secondary damage to the tooth flanks

Fig. 4.37b. In modern gearboxes, monitoring is carried out using acoustic emission. This
provides the possibility to bring a gearbox to a controlled stop in the event of a crack,
even before a tooth breaks off.
In the case of such heavily damaged tooth flanks, the primary tooth breakage is rarely
identifiable, as it is usually also destroyed by the secondary damage.
The second example, unlike the previous case, clearly shows the fatigue fracture sur-
face with numerous beachmarks, even though this planet gear was operated for a longer
time after the crack started, Fig. 4.38. At the crack start, there was also a non-metallic
influence, which was heavily worn due to the relative movement of both crack surfaces.
Here too, the examination was continued in the scanning electron microscope and ele-
ment distribution images were created. The representation of the backscattered electron
image is particularly suitable for worn inclusions, Fig. 4.39a, b. Almost no contrast for-
mation is recognizable in the secondary electron image.
The type of inclusion was again determined by EDX analyses. A macroscopic alu-
minum oxide inclusion was present, Fig. 4.39c, d.

Fig. 4.38 Single tooth breakage with clear beachmarks (a) Breakout at a single tooth; (b) Fatigue
fracture with origin at a macro inclusion
118 4 Material Influence Area

Fig. 4.39 Element distribution images of a worn non-metallic inclusion (a) Secondary electron
image; (b) Backscattered electron image (c) Element distribution aluminum; (d) Element distribu-
tion oxygen

Example 12—Flake formation on a gear ring

Component Gear ring

Material: 42CrMo4, Material No. 1.7225
Error pattern Flake formation
Heat treatment Quenching and tempering Tempering

During the crack test after mechanical processing, several cracks were discovered in the
tooth base area of a gear ring, Fig. 4.40. The gear ring was made from the heat-treatable
steel 42CrMo4, 1.7225, and was in the pre-quenched and tempered state. The intended
inductive surface hardening had not yet taken place.
The examination of the chemical composition showed no abnormalities of the speci-
fied alloy elements. However, when determining the hydrogen content below the tooth
4.5 Damage Examples—Material 119

Fig. 4.40 Segment of the gear ring and 2 mm long crack in the tooth base

base, a strongly fluctuating total hydrogen content of 0.3 to 1.8 ppm was measured dur-
ing the melt extraction. The proportion of diffusible hydrogen was 0.3 ppm.
The microsection for the metallographic examination was taken perpendicular to the
crack formation. There was a quenching and tempering microstructure with clear segre-
gations. The crack ran in a positively segregated line, Fig. 4.41.
The investigations were continued with a tensile test and the results of the tensile test
are shown in Table 4.8. While the values for the 0.2% yield point and tensile strength
safely met the requirements of DIN EN 10083-3:2007-01, the elongation only reached
the minimum value and the necking was significantly lower than required.
The macroscopic examination of the fracture surface revealed the white spots typical
for flakes, Fig. 4.42. When examining the flakes in the scanning electron microscope,
the typical image of a “fish eye” was shown, Fig. 4.43. This fish eye was characterized
by a transcrystalline fracture structure, which was arranged concentrically around a non-
metallic globular inclusion and clearly stood out from the rest of the fracture surface.
The splitting at the globular inclusion was clearly visible.

Fig. 4.41 Cut crack and crack run-out in a positively segregated line
120 4 Material Influence Area

Table 4.8  Results of the tensile test

Sample dimensions Test results
Diameter Cross section Measuring length Rp0,2 Rm A Z
[mm] [mm2] L0 [mm] [MPa] [MPa] [%] [%]
Requirements according to DIN EN 10083-3:2007- min. 500 750–900 min. 14 min. 55
01
for the thickness 100 < t <=160 mm
9.98 78.23 50 619 826 14 25

Fig. 4.42 Fracture surface of the tensile specimen with noticeable white spots

Example 13—Fracture on a fork tine during straightening

Component Fork tine

Material: 36MnB4, Material No. 1.5537
Error pattern Fracture during straightening
Heat treatment Quenching and tempering Tempering

The boron-alloyed steel 36MnB4, 1.5537, was used to manufacture fork tines. For this
purpose, rectangular dimensions of 200 × 90 mm were ordered from a steel dealer,
which originated from continuous casting production. As part of the manufacturing pro-
cess, flame cutting was carried out to create the taper at the fork tip (tapering). This was
4.5 Damage Examples—Material 121

Fig. 4.43 Transcrystalline fracture surface at a globular non-metallic inclusion and honeycomb
fracture on the adjacent fracture surface

followed by quenching and tempering. Due to the unavoidable distortion, the fork tines
had to be straightened after tempering. During this process, a fracture of the fork tine
occurred.
Pores were visible on the underside of the taper at the flame cutting surface and at a
separation cut, Fig. 4.44. The metallographic examination showed strong selective edge
oxidation and pronounced copper-nickel segregation at the former austenite grain bound-
aries, Fig. 4.45.
Optical emission spectroscopy determined a copper content of 0.19% and a nickel
content of 0.13%. EDX analyses at the grain boundaries revealed copper contents of
up to 2.3% and nickel contents of up to 5.4%. Since the quenching and tempering was
carried out under protective gas, both the oxidation at the pore surfaces and the copper-
nickel segregation occurred during flame cutting.
This Cu-Ni segregation occurs when iron is oxidized at high temperatures and the
oxidation-stable elements copper and nickel thus accumulate on the surface and diffuse
into the edge zone via the grain boundary.

Fig. 4.44 Pores on the taper surface and in the separation cut
122 4 Material Influence Area

Fig. 4.45 Strong edge oxidation and copper-nickel segregation at pores

The pores in combination with the Cu-Ni segregation caused a strong notch effect
during straightening, which led to the fracture. From the 3.1 test certificate of the steel
producer, it was found that the hot forming degree was 1.9 times. This low degree of
hot forming is insufficient for a complete closure of the shrinkage cavities. The forming
degree should be at least 3.5 times.

References

1. Stahl-Eisen-Liste 11. Auflage, Verlag Stahleisen, Düsseldorf 2003. ISBN 3-514-00701-2

2. Datenbank StahlDat https://matplus.de/de/stahldat-sx/
3. Bleck, W. und E. Moeller: Handbuch Stahl, Auswahl, Verarbeitung, Anwendung. Carl Hanser
Verlag München 2018. ISBN 978-3-446-44961-9
4. Hegemann, K-R. und R. Guder: Stahlerzeugung. SpringerLink 2020, ISBN 978-3-658-29091-
7
5. Oeters, F.: Metallurgie der Stahlherstellung. Springer-Verlag Berlin, Heidelberg, New York,
London, Paris, Tokyo, Hong Kong, Verlag Stahleisen Düsseldorf 1989, ISBN 978-3-642-
51165-3
6. Jahrbuch Stahl 2017, Verlag Stahleisen GmbH, Düsseldorf, ISBN 978-3-514-00825-0
7. Bleck, W.; Pariser, G.; Trute, S.: Herstellung und Verarbeitung moderner Stahlwerkstoffe.
HTM Z. Werkst. Waermebeh. Fertigung 58. (2003) 4, S. 181–188
8. Bleck, W.: Werkstoffkunde Stahl für Studium. Verlag Mainz, Wissenschaftsverlag Aachen.
ISBN 3-89653-820-9
9. DIN 50602:1985-09: Mikroskopische Prüfung von Edelstählen auf nichtmetallische Ein-
schlüsse mit Bildreihen
10. ISO 4967:1998-10-1: Determination of content of nonmetallic inclusions – Micrographic
method using standard diagrams
11. ASTM E-45a: 2018: Standard Test Methods for Determining the Inclusion Content of Steel
12. DIN EN 10247:2017-09: Metallographische Prüfung des Gehaltes nichtmetallischer Ein-
schlüsse in Stählen mit Bildreihen
References 123

13. Buckstegge, J.: Erfahrungen beim Einsatz der Ultraschall-Tauchtechnik zur Bestimmung des
Makroreinheitsgrades von Wälzlagerstählen. HTM Härterei-Techn. Mitt. 54 (1999) 4; S. 250–
258
14. Hering, J.: Ultraschall-geprüfte Stahlreinheit – Verfahren für die Praxis, Ergebnisse aus der
Praxis. HTM Härterei-Techn. Mitt. 54 (1999) 3; S. 259–265
15. Thiery, D.; Delhaes, Chr.: Die Ultraschall-Tauchtechnik als Mittel zur Bestimmung des mak-
roskopischen Reinheitsgrades von Stählen. HTM Härterei-Techn. Mitt. 55 (2000) 3; S. 160–
165
16. DIN EN ISO 642:2000-01: Stahl – Stirnabschreckversuch (Jominy-Versuch)
17. DIN EN ISO 683-2:2018-09; Für eine Wärmebehandlung bestimmte Stähle, legierte Stähle
und Automatenstähle – Teil 2: Legierte Vergütungsstähle
18. DIN 17115:20212-07: Stähle für geschweißte Rundstahlketten und Ketten-Einzelteile
19. DIN 17212:1972-08: Stähle für Flamm- und Induktionshärten; Gütevorschriften
20. DIN EN 10083-3:2006-10: Vergütungsstähle – Teil 3: Technische Lieferbedingungen für legi-
erte Stähle
21. DIN EN ISO 18265:2004-02: Metallische Werkstoffe – Umwertung von Härtewerten
22. DIN EN 1564:2012-01: Gießereiwesen – Ausferritisches Gusseisen mit Kugelgraphit
23. Haase, S.: Guß- und Gefügefehler. Verlag Schiele & Schön, Berlin 2003. ISBN 3-7949-0698-5
24. DIN EN ISO 945-1:2019-10: Mikrostruktur von Gusseisen – Teil 1: Graphitklassifizierung
durch visuelle Auswertung
25. DIN EN ISO 683-3:2019-04; Für eine Wärmebehandlung bestimmte Stähle, legierte Stähle
und Automatenstähle – Teil 3: Einsatzstähle
Manufacturing Influence Area Before
Heat Treatment 5

Contents

5.1 Uneven Material Removal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

5.2 Residual Stresses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
5.3 Surface Hardening. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
5.4 Non-removed Functional Layers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
5.5 Passive Layers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
5.6 Damage Examples—Manufacturing Before Heat Treatment . . . . . . . . . . . . . . . . . . . . . . . 133
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

The influences of the manufacturing processes are divided into a manufacturing before
heat treatment and one after. In both cases, only those influencing factors are described
that have led to errors or damage. It is assumed that the raw materials are in accord-
ance with specifications and free of defects, as the influencing factor material defects has
already been described in the last chapter.
The error sources listed in Fig. 5.1 provide a first overview.

5.1 Uneven Material Removal

Hot-formed raw materials naturally have scale formations, decarburizations and minor
imperfections on the surface. If, on the other hand, a decarburization-free surface is
required already in the delivery state, this can be achieved for round products by peeling.
For flat products, grinding is necessary and numerous very different dimensions can be
delivered in the ground state.

© The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, 125
part of Springer Nature 2024
P. Sommer, Mistakes Before, During and After Heat Treatment of Steel,
https://doi.org/10.1007/978-3-658-44160-9_5
126 5 Manufacturing Influence Area Before Heat Treatment

Manufacturing defects before heat treatment

Uneven removal
Residual stresses due to One-sided decarburization
mechanical processing
Warpage due to uneven Passive layers
residual stress distribution
Soft spots in the
Nitriding/nitrocarburizing
Flame cutting, laser cutting
Edge hardening Eroding
Residual stresses due to erosion
Functional coatings from
cold forming
Unwanted edge layers

Fig. 5.1 Typical error sources in manufacturing before heat treatment

The permissible decarburizations, imperfections on the surface and dimensional toler-

ances of untreated raw materials are specified for long and flat products either directly
in the product standards or generally in DIN EN 10021:2007-03 [1] or ISO 404:2013
[2]. Information for forged components can be found, for example, in DIN EN 10222-
1:2002-07 [3] or DIN EN10243-1 [4].
It sounds plausible and self-evident that the affected surface must be removed. The
same self-evidence applies to the demand that this must be done evenly on all surfaces.
If this is not the case, surfaces with different carbon contents are present. This means that
in a subsequent hardening process, the martensite formation on the decarburized surface
begins earlier than on the non-decarburized surface, which leads to an increase in the
risk of distortion. The thinner the dimension, the more severe this effect will be. Hand-
saw blades made of high-speed steels are a typical example, as the hardening process
is mainly carried out in vacuum furnaces without changing the edge composition. The
result can be bent handsaw blades with different hardness values.
Decarburizations can also occur during intermediate annealing in a decarburizing
atmosphere. Removal of the decarburization is also necessary for these treatment condi-
tions. Particularly severe property deficits occur in induction hardening if decarburized
surfaces are present. The rapid heating and short holding time does not lead to a concen-
tration balance, which is why soft surfaces are formed.

5.2 Residual Stresses

Residual stresses are, by definition, stresses that exist in a workpiece without external
load and after temperature equalization.
5.2 Residual Stresses 127

Residual Stresses in pre-material deliveries

Residual stresses may already be present in the pre-material as it is delivered. Bars that
are straightened after hot rolling and then not stress-relief heat-treated exhibit residual
stresses due to bending.
Rods with special requirements for surface quality can be produced by a peeling pro-
cess. Since peeling is a cutting process, plastic deformations on the surface are to be
expected. The machining parameters must be chosen during peeling in such a way that
friction martensite layers are avoided. However, if such layers occur, they are under such
high residual stresses that cracking can occur, Fig. 5.2.

Residual Stresses in machining

Machining is one of the most important processing methods. Soft state machining is typ-
ically done by cutting with a geometrically defined cutting edge. The machining methods
peeling, drilling, turning, and milling are the classic representatives of these machining
methods.

Grinding is a cutting process with a geometrically undefined cutting edge and is pref-
erably used in the hard state. This machining method and its effects are discussed in
Chap. 7.
The performance of cutting tools, such as in milling, has been greatly increased by
coating processes or the use of hard metals with and without cubic boron nitride (CBN).
This led to ever higher machining speeds, which did not leave the residual stress profile
of the components to be machined unaffected.
In [6] a simulation method for determining the max. temperature in front of the cut-
ting edge was described. Accordingly, at appropriate cutting speeds, temperatures can
reach well above Ac1, Fig. 5.3. This leads to the formation of austenite and subsequently
even with self-cooling to a hardening microstructure. Tensile residual stresses, which are

Fig. 5.2 Friction martensite with crack formation on the surface of a quenched and tempered rod
after peeling
128 5 Manufacturing Influence Area Before Heat Treatment

880.000 NT11 600

Tensile stresses
808.333 500

Residual stresses [Mpa]

736.667
400
Temperature [°C]

665.000
593.333 300
521.667
450.000 200
378.333
306.667
100
235.000 0
163.333
91.667 -100
20.000 -200
Max: 843.395
-300
Node: WP-1.2552 Compressive stresses
y -400
Max: 843°C 0 10 20 30 40 50 60
z t
Depth [µm]

Fig. 5.3 Calculated temperature in front of the cutting edge and resulting residual stresses [6]

Machining

Inhomogeneous plastic deformation Inhomogeneous plastic deformation

without material separation (crushing) due to material separation

by forces by forces through due to predominant

perpendicular to the perpendicular and Heat development heat development
surface parallel to the surface

Pressure - Print Tension - residual Train -

residual stresses residual stresses stresses residual stresses

Residual stresses of the crushed surface Residual stresses of the uncrushed surface

Residual stresses of the braced surface

Fig. 5.4 Classification of the residual stress generating processes in machining [7]

compensated by compressive residual stresses below the surface, arise on the surface due
to the hardening.
It must also be assumed that this residual stress profile exemplified in Fig. 5.3 is not
uniformly present on the entire surface. The residual stresses can be largely eliminated
by a proper stress relief annealing. If no stress relief annealing is carried out, the residual
stresses are only reduced when heating for hardening. This can lead to significant shape
changes, especially if the residual stresses are very differently developed.
If the cutting process leads to the formation of friction martensite, this can also lead to
crack formation in the friction martensite layer. If these cracks are not detected, because
5.3 Surface Hardening 129

the bore surface is not easily visible, for example, damage development in a subsequent
hardening treatment is practically preprogrammed.
The diversity of the causes of residual stress was described, for example, in [7] and
summarized in Fig. 5.4.

5.3 Surface Hardening

Surface hardening can occur when shaping is done by thermal processes, e.g. flame
cutting, plasma or laser cutting, or wire erosion. In all cases, these separation processes
require local melting or burning. This leads to heat input at the cutting edge and often to
hardening at the edge.
The greatest heat input is to be expected with autogenous flame cutting. A burning
gas-oxygen flame generates the necessary heat to burn the steel, and the oxidation prod-
ucts are blown out of the cutting joint by an oxygen jet [8].
Figure 5.5 shows the schematic representation of the processes during flame cutting.
The heating flame (1) locally heats the material to ignition temperature T > Tz. The cut-
ting oxygen begins to burn the material locally (2), Fig. 5.5a. The burner moves at the
cutting speed v, thereby melting and burning material. This creates the cutting joint with
the width f and the cutting length l, Fig. 5.5b.

a b

Fig. 5.5 Schematic processes during flame cutting [8]

130 5 Manufacturing Influence Area Before Heat Treatment

Similar to welding, there is therefore a heat-affected zone and hardenings are pos-
sible. In particular, steels with increased hardenability or higher carbon contents form a
pronounced hardening on the flame cut surface. This can lead to high residual stresses
and crack formation, similar to friction martensite layers. To avoid these unwanted reac-
tions, preheating of the flame cut surfaces is indicated in these cases.
The heat input during plasma and plasma cutting is lower, but even with these thermal
cutting processes, hardening occurs if the steel has sufficient hardenability.
The wire erosion or spark erosion cutting with wire completes the processes of ther-
mal cutting. This process can be used regardless of the hardness of the steel. Therefore,
the range of applications is very large. Similar to all other described thermal cutting pro-
cesses, there is also a heat-affected zone here, but it is much smaller compared to the
other processes. Nevertheless, new hardening layers can also form on the cut surface
here.
The residual stresses occurring in all thermal machining processes can be largely
reduced by stress relief annealing, as with the cutting processes. If this annealing treat-
ment is omitted, the residual stresses are reduced during subsequent austenitization. This
can lead to changes in dimensions and shape and in borderline cases also to crack forma-
tion. When assessing damage after heat treatment, direct evidence of too high or une-
venly distributed residual stresses is no longer possible. The aim here is to convince all
parties involved that stress relief annealing with subsequent minor fine machining is the
best possible prerequisite for low-distortion hardening.

5.4 Non-removed Functional Layers

To improve the cold formability of wires, profiles, and sheets, they are often coated with
a phosphate layer. In addition to improving cold formability, this usually also improves
corrosion resistance. A conversion layer of metal phosphates is formed, which due to its
topography is very receptive to oils or lubricants. Zinc and manganese phosphating are
used.
The zinc phosphate layers are treated with various lubricants and soaps in addition to
their good sliding and lubricating properties, depending on the application. This serves to
increase the operating temperature of these phosphate layers and to improve their sliding
properties.
After cold forming, the phosphate layers must be removed again before a protec-
tive gas heat treatment, as the zinc phosphate layer decomposes to form metallic zinc in
neutral and reducing furnace atmospheres. The zinc transitions to the gas phase at suf-
ficiently high furnace temperatures due to its low boiling point of 907 °C and deposits at
colder locations. The remaining PO4 radical is reduced to phosphorus. This phosphorus
then diffuses into the steel surface [9].
5.4 Non-removed Functional Layers 131

Phosphorus [wt.%] Phosphorus [wt.%]

1 4 6 8 10 20 25 30 35 0.2 0.4 0.6 0.8
1600 1500
1536 Fe3P Fe2P FeP
1392 1400
1400 1370°C
1392

1262°C
40
1300
1200 24.5 1156°C
Temperature [°C]

Temperature [°C]
1200
1048°C
4.52 19.62
1000 -Fe 0.55 1.23

911 1100
-Fe
800
Tc 745°C
1.65 1000

600 911
900

Tc (Fe3P) 443°C
400 800
Fe 10 20 30 40 50 Fe 0.4 0.8 1.2 1.6
Phosphorus [at.%] Phosphorus [at.%]

Fig. 5.6 Binary phase diagram iron-phosphorus according to [10]

Fig. 5.7 Detection of phosphorus diffusion; (a) quenching and tempering microstructure with
delta ferrite; etching: 3% HNO3; (b) phosphorus diffusion without ferrite seam; etching: Oberhof-
fer
132 5 Manufacturing Influence Area Before Heat Treatment

This can result in phosphorus contents of 0.5 to 0.8% in the surface layer. As can be
seen in the binary phase diagram, phosphorus severely constricts the austenite region,
leading to ferrite layers on the surface, Fig. 5.6. This ferrite can completely dissolve the
phosphorus as a mixed crystal and is often referred to as delta ferrite.
Metallographically, the phosphorus-stabilized ferrite seam can be distinguished from
a decarburization at the sharply defined transition to the quenching and tempering micro-
structure, Fig. 5.7a. Steels with higher carbon content, e.g. the microstructure of the steel
C75S shown in Fig. 5.7b, do not show any ferrite. However, phosphorus diffusion can be
reliably detected by an Oberhoffer etching.

5.5 Passive Layers

The result after nitriding or nitrocarburizing depends both on the chosen process param-
eters and on the condition of the surface. Many users assume that only the heat treatment
process is responsible for the result. However, to achieve a uniform compound- and dif-
fusion layer, a clean, residue-free and metallically bright surface is required [11]. This
applies equally to nitriding and nitrocarburizing processes in gas atmospheres and in
plasma. Passive layers a few μm thick are sufficient to hinder or completely block nitro-
gen uptake. The detection of barrier layers is made difficult by their small thickness and
the usual cleaning methods cannot guarantee safe removal.
In [11] the possible effects of manufacturing steps on the process capability of com-
ponents are listed, Table 5.1.

Table 5.1  Comparison of manufacturing steps and possible effects on the process capability of
components [11]
Manufacturing step Manufacturing step affects Responsibility
Steel production, semi-finished Chemical composition of the Manufacturer
product, heat treatment steel, microstructure/resid-
ual stress
Mechanical processing Residual stress/boundary layer Manufacturer
Heat treatment Microstructure, residual stress, Manufacturer
boundary layer
Cleaning/Preservation Boundary layer Manufacturer
Cleaning before nitriding/ nitro- Boundary layer Heat treater/Manufacturer
carburizing
Nitriding/Nitrocarburizing (pro- Nitriding result Heat treater
cess capability of the plant)
Testing (process capability of the Evaluation of the nitriding result Heat treater
measuring tools and methods)
5.6 Damage Examples—Manufacturing Before Heat Treatment 133

5.6 Damage Examples—Manufacturing Before Heat

Treatment

Example 1—Soft surface of a bolt after induction hardening

Component: Bolt ∅ 15 mm
Material: C45, Material No. 1.0503
Error pattern: Soft surface after induction hardening
Heat treatment: Induction hardening

The first example of manufacturing defects before heat treatment shows that decarbur-
ization can occur not only as a result of hot forming, but also intermediate annealing
in humid atmospheres can lead to significant decarburization. If the decarburization is
not mechanically removed after such annealing, completely inadequate hardness profiles
occur during induction hardening.
Bolts, ∅ 15 mm, made of unalloyed heat-treatable steel C45 were induction hardened.
The target specification for the case hardening depth was SHD = 1.1 +1.5 mm. The sur-

Fig. 5.8 Microstructure of the case hardening from the release sample
134 5 Manufacturing Influence Area Before Heat Treatment

67.2 µm

Fig. 5.9 Decarburized surface layer of an induction hardened bolt

face hardness was specified with 55 + 6 HRC. The bolts were in the normal annealed
state and a flawless hardening microstructure was achieved during an initial sample
inspection, Fig. 5.8. The microstructure images also show the typical, few μm-thick
oxide layer after induction hardening without protective gas.
In a subsequent batch with identical induction hardening settings, the surface hard-
ness test yielded completely inadequate values, which is why an investigation of a bolt
from this delivery was commissioned. The result was very clear, as can be seen from
the microstructure images in Fig. 5.9. On the surface, there was a completely decarbur-
ized surface layer with pure ferrite. The decarburized suface layer had an extent of about
70 μm. Behind the decarburized surface layer, there was a deep decarburization, which
can be seen from the hardness progression curve in Fig. 5.10. Only after about 0.7 mm
was the full hardness achieved.

Example 2—Distortion of spring steel clips after hardening and tempering

Component: Spring steel clips

Material: C75S, Material No. 1.1248
Error pattern: Distortion after hardening and Tempering
Heat treatment: Salt bath hardening and tempering

Wire and sinker EDM is performed in manufacturing technology both in the soft and
hard state. The error frequency proved to be significantly higher in all conducted investi-
gations for EDM treatments in the hard state, which will be explained in Chap. 7.
The next example, however, deals with an unnoticed error of an EDM treatment in
the soft state, which only became apparent during the final hardening. This is a typi-
cal example of a latently hidden error. Only the required temperature-time cycles for
quenching and tempering made the error visible.
5.6 Damage Examples—Manufacturing Before Heat Treatment 135

900

800

700

600
Hardness [HV 1]

500

400

300

200

100

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
Edge distance [mm]

Fig. 5.10 Hardness profile perpendicular to the surface of the induction hardened bolt

From a cold strip with a strip thickness of 0.3 mm of the spring steel C75S, multi-
layer spring steel clips were worked out by wire erosion, Fig. 5.11. The cold strip was in
the +LC state (soft annealed and slightly rerolled). The spring steel clips produced in this

Fig. 5.11 Complained spring steel clips due to strong distortion

136 5 Manufacturing Influence Area Before Heat Treatment

way were then hardened and tempered without further intermediate treatment. Due to the
strong distortion, they were subsequently complained about.
The heat treatment was carried out in a salt bath at a hardening temperature of
820 °C. The components to be heated were previously dried and preheated at a tem-
perature of 300 to 400 °C and then immersed in the salt melt. The preheating is car-
ried out for safety reasons to avoid moisture entry into the salt bath. On the other hand,
this avoids a shock heating of these thin spring steel clips. The spring steel clips were
charged hanging through the hole.
The quenching was carried out in a second salt melt heated to 180 °C. After this, the
spring steel clips were cooled to room temperature. The final step of the heat treatment
was tempering at 400 °C for 1 hour in a circulating air oven.
After hardening and tempering, the spring steel clips were heavily distorted, which
can already be seen in Fig. 5.11. However, it was surprising that the distortion did not
have the typical appearance of a classic hardening distortion. With uneven heating or
cooling, a curvature of the spring over the flat side should have occurred. Instead, there
was a curvature of the spring steel clips over the narrow side. There was also no pre-
ferred orientation. The direction of curvature was completely random.
The components treated in this way were complained about at the executing salt bath
hardening company and sent for examination. The investigations were to clarify whether
microstructural inhomogeneities have led to different residual stresses, which could be
responsible for the shape change.
From the numerous spring steel clips, a particularly heavily bent clip was selected.
The residual stresses were measured on both flat sides with an X-ray diffractometer. The
following results were obtained in the measurement:
Side 1: −16.0 MPa ±15.2; Side 2: −21.6 MPa ±11.4
From these measurement results it could be read that there was no significant difference
in the residual stresses between both sides and practically no significant residual stresses
were present on both sides. The results corresponded to typical values after the previ-
ously described heat treatment.
The distribution of hardness values was also checked, Table 5.2. The hardness val-
ues in Tab. 2 correspond to the original reading values. The mean value was corrected
according to the specification in DIN EN ISO 6507-1—Method M2 [12] with the
measurement deviation b determined in accordance with the standard from long-term

Table 5.2  Results of the Vickers hardness measurements (core)

Measured hardness values [HV 1] Aver- b corr. UCorr Test result
1st 2nd 3rd 4th 5th age [HV Aver- [HV [HV 1]
value value value value value value 1] age 1]
[HV 1] value
[HV 1]
464 466 466 473 468 464 −6,2 470 ± 11,1 11,1 ± 11,1
5.6 Damage Examples—Manufacturing Before Heat Treatment 137

Fig. 5.12 Uniform quenching and tempering microstructure without surface changes of two
examined spring steel clips

e­ quipment monitoring. The standard measurement uncertainty Ucorr was determined and
documented.
There was a very uniform hardness of 470 ±11.1 HV 1. This mean value corre-
sponded to the expected hardness after hardening and tempering at 400 °C.
Finally, a metallographic microstructure evaluation was also carried out on two spring
steel clips, Fig. 5.12. The quenching and tempering microstructure was homogeneous
138 5 Manufacturing Influence Area Before Heat Treatment

across the entire cross-section in both samples examined. The few μm thick oxide layer
was formed during tempering in the air circulation oven.
The formation of the direction of deformation and the flawless and homogeneous
microstructure formation argued against a faulty tempering treatment. Segregation-
related microstructural inhomogeneities were also not present.
Therefore, the most likely cause of the error was assumed to be an uneven distribution
of residual stresses at the cut edges during wire EDM.
The detailed presentation of all investigation steps makes it clear that a direct proof
of the cause of damage was not possible. Surface hardenings during wire EDM with the
formation of high and unevenly formed residual stresses are no longer detectable after a
tempering treatment. During heating, both any existing re-hardening layers are removed
and the residual stresses are released. Untreated reference parts were no longer available.
The cause assessment based on indications is a frequently occurring basis for deci-
sion-making in damage determination. The financial value of the spring steel clips was
low, but the effects were significant, as numerous systems could not be delivered on
schedule.

Example 3—Cracks in Shaft Retaining Rings After Heat Treatment

Component: Shaft Retaining Rings

Material: C75, Material No. 1.0605
Error Image: Cracks After Heat Treatment
Heat Treatment: Hardening and Tempering

The positive changes in properties due to phosphate layers have already been described
in the introduction (5.4). At the same place, however, it was also pointed out that phos-
phate layers must be removed before heat treatment under a protective gas atmosphere.

284.3 µm 248.6 µm

Fig. 5.13 Intergranular cracks up to a depth of approx. 0.30 mm

5.6 Damage Examples—Manufacturing Before Heat Treatment 139

Fig. 5.14 Intergranular cracks (a) after etching in 3% HNO3; (b) after an Oberhoffer etching

If removal is omitted or hardening is accidentally carried out without dephosphating, this

leads to a significant uptake of phosphorus.
In the damage case presented here, shaft retaining rings made of the material C75,
1.0605, were austenitized in a conveyor belt furnace at 840 °C, quenched in oil and sub-
sequently tempered at 320 °C. After hardening and tempering, numerous cracks were
visible on the ring surface. The crack depth was very similar in all cases, with 0.25 to
0.30 mm, Fig. 5.13.
With the carbon content present here, there was no formation of phosphorus-stabi-
lized ferrite, Fig. 5.14a. Only the etching contrast according to Oberhoffer demonstrated
the grain boundary-oriented phosphorus diffusion, Fig. 5.14b. The high phosphorus
content embrittled the grain boundaries and the quenching stresses unavoidable during
quenching led to crack formation.
Energy-dispersive X-ray analysis measured a phosphorus content immediately below
the surface of 0.61 wt.-%.

Example 4—Different Initial Hardness Values in Throttle Discs

Component: Throttle Discs Ø 12 mm × 3.5 mm

Material: 100Cr6, Material No. 1.3505
Error Image: Different Initial Hardness Values
Heat Treatment: Vacuum Hardening

The heat treatment plant received a large quantity of small choke discs Ø 12 mm × 3.5
mm made of the material 100Cr6, 1.3505, for hardening and tempering. Due to the small
dimensions, the heat treatment plant decided on a vacuum hardening 850 °C → 5 bar
nitrogen with subsequent deep freezing. During the inspection of the initial hardness,
two different hardness levels were measured rather randomly, and an investigation was
initiated as a result.
140 5 Manufacturing Influence Area Before Heat Treatment

Fig. 5.15 Martensite without pearlite in the series with the higher hardness

Fig. 5.16 Martensite with small amounts of pearlite in the series with the lower hardness

The hardness check revealed that a subset had an average hardness of 730 HV 10,
while the second subset had an average hardness of 695 HV 10. The variations were
comparably large at ± 9 HV 10 for both hardness levels.
To check the residual austenite content, an X-ray determination was carried out. The
residual austenite content was below the detection limit of 2% for both hardness levels.
Therefore, the choke discs were technically free of residual austenite.
However, differences in the microstructure were found during the metallographic
examination. The choke discs with the higher hardness were completely transformed into
martensite, Fig. 5.15, while a small amount of pearlite was present in the choke discs
with the lower hardness level, Fig. 5.16. Since these were components from the same
hardening batch, which were randomly distributed, a different quenching speed was
ruled out as a cause.
5.6 Damage Examples—Manufacturing Before Heat Treatment 141

Table 5.3  Analysis results and target values according to DIN EN ISO 683-17 [13]
Target C Si Mn P S Cr Ni Mo Cu Al
analysis % % % % % % % % % %
100Cr6, 0.93 0.15 0.25 max. max. 1.35 – max. max. max.
1.3505
acc. to 1.05 0.35 0.45 0.025 0.015 1.60 – 0.10 0.30 0.050
DIN EN
ISO
683-17
Permissi- ± 0.03 ± 0.03 ± 0.04 ± ± ± 0.05 – ± 0.03 ± 0.03 ±
ble limit 0.005 0.005 0.01
deviation
of the
piece
analysis
Throt- 0.88 0.35 0.25 0.007 0.005 1.33 0.09 0.02 0.06 0.035
tle discs
with high
hardness
Throt- 0.94 0.28 0.24 0.009 0.006 1.41 0.07 0.02 0.10 0.034
tle discs
with low
hardness
All information in mass-%

Since initially only one component sample was analyzed, analyses of choke discs
with high and low hardness were now specifically determined using optical emission
spectrometry (OES), Table 5.3.
The result was surprising for the hardening shop, as the analysis results showed two
different melting batches of a delivery lot. The throttle discs with the higher hardness
were below the minimum content in carbon content, even taking into account the permis-
sible limit deviation, which seemed contradictory at first glance.
The carbides in the rolling bearing steel 100Cr6 are not pure iron carbides, but iron-
chromium mixed carbides. The higher carbon content thus leads to a higher amount of
carbide, which also causes a larger amount of bound chromium in the carbides. At the
chosen austenitizing temperature of 850 °C, there was still a larger amount of undis-
solved carbides and thus a larger amount of bound chromium content in the components
with higher carbon content. The matrix therefore had slightly lower chromium contents,
which slightly increased the upper critical cooling rate for martensite formation. How-
ever, the very mild nitrogen gas pressure cooling was sufficient to form a small amount
of pearlite.
142 5 Manufacturing Influence Area Before Heat Treatment

Fig. 5.17 Uneven formation of the compound layer (red layer = copper support foil)

Fig. 5.18 Cracks in the compound layer as a result of incorrect preparation

An austenitizing temperature 10 °C higher would not have made this small difference
in hardenability visible. The same applies to oil quenching, where the higher cooling rate
would have been sufficiently large for complete martensite formation even with lower
hardenability.
For all steels, permissible ranges are set for each alloying element. In addition to the
ranges of the melt analyses, the material standards also specify permissible limit devia-
tions for piece analyses. These limit deviations are different for both the respective alloy-
ing elements and the absolute content. Narrowed ranges are possible, but usually only
achievable when accepting large delivery quantities. Differences in hardenability are
possible within the permissible ranges.
Therefore, a batch-pure delivery is generally required, or different melting batches of
the same components must be kept separate. The manufacturing error was therefore in
the mixing of two melting batches.
5.6 Damage Examples—Manufacturing Before Heat Treatment 143

Table 5.4  Analysis results and evaluation of the material DC04, 1.0338, according to DIN EN
10139:2020-06 [16]
Target analysis C Si Mn P S Cr Ni Mo Al
DC04, 1.0338 % % % % % % % % %
acc. to DIN EN max. – max. max. max. – – – –
10139 :2020-06
0.08 – 0.40 0.030 0.030 – – – –
Measurement results 0.06 0.02 0.21 0.007 0.012 0.03 0.04 0.003 0.036
All data in mass-%

Example 5—Uneven compound layer on a housing

Component: Housing with 0.8 mm sheet thickness

Material: DC04, Material No. 1.0338
Error pattern: Uneven compound layer
Heat treatment: Gas nitrocarburizing

When nitrocarburizing housings made of cold-formed sheets of the material DC04,

1.0338, there were strong fluctuations in compound layer thicknesses after a supplier
change, Fig. 5.17. In some places, the target specification of 6 to 14 μm was achieved, in
many places the compound layer was thinner or not present at all.
When examining nitrocarburized components metallographically, it is recommended
to stick a support foil, e.g. copper, onto the component surface [14]. The compound layer
has a high hardness and can therefore form cracks or completely chip off if prepared
improperly [15], Fig. 5.18.
The material DC04 is an unalloyed steel, which is why the formation of a pas-
sive layer from the material composition is not to be expected, Table 5.4. Rather, it is
assumed that passive layers were unintentionally applied to the surface during manufac-
turing, which could not be removed during cleaning before nitrocarburizing.
In the course of the conducted investigations, no residues of passive layer-forming
substances could be detected in the EDX analysis. However, since there were highly var-
ying compound layer thicknesses at very close distances, passive layers were evaluated
as the most likely cause.

Example 6—Insufficient Surface Hardness of a Tool

Component: Tool
Material: 31CrMoV9, Material No. 1.8519
Error pattern: Insufficient surface hardness
Heat treatment: Plasma nitriding
144 5 Manufacturing Influence Area Before Heat Treatment

Fig. 5.19 Plasma nitrided tool with noticeable color differences on the nitrided surface

8.44 µm

900 900

800 800

700 700
Hardness [HV 0.5]

Hardness [HV 0.5]

600 600

500 500

400 400

300 300

200 200

100 100

0 0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Edge distance [mm] Edge distance [mm]

Fig. 5.20 Microstructure documentation and hardness progression measurement in the area of
high hardness and low hardness

The material 31CrMoV9, 1.8519, was standardized as nitriding steel in DIN EN

10085:2001-07 [17] at the time of the nitriding treatment described here. This standard
has since been replaced by the international DIN EN ISO 683-5:2021-08 [18].
A tool made of this nitriding steel was in the quenched and tempered state and was
delivered for plasma nitriding. After completion of the nitriding treatment, the tool
5.6 Damage Examples—Manufacturing Before Heat Treatment 145

Table 5.5  Results of the Element contents [wt.-%]

EDX measurements with two Element 20 kV 10 kV
different acceleration voltages
Nitrogen 1.35 < Detection limit
Oxygen 2.22 18.68
Silicon 0.71 3.27
Chromium 2.72 3.17
Manganese 0.76 < Detection limit
Iron Remainder Remainder

Fig. 5.21 Pin (left) and brittle fracture surface (right)

became noticeable due to a different discoloration on the surface, Fig. 5.19. Therefore,
a small load hardness test was carried out and hardness values between 350 and 860 HV
1 were determined. These very large hardness differences were the reason for a further
investigation.
The metallographic examination clearly revealed the cause of the large hardness dif-
ferences. In the area of high hardness, an intact nitriding layer with an approx. 8 μm
thick compound layer was present, Fig. 5.20. At the position with the low hardness, nei-
ther a diffusion layer nor a compound layer was present. Hardness progression measure-
ments to determine the nitriding hardness depth confirmed the metallographic findings.
In the area of the intact nitriding layer, a near-surface hardness of 860 HV 0.5 and a
nitriding hardness depth NHD = 0.40 mm were determined. Below the macroscopically
noticeable surface, no increase in hardness was detected.
The investigations were continued in the scanning electron microscope (SEM) using
energy-dispersive X-ray analysis (EDX). In the area of low hardness, spectra were cre-
ated with two different acceleration voltages. This allows different depth ranges to be
captured. At the acceleration voltage of 20 kV, the electrons penetrate approx. 1.5 μm
into the surface, at 10 kV approx. 0.5 μm.
146 5 Manufacturing Influence Area Before Heat Treatment

The absolute values of the element specifications are not of primary importance.
Rather, the comparison of both measurements provides a hint at possible passive layers,
Table 5.5. The results at the acceleration voltage of 10 kV show a strong increase in the
elements oxygen and silicon compared to the measurement at 20 kV. This suggests a sili-
con oxide layer on the surface.

Example 7–Brittle fracture of a pin

Component: Pin ∅ 26 mm
Material: 34CrNiMo6, Material No. 1.6582
Error image: Brittle fracture
Heat treatment: Salt bath nitrocarburizing

A pin ∅ 26 mm with a simple geometry had broken brittle due to bending stress,
Fig. 5.21. The supplied drawing specified the material 34CrNiMo6 +QT, 1.6582, for
the pin and a salt bath nitrocarburizing with a compound layer of 10 + 10 μm was pre-
scribed. The hardness in the base material was not specified.
As part of a damage case investigation, the chemical composition is preferably
checked immediately after the incoming documentation. This was also done in this case
and the results are listed in Table 5.6. It was very quickly recognizable that the intended
heat-treatable steel 34CrNiMo6, 1.6582, was not used. The pin was made from the case-

Table 5.6  Results of the emission spectroscopic examination and comparison with two standard
materials
Target analysis C Si Mn P S Cr Ni Mo Al
34CrNiMo6, % % % % % % % % %
1.6582 0.30 max. 0.50 max. max. 1.30 1.30 0.15 –
acc. to DIN EN
10083-3 0.38 0.40 0.80 0.025 0.035 1.70 1.70 0.30 –

20MnCr5, 0.17 max. 1.10 max. max. 1.00 – – –

1.7147 0.22 0.40 1.40 0.025 0.35 1.30 – – –
acc. to DIN EN
10084
Measurement 0.20 0.22 1.20 0.007 0.007 1.04 0.12 0.02 0.017
results
Trace elements Ti Cu W Co V Nb Sn B N
% % % % % % % % %
Measurement 0.002 0.22 <Nwg. 0.009 0.003 <Nwg. 0.013 <Nwg. 0.009
results
< Nwg.: below detection limit. All data in mass-%
5.6 Damage Examples—Manufacturing Before Heat Treatment 147

Fig. 5.22 Case hardening layer instead of a compound layer; (a) case hardening layer with Cu
support foil; (b) non-martensitic surface layer

900

800
0.83 mm
700

600
Hardness [HV 1]

500

400

300

200

100

0
0.0 0.5 1.0 1.5
Edge distance [mm]

Fig. 5.23 Determination of the case hardening depth

hardening steel 20MnCr5, 1.7147. The standards valid at the time of the damage were
used for comparison of the analyses.
In the next examination section, a metallographic examination was carried out.
Expecting a thin compound layer, the bolt surface was supported with a copper foil,
Fig. 5.22. The result was surprising, there was no compound layer, but a case hardening
layer. On the component surface, the typical non-martensitic surface layer after carbur-
izing in an endothermic furnace atmosphere was present.
148 5 Manufacturing Influence Area Before Heat Treatment

The non-martensitic surface layer is a result of the unavoidable edge oxidation. The
atomic oxygen released during CO splitting is partially absorbed by the surface and
forms oxides with the oxygen-affine elements chromium, manganese, and silicon. As a
result, the oxide-affected surface layer is depleted of hardenability-increasing alloying
elements, and an oil cooling is no longer sufficiently fast to form martensite.
The case hardening depth was 0.83 mm, Fig. 5.23. The maximum hardness measured
in the hardness profile was 848 HV 1, and a surface hardness of 65 HRC was also meas-
ured using the HRC test.
From the examination results, it was therefore evident that not only a different mate-
rial was used, but also an unintended case hardening was carried out. Therefore, it was
not a pure material mix-up, but a different, non-drawing-compliant solution was deliber-
ately chosen.
The very high surface hardness also proved that either no tempering was carried out
at all, or it was insufficient. Overall, a highly brittle condition was present, which had a
high probability of fracture under bending loads.

Fig. 5.24 Cracking in the inductively hardened area

5.6 Damage Examples—Manufacturing Before Heat Treatment 149

Fig. 5.25 Fineacicular martensite in the inductively hardened surface layer

Fig. 5.26 Surface defects on the bore surface in the hardened area

Fig. 5.27 Friction martensite and cracking on the bore surface below the hardened area
150 5 Manufacturing Influence Area Before Heat Treatment

Example 8—Cracking on Adjustment Screws

Component: Adjustment Screws

Material: C53GG, Material No. 1.1213
Error Image: Cracking
Heat Treatment: Inductive Hardening

To increase wear resistance, adjustment screws made of material C53G, 1.1213, were
inductively hardened in the head area. In this process, cracking occurred in several parts
in the inductively hardened area, Fig. 5.24. Since the cracking was present after the
inductive hardening, it was clear to the client that it was a hardening stress crack and the
parts were claimed. The hardening shop checked all setting data and could not find any
indications of faulty inductive hardening. Therefore, an independent investigation was
commissioned.
The determination of the chemical composition confirmed the unalloyed heat-treata-
ble steel C53G without abnormalities.
The martensitic hardening microstructure was very fineacicular, indicating a properly
low hardening temperature and holding time, Fig. 5.25. The bore surface showed numer-
ous unstructured cracks and surface defects, Fig. 5.26. Similar cracks were also pre-
sent on the unhardened bore surface, Fig. 5.27. Here it was also recognizable what had
caused the cracks. During deep hole drilling, the surface temperatures were so high that
the ferrite-pearlite microstructure of the initial state was austenitized and transformed to
martensite. The tensile residual stresses formed in the friction martensite layer led to
cracking. The friction martensite was naturally no longer present in the hardening layer.
A bore surface prone to cracking due to friction martensite formation creates a high
probability of cracking even with proper inductive hardening. Although there was a hard-
ening stress crack, the cause was in the faulty borehole execution.

References

1. DIN EN 10021:2007-03: Allgemeine technische Lieferbedingungen für Stahlerzeugnisse

2. ISO 404:2013, Steel and steel products – General technical delivery requirements
3. DIN EN 10222-1:2002-07: Schmiedestücke aus Stahl für Druckbehälter – Teil 1: Allgemeine
Anforderungen an Freiformschmiedestücke
4. DIN EN 10243-1:2000-06: Gesenkschmiedeteile aus Stahl – Maßtoleranzen Teil 1: Warm-
hergestellt in Hämmern und Senkrecht-Pressen
5. Scholtes, B.; Ellermann, A.; Baunack, D.; Kraus, A.; Zinn, W.: Werkstofftechnische Grundla-
gen des Biegerichtens. HTM J. Heat Treatm. Mat. 70, S. 4–18
6. Schulze, V.; Osterried, J.; Strauß, T.; Zanger, F. Analysis of surface layer characteristics for
sequential cutting operations. HTM J. Heat Treatm. Mat. 67 (2012) 6, S. 347–356
7. Bernhard Syren, Helmut Wohlfahrt, Eckard Macherauch: Zur Entstehung von Bearbeitungsei-
genspannungen. Archiv für das Eisenhüttenwesen (1977) 48. Jahrgang, S. 421–426
References 151

8. Fritz, A. F.: Fertigungstechnik. Springer Vieweg-Verlag 2018, Berlin

9. Rausch W.: Die Phosphatierung von Metallen; Leuze-Verlag Saulgau 1974
10. Kubaschewski, O.; „Iron – Binary Phase Diagrams“; Springer-Verlag, 1. Auflage; 1982
11. Huchel, U.; Klümper-Westkamp, H.; Sommer, P.; Spengler, A.: Notwendige Anforderungen
an Bauteile zur Sicherung der Prozessfähigkeit beim Nitrieren und Nitrocarburieren. HTM J.
Heat Treatm. Mat. 64; (2009) 4. S. 222–225
12. DIN EN ISO 6507-1:2018-07: Metallische Werkstoffe – Härteprüfung nach Vickers – Teil 1:
Prüfverfahren
13. DIN EN ISO 683-17:2015-02: Für eine Wärmebehandlung bestimmte Stähle, legierte Stähle
und Automatenstähle – Teil 17: Wälzlagerstähle
14. Riesbeck, S.; Baudis, U.: Universelle Präparationsmethode für randscharfe Schliffproben zur
Ermittlung der Verbindungsschichten an nitrierten und nitrocarburierten Werkstücken. Der
Wärmebehandlungsmarkt (2007) 2, S. 5–8
15. Sommer, P.: Failure Analysis of Nitrided Components. HTM J. Heat Treatm. Mat. 73 (2018)
5, pp. 258–267
16. DIN EN 10139:2020-06: Kaltband ohne Überzug aus weichen Stählen zum Kaltumformen –
Technische Lieferbedingungen
17. DIN EN 10085:2001-07: Nitrierstähle – Technische Lieferbedingungen
18. DIN EN ISO 683-5:2021-08: Für eine Wärmebehandlung bestimmte Stähle, legierte Stähle
und Automatenstähle – Teil 5: Nitrierstähle
Heat Treatment Influence Area
6

Contents

6.1 Process Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

6.2 Thermochemical Heat Treatments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250

One of the most distinctive characteristics of steels is their heat treatability. A soft struc-
tural state can be produced for mechanical processing. Under operating conditions, wear
resistance and power transmission are essential properties that can be optimized through
targeted heat treatment. The unalloyed case-hardening steel C15E, 1.1171, has a hard-
ness of 140 HV in the annealed state and can reach hardness values of 850 HV on the
surface through carbonitriding and case hardening. This corresponds to a factor of 1:6.
The hot-work steel X37CrMoV5-1, 1.2343, is pre-hardened or annealed before mechani-
cal processing and the hardness is then 220 HV. After nitriding treatment, the hardness is
up to 1100 HV, which means a max. hardness increase by a factor of 5.
The heat treatability of steels has led to an enormous variety of possible heat treatments.
In Fig. 6.1, the heat treatment processes are divided into three main groups and subgroups.
Numerous heat treatment processes can be found in each individual group. Detailed
descriptions of the individual heat treatments can be found in the literature, e.g. [1–5].
In the treatments from the hot forming heat, no separate heating for the heat treat-
ment is necessary. The components are first hot formed and then often press hardened.

© The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, 153
part of Springer Nature 2024
P. Sommer, Mistakes Before, During and After Heat Treatment of Steel,
https://doi.org/10.1007/978-3-658-44160-9_6
154 6 Heat Treatment Influence Area

Heat treatment processes for steels

Treatments from the Thermal Thermochemical
hot forming heat processes processes

Precipitation hardening Annealing treatments Case hardening

Hardening, press Nitriding

Surface hardening
hardening Nitrocarburizing
Hardening and Boriding
tempering with
thorough heating

Fig. 6.1 Grouping of different heat treatment processes

­ owever, precipitation hardening requires precipitation hardening ferrite-pearlite steels

H
(AFP steels). A controlled cooling from the hot forming heat leads to a precipitation of
finest carbides or carbonitrides, which lead to a hardness increase.
In the thermal processes, no change in the surface layer composition is intended.
Decarburizations, on the other hand, are particularly accepted in the case of pre-material
heat treatments.
On the other hand, in the thermochemical processes, the change in the surface layer
composition is specifically aimed for. The elements carbon, nitrogen, and boron are the
most significant elements in this heat treatment group. These elements diffuse into the
surface and increase the hardness and wear resistance.
Each individual heat treatment process can also be carried out in various heat treat-
ment furnaceswith specific process parameters, Fig. 6.2. Therefore, the information on
the respective plant and process parameters is also important for the heat treatment errors
to be described.
A detailed description of the individual heat treatment processes will not be discussed
here, as this has already been done in numerous publications. Instead, the potential
sources of error in the heat treatment processes will be described. If there has been a
faulty treatment, the following scenarios are possible:

1. The error in the process management is detected directly during or after the heat treat-
ment, thus preventing the accidental delivery of faulty components.
2. If the error has occurred in only a few components within a large total quantity, this
could go unnoticed and faulty components would be delivered and possibly installed.
6.1 Process Parameters 155

Heat treatment furnaces

Heating Charging Oven room Atmosphere Medium
Electric Batch oven Chamber furnace without shielding gas Gas atmosphere
Gas Continuous oven Bell type furnace with shielding gas Vacuum
Oil Rotary hearth oven Shaft furnace Vacuum Molten salt
Direct heating Bogie hearth furnace Plasma
Manual charging
Indirect heating Tub furnace Lead
Semi-automatic
Fluidized bed
charging
Fully automatic
charging

Fig. 6.2 Variety of heat treatment furnaces

3. A systematic check according to a specified test plan was not carried out, leading to a
high risk of overlooking faulty components.
4. The testing of reference parts, e.g., test coupons in case hardening, are not representa-
tive of the treated components.
5. A check of the required properties, e.g., notch toughness, is not possible or is not car-
ried out.
6. Missing target specifications or heat treatments that lead to changes in material-spe-
cific properties, e.g., corrosion resistance, are not clarified before the heat treatment is
carried out.
7. Geometric risks, e.g., abrupt cross-sectional transitions, damages or corrosion depos-
its are not recognized or clarified before the heat treatment is carried out.

The error examples still to be described make it clear that all the aforementioned variants
have occurred in practice and have been submitted for investigation. The sources of error
described in the following sections refer to the process variables of heat treatment. They
are neither material nor component specific. The list of error patterns does not claim to
be complete, but they are typical errors that were selected from over 10,000 damage case
investigations in our laboratory.
In Fig. 6.3, typical causes of error are summarized in four groups.

6.1 Process Parameters

For all heat treatments, the process parameters temperature, time, and atmosphere apply.
The temperature is not only to be evaluated as an absolute value, but also its temporal
change or heating and cooling speed. Influencing these process parameters are
156 6 Heat Treatment Influence Area

Heat treatment errors

High thermal False Temperature
Lack of care
stresses atmospheres error
Heating for Unwanted Temperatures Charging
austenitizing carburization too high too dense
Quenching stresses Unwanted Temperatures Local carburizing
due to abrupt media embroidery too low protection
Heating for Other surface Incomplete Other heat
tempering reactions heating treatment defects

Fig. 6.3 Overview of typical heat treatment errors

• the geometry of the components to be treated,

• the number of components to be treated in a batch,
• the type of batch loading,
• the chosen or available furnace system,
• the chosen duration after temperature equalization as well as the
• dynamics of the atmosphere control.

6.1.1 The Heating Phase

The temperature range of possible heat treatment processes ranges from −196 °C for
deep freezing to 1250 °C for austenitizing high-speed steels. A distinction must also be
made between the respective target temperature and the temperature gradient until the
target temperature is reached. This section will initially only discuss heating to the target
temperature.
Small-volume components are often treated as bulk material in continuous furnaces
or in batches in baskets. In conveyor belt furnaces, the components can be automatically
applied via a vibration chute and precisely dosed using a weighing cell. However, man-
ual application is also possible. The treatment throughput and the belt speed are coordi-
nated with each other. The mentioned process variables treatment quantity, belt speed,
temperature, and furnace atmosphere are specified and documented in modern systems
via a process control computer. This already excludes important sources of error, assum-
ing that all process variables have been correctly set for the treatment batch.
The uniformity of the heap height cannot be fully automated for verification. If heap
formations occur, there is a risk that the components lying in the center of the heap will
not be sufficiently heated to the hardness or tempering temperature (Fig. 6.4).
Bulk goods can of course also be heat treated in batch-operated furnace systems. The
charging baskets are usually filled manually and the operator must ensure an even distri-
bution and compliance with the max. permissible heap height per basket. Furthermore, it
6.1 Process Parameters 157

Generally too dense fill

Incomplete
Belt speed too high
austenitization

Isolated heap formations

Fig. 6.4 Possible causes for incomplete austenitization in conveyor belt furnaces

must be considered whether and, if necessary, how many charging baskets can be stacked
on top of each other without causing too large temperature differences and resulting
unacceptable hardness variations.
The error pattern of incomplete austenite formation is not only found in bulk goods,
but can occur in practically all heat treatments. Eckstein has already adopted the correc-
tion factors according to Ordinanz [6] for different furnace loads and published them in
[7], Fig. 5.5.
The graphical representation in Fig. 6.5 clearly shows that differences in the heating
speed of up to 400% can occur depending on the shape and charging density. This con-
sideration is relevant for all batch-operated furnace systems.
Insufficient heating can also occur during tempering. This then leads to larger hard-
ness variations and possibly to losses in toughness.
Inductive heating is always a single hardening process, which is why the two previ-
ously described considerations do not apply [8]. Nevertheless, even with inductive heat-
ing of the entire cross-section or only the surface layer, insufficient austenite formation
is possible. Especially with short-term surface heating, a precise coordination of the pro-
cess parameters with the respective initial state is important. Pre-hardened components
have a uniform carbide distribution with small carbide size. After the formation of aus-
tenite, short diffusion paths are then available, which favor a uniform distribution of car-
bon in the austenite.
Ferrite-pearlite microstructures, especially after forging processes with high final
forging temperature, on the other hand, have a wide-mesh ferrite network. With such a
microstructure, the diffusion paths are considerably larger, which makes the formation of
a homogeneous austenite difficult, sometimes even impossible.
When heating up components, temperature differences within a component are hardly
avoidable. The very slow heating during careful stress relief annealing is excluded here.
Temperature differences within a component always lead to thermal stresses. Simpli-
fied, it can be estimated that a temperature difference of 100 °C causes a stress of 250
MPa. For reasons of equilibrium, this amount is divided into equal amounts of tensile
and compressive stresses.
If the temperature differences during heating are so large that the resulting thermal
stresses exceed the yield point, plastic deformation and stress relief occur. This perma-
158 6 Heat Treatment Influence Area

Arrangement in Factor Arrangement in Factor

the oven the oven

. .

. .

. .
. .

. .

. .

. .

Fig. 6.5 Correction factors for the heating duration with different furnace loading according to
Ordinanz [6]

nent deformation is perceived as distortion after heat treatment. If plastic deformation is

not possible due to high hardness, cracking can also occur if the stress amounts are suf-
ficient. For large-volume components and large cross-sectional differences, a multi-stage
heating with holding phases for temperature compensation is recommended.
When heating up to the hardening temperature, a soft, mechanically workable mate-
rial condition is present in most cases. This is also reflected in the number of investigated
heating stress cracks. Although these have been detected a few times, in these cases there
were surface hardenings due to thermal machining processes such as flame or laser cut-
ting, or very unfavorable geometries or design features. The crack progression in the soft
material condition is usually transcrystalline.
The risk of a heating stress crack when heating up for tempering is much greater,
as a high-strength and brittle material condition is present after quenching. Unfavora-
ble geometries, e.g. too tightly dimensioned radii, large mass differences, notches etc.,
increase the risk of crack formation.
6.1 Process Parameters 159

A careful evaluation of the crack flanks is necessary for identifying the time of crack
formation. If a heating stress crack occurs during heating to the hardening temperature,
the crack flanks must also show decarburization in a decarburizing atmosphere. If heated
in an endothermic atmosphere, the crack flanks will show comparable intergranular oxi-
dation as the surfaces. Only in vacuum hardening would no changes be detectable on the
crack flanks.
A heating stress crack during heating for tempering typically shows an intergranu-
lar crack progression. If the tempering treatment is carried out under air, oxides in the
crack gap are to be expected. Even with tempering treatments under an inert atmosphere,
e.g., nitrogen, oxide formation in the crack may be present if, for example, the batch
was cooled in air. On the other hand, no significant decarburizations are to be observed
at usual tempering temperatures of up to 650 °C. The intergranular crack progression
and the subsequent reactions in the crack gap are also given in a quenching stress crack,
which makes a distinction between the two points of origin difficult.
The evaluation of a documented temperature-time course can provide valuable infor-
mation. Numerous crack formations were observed during heating for tempering when
the required tempering temperature was already present in the furnace and the compo-
nent or batch was introduced cold into the furnace. There were also cases where large
components with a single weight of 840 kg were charged into an initially unheated fur-
nace. The direct gas heating then realized a heating to 490 °C within 20 minutes. This
process management is equivalent to charging a fully preheated furnace.

6.1.1.1 Damage Examples—Heating

Example 1—Missing Spring Action in Safety Clips

Component: Cold formed safety clips

Material: C45E, Material No. 1.1231
Error Image: A missing spring action was detected during assembly
Heat Treatment: Bainite hardening in salt bath

Cold formed safety clips with a sheet thickness of 0.5 mm made of material C45E,
1.1231, were cold formed in the GKZ-annealed state and then transported to the harden-
ing shop. Here, the components were austenitized in a conveyor belt furnace and then
isothermally transformed in the bainite stage in a salt bath. A hardness of 450 +70 HV 1
was required. After the hardness was measured on some safety clips and the target hard-
ness was reached, the components were released and sent back. The manufacturer of the
safety clips supplied his customers without further testing.
A missing spring action was detected during later assembly. The hardness was meas-
ured on this spring at the end customer, which was significantly below the target specifi-
cation and the attested hardness values in the supplier certificate of the hardening shop.
160 6 Heat Treatment Influence Area

Fig. 6.6 Microstructure images of safety clips from a treatment batch (a) Decarburization-free
bainite microstructure; (b) Recrystallized microstructure with different grain size; (c) Microstruc-
ture completely transformed in the lower bainite stage; (d) Large proportions of untransformed
GKZ annealing microstructure

The entire batch was blocked and a metallographic examination and a hardness test were
carried out.
Most safety clips showed a flawless bainite microstructure with a hardness at the
upper tolerance limit, Fig. 6.6a, c.
The soft safety clips revealed a recrystallized microstructure with varying grain size
at low magnification, Fig. 6.6b, d. This rather random state of recrystallization was the
result of different degrees of cold forming. At the outer bending line and the inner radius,
there was cold deformation in the critical forming range, which led to a coarse-grained
recrystallization. The central area, on the other hand, was much finer, indicating a degree
of deformation below the critical range. After exceeding the Ac3 transformation temper-
ature, new grain formation would have occurred and the differences would have been
eliminated, which was also evident in the bainite microstructure.
At higher magnification, significant proportions of GKZ annealing microstructure
were still visible on the faulty safety clips. Only at the grain boundaries was there a
changed microstructure, indicating the beginning of austenite formation. During the sub-
sequent cooling in the salt bath, fine-striped pearlite formation occurred at the former
austenite grain boundaries.
6.1 Process Parameters 161

A predominantly still existing GKZ annealing microstructure was a clear error pat-
tern of incomplete austenitization. The shadowing effect of a heap, even with very thin-
walled components, should therefore not be underestimated.

Example 2—Plastic Deformation on Support Pins

Component: Support pinsØ 10 mm × 34 mm

Material: 100Cr6, Material No. 1.3505
Error pattern: Plastic deformation due to insufficient hardness
Heat treatment: Hardening and tempering with oil quenching

Support pins made of the rolling bearing steel 100Cr6, 1.3505, were manufactured from
bar material with a dimension of Ø 10 mm × 34 mm. A protective gas-operated batch
furnace with oil quenching for hardening and tempering was available in the company’s
own hardening shop.
Due to a significantly increased delivery quantity requirement from the end customer,
the throughput was increased and the number of wire mesh charging baskets stacked on
top of each other was increased from 5 to 7, without adequately adjusting the treatment
duration. The bulk density in each individual basket was not changed, but the total mass
was increased by 40%. This should have led to a reassessment of the process parameters
and the treatment results. The removal of a single sample from an unspecified position
was not suitable for assessing or verifying the changed process.
As part of a first batch, components must be taken from different positions and sam-
pled. During the result check, the “best” and “worst” position must then be documented.
For subsequent batches, at least these two positions and possibly another random, but
documented position must be checked.
Since the randomly checked position had reached the required hardness, the entire
batch was released. The support pins were then installed in a combustion engine and a
plastic deformation occurred during the functional test. In the meantime, numerous sup-
port pins had already been installed, the hardness of which was unknown. The deformed
pin had a hardness of 205 HV, with a target specification of 700 +100 HV 10. All sup-
port pins already installed from the relevant delivery batch had to be removed again. The
microstructures shown in Fig. 6.7a-d show the enormous range of the austenitization
state and the resulting hardness scatter in the faulty batch, Table 6.1.

Example 3—Very Soft Nuts

Component: Nuts M10according to DIN EN ISO 898-2 of strength class 10

Material: 20MnB4, Material No. 1.5525
Error pattern: During assembly, very soft nutswere detected
Heat treatment: Tempering with oil quenching in the conveyor belt furnace
162 6 Heat Treatment Influence Area

Fig. 6.7 Microstructures from the faulty treatment batch (a) GKZ annealing microstructure in the
delivery state; (b) High number of undissolved carbides and low pearlite formation; (c) Undis-
solved carbides and martensite formation; (d) Perfect hardening microstructure from the same
treatment batch

Table 6.1  Results of the hardness measurement on five different support pins

Examined sup- measured hardness values [HV 10] Average value
port pin 1st value 2nd value 3rd value 4th value 5th value [HV 10]
Part 1 680 675 681 670 676 676
Part 2 292 265 277 280 284 280
Part 3 334 330 319 322 341 329
Part 4 532 540 537 534 529 534
Part 5 213 211 209 214 217 213

When quenching and tempering bulk goods, the even distribution on the conveyor belt is
extremely important to guarantee uniform and complete heating. If unnoticed pile forma-
tion occurs, this can lead to individual components not reaching the austenitization tempera-
ture. Even with very careful inspection according to a set inspection plan, finding individual
screws with a faulty quenched and tempered microstructure is almost impossible.
6.1 Process Parameters 163

Table 6.2  Results of the hardness measurement on three different nuts

Examined nut measured hardness values [HV 10] Average value
1st value 2nd value 3rd value 4th value 5th value [HV 10]
Part 1 156 155 154 155 158 156
Part 2 157 157 160 161 157 158
Part 3 155 157 154 155 156 155

When components are removed from the quenching bath or after tempering, there is
no longer any reference to pile formation. Only the use of the components then repre-
sents a 100% inspection, which can naturally lead to very large problems.
In the following example, M10 nuts according to DIN EN ISO 898-2 [9] made of
20MnB5 material were produced by cold forming and quenched and tempered in a conveyor
belt furnace system. A GKZ-annealed condition is required to enable flawless cold forming.
The heat-treatable steel used is standardized in DIN EN ISO 683-2:2018-09 [10]. The
hardening temperature is given in this standard as 880–920 °C. This relatively high hard-
ening temperature is necessary because the Ac3temperature is at 865 °C due to the low
carbon content. The Ac1 temperature is at 760 °C.
Only when installing this treatment batch were soft nuts discovered, Table 6.2. The
subsequent metallographic examination and hardness test showed that the soft nuts
had only slightly exceeded the Ac1 temperature. The remnants of carbides of the GKZ
annealing microstructure confirmed this. Consequently, the proportion of undissolved
ferrite was very large, Fig. 6.8a-d. Accordingly, the hardness was only slightly above that
of the GKZ annealing microstructure.
Such a mistake is particularly unpleasant, as neither the temperature-time documenta-
tion nor the regular component inspection after quenching and after tempering guarantee
detection.

Example 4—Insufficient Hardness on Ball Discs

Component: Ball discsForm C21 according to DIN 6319, ∅ 36 mm

Material: 19MnB4, Material No. 1.5523
Error pattern: Hardness after quenching and tempering not achieved
Heat treatment: Tempering with oil quenching

Ball discs—Form C21 with an outer diameter of 36 mm were manufactured from hot-
rolled bars by a machining process. The bars made of the material 19MnB4, 1.5523,
were in the untreated state +U. At the time of production, this heat-treatable steel was
standardized in DIN EN 10269:2006-07 [11] with the following target analysis. This
standard has since been withdrawn and in DIN EN 10269:2014-02 [12] the material
19MnB4 was removed and the material 20MnB4 was added.
164 6 Heat Treatment Influence Area

Fig. 6.8 Microstructure images of the faulty treatment batch (a) Galvanized surface; (b) Clear
line structure; (c) High proportion of undissolved ferrite+carbide; (d) GKZ annealing microstruc-
ture still clearly visible

In addition to the analysis specification from the standard, the titanium content is
indicated, Table 6.3. In the withdrawn standard, but also in current editions, the titanium
content is not specified for boron-alloyed steels. The small addition of titanium binds the
nitrogen in the steel to titanium carbonitrides, thus preventing the formation of boron
nitride. Only the boron dissolved in the lattice leads to the desired increase in hardenabil-
ity, which is why small additions of titanium are necessary.

Table 6.3  Analysis result of the ball disc and comparison with the target analysis
Target analysis C Si Mn P S Al B Ti
19MnB4, 1.5523 % % % % % % % %
acc. to DIN EN 0.17 max. 0.80 max. max. min. 0.0008 –
10269:2006
0.24 0.40 1.15 0.030 0.035 0.020 0.0050 –
Measurement results 0.21 0.05 0.99 0.011 0.010 0.03 0.0033 0.025
All data in mass-%
6.1 Process Parameters 165

Fig. 6.9 Microstructural images after the quenching and tempering of ball discs (a) cross-section
ball disc; (b) lines with undissolved ferrite; (c) homogeneous quenched and tempered microstruc-
ture without ferrite; (d) high ferrite content and tempered martensite

The initial microstructure was a lamellar ferrite-pearlite microstructure, which is typi-

cal for hot-rolled bars of this material. The austenitization was carried out in an internal
quench furnace under neutrally adjusted endothermic protective gas. The ball discs were
charged in several stacked wire mesh baskets. There were no overlaps within a level.
Since the specified tensile strength of 800 to 950 MPa could not be measured on the
ball discs, a conversion according to DIN EN ISO 18265 [13] into a hardness specifica-
tion of 256 to 303 HV 10 was agreed upon.
After the completion of the entire compensation treatment, a hardness test on ball
discs from each batch basket revealed insufficient hardness of 212 to 236 HV 10. This
was followed by a metallographic examination, Fig. 6.9a-d. The ferrite lines of the for-
mer ferrite-pearlite microstructure were not completely dissolved in significant quanti-
ties. In contrast, the former pearlite lines had a flawless compensation structure.
Since all treatment layers showed a comparable result, a too low temperature was
assumed as the cause of the error. The required austenitizing temperature of 880 °C was
either not reached or not specified. The low-carbon ferrite lines did not reach the AC3
temperature.
166 6 Heat Treatment Influence Area

Example 5—Polygonal Ferrite in the Core of a Pinion Shaft

Component: Pinion shaft

Material: 20MoCrS4, Material No. 1.7323
Error pattern: Polygonal ferrite in the core after case hardening
Heat treatment: Case hardening and tempering (Direct hardening)

After carburizing, direct hardening, single hardening, hardening after isothermal trans-
formation or double hardening can be carried out. The procedures are described in detail
in DIN 17022-3:1999-04 [14].
Pinion shafts made of the case-hardening steel 20MoCrS4, 1.7323, should be directly
hardened after carburizing in an internal quench furnace. Before quenching in oil, the
temperature is often reduced to 820 to 840 °C, which reduces the temperature difference
between the components and the quenching medium. This has a positive effect on distor-
tion.
During the cooling phase to the quenching temperature of 820 °C, there was a burner
failure and the furnace temperature dropped to about 700 °C. The burner could be made
functional again and the batch was heated to 820 °C, held for a sufficient length of time,
and then quenched in oil.
In the subsequent metallographic examination, significant amounts of polygonal fer-
rite were found in the core area of the pinion shafts, while the hardening microstructure
in the case hardening layer was flawless, Fig. 6.10a-d.
The review of the temperature-time documentation confirmed the temperature profile
and a sufficiently long holding time after reheating to 820 °C. However, a look at the
TTT diagrams of this steel made it clear why the ferrite was not completely dissolved
during reheating.
The cooling from 920 °C after carburizing to 820 °C does not lead to the formation of
polygonal ferrite, even with a long holding time. This was evident from the isothermal
TTT diagram of this steel. On the other hand, polygonal ferrite was formed during the
unintentional cooling to 700 °C due to the burner defect. When reheating to the origi-
nally intended hardening temperature of 820 °C, the Ac3 temperature, which is at 855 °C,
was not reached.
Therefore, it makes a difference whether the hardening temperature of 820 °C is
reached from the higher carburizing temperature or whether the components are heated
from 700 °C to 820 °C. During cooling, the formation of pre-eutectoid ferrite is greatly
delayed, while during reheating, the existing ferrite is not completely dissolved.
The microstructure in the case-hardened surface layer, due to the higher carbon con-
tent, had a significantly lower transformation temperature, which is why a flawless mar-
tensitic hardening microstructure with low residual austenite proportions was present in
the case hardening layer.
6.1 Process Parameters 167

Fig. 6.10 Microstructural images of the core and surface microstructure of the case-hardened pin-
ion shaft (a) mixed microstructure of martensite and ferrite in the core area; (b) high proportions
of polygonal ferrite in the core area; (c) homogeneous hardening microstructure in the case hard-
ening layer; (d) martensite and approx. 10 μm deep non-martensitic edge layer

The non-martensitic microstructure seam on the component surface is a reaction of

edge oxidation and is unavoidable in carburizing treatments in endothermic atmospheres.
The penetrating oxygen of the split carbon monoxide binds the hardenability-increasing
alloying elements and the cooling speed in the oil is then no longer sufficient for mar-
tensite formation.

Example 6—Insufficient Hardening Depth on a Forged Part

Component: Forged partwith forged pearlitic ferrite-pearlite microstructure

Material: C45, Material No. 1.0503
Error image: Hardening depth not achieved after induction hardening
Heat treatment: Inductive surface hardening
168 6 Heat Treatment Influence Area

Constant heating rate [°C/sec]

2400 1000 300 100 30 10 3 1 0,22 0,05
1100

Material C35G,1.1183
1050
homogeneous austenite
1000
Ac3 inh
(ho omo
950 mo ge
Temperature [°C]

ge neo
ne
ou us a
s a us
900 us ten
ten ite
ite
)
850 Ac3 Ferrite
+ pearli
(inhom te + aus
ogeneo te
us auste nite
nite)
800 (inhomog
eneous au
stenite)
Ac1
750

Ferrite + carbide (martensite) or ferrite + pearlite

700
0.1 1 10 100 1000 10000 1E5
Time [sec]

Fig. 6.11 Continuous TTA diagram of the steel C53G [15]

A forged part made of the unalloyed heat-treatable steel C45, 1.0503, with a forged
pearlitic ferrite-pearlite microstructure showed insufficient hardening depth after induc-
tion hardening. The entire batch was blocked until the cause was clarified.
In inductive surface hardening, the austenitization duration is always very short,
which is why the initial microstructure is of particular importance. Pre-hardening is the
optimal initial state, as the carbides are finely and evenly distributed here. Compared
to a ferrite-pearlite microstructure, the transformation temperatures are significantly
lower at the same heating rate, which can be seen from the following TTT diagram [15],
Fig. 6.11.
A coarse forged pearlitic ferrite-pearlite microstructure therefore requires special
attention when choosing the hardening temperature and holding time. The formation of
austenite begins when the Ac1temperature is reached inside the grain. The carbon dis-
solved in the austenite then has to cover large diffusion paths to the far-off ferrite net-
work, as a rather coarse grain structure is present.
Therefore, if the hardening temperature was too low and/or the holding time was too
short, homogeneous austenite did not form. In the area of the former ferrite network,
the carbon content was lower, which led to an inhomogeneous hardening microstructure
with locally very large hardness differences, Fig. 6.12a-d.
6.1 Process Parameters 169

Fig. 6.12 Microstructure images before and after induction hardening (a) Coarse-grained ferrite-
pearlite microstructure in the initial state; (b) Fine-striped pearlite and net-like arranged ferrite; (c)
Former ferrite network still clearly visible (d) 751 HV 0.2 in the grain interior; 636 HV 0.2 at the
grain boundaries

The unfavorable initial microstructure resulted in the achieved surface hardness being
at the lower limit and the required hardening depth of 2.5 +1.0 mm being undershot at
1.9 mm.
Improving the conditions for inductive surface hardening by pre-hardening often fails
due to the cost of this additional treatment. Therefore, the only option is to gradually
adjust the hardening temperature and holding time. This optimization requires very fine
adjustment of the parameters, as overheating can quickly occur due to the high power
density, which then has other disadvantages or can also lead to cracking.

Example 7—Insufficient Hardness on Wear Plates

Component: Wear plateswith a material thickness of 22 mm

Material: X210Cr12, Material No. 1.2080
Error pattern: Insufficient hardnessafter hardening and tempering
Heat treatment: Vacuum hardening and tempering
170 6 Heat Treatment Influence Area

Vacuum hardening offers optimal conditions for low-distortion hardening through gas
pressure cooling. Since gas pressure cooling quenches more gently than liquid quench-
ing media, steels with high hardenability are required for vacuum hardening. For higher
alloyed cold work steels with correspondingly good hardenability, vacuum hardening is
the most frequently used heat treatment process. Components or tools that are nitrided,
eroded or coated after hardening and tempering are usually secondary hardened from an
elevated hardening temperature, which allows a tempering temperature of min. 500 °C.
Microstructural changes are then not to be expected in the subsequent thermally stressful
treatment.
However, the error example described here of wear plates made of the material
X210Cr12, 1.2080, came from a vacuum hardening at a low hardening temperature of
950 °C. The obligatory hardness test revealed a strong hardness scatter. Wear plates with
59 HRC were measured, while other wear plates of the same type only showed 40 HRC.
Due to these large hardness differences, an investigation was carried out on two wear
plates on which the large differences were measured.
First, both wear plates were analyzed to rule out a material mix-up, Table 6.4. The
material X210Cr12, 1.2080, is no longer included in the current tool steel standard DIN
EN ISO 4957:2018-11 [16], which is why the reference to the predecessor standard was
given, which was also valid at this time.
The analyses of the two wear plates showed a very good match, which is why a mate-
rial mix-up was ruled out.
The metallographic examination made the cause of the error clear. In the microstruc-
ture of the soft wear plate, significant portions of the GKZ annealing microstructure
were still present, Fig. 6.13a-d, while the second wear plate showed a flawless hardening
microstructure. It was therefore clear that the soft wear plate did not reach the specified
hardening temperature of 950 °C.
The reasons for this could not be researched, as the heat treatment plant did not pro-
vide any further information and no batch photo was available. It would be conceivable
that several wear plates were stacked on top of each other without any space in between,
which would then have led to massive shading.

Table 6.4  Analysis results of the two wear plates

Target analysis C Si Mn P S Cr Mo V Al
X210Cr12, 1.2080 % % % % % % % % %
acc. to DIN EN ISO 1.90 0.10 0.20 max. max. 11.00 – – –
4957:2001-02
2.20 0.25 0.47 0.030 0.030 13.00 – – –
Component with 2.05 0.25 0.47 0.018 0.002 11.02 0.17 0.066 0.021
59 HRC
Component with 2.07 0.25 0.47 0.018 0.002 11.03 0.17 0.067 0.021
40 HRC
All data in mass-%
6.1 Process Parameters 171

Fig. 6.13 Microstructural images of the vacuum-hardened plates made of high-alloy cold work
steel X210Cr12 (a) Primary carbide distribution of the shattered carbide network; (b) Predomi-
nantly GKZ microstructure with pearlite and martensite; (c) Microstructure of the second plate
from the same furnace batch; (d) Flawless hardening microstructure from the same furnace batch

Example 8—Crack Formation on a Tool

Component: Tool
Material: X153CrMoV12, Material No. 1.2379
Error pattern: Crack formation after hardening and tempering
Heat treatment: Vacuum hardening and tempering

The cold work steel X153CrMoV12, material no. 1.2379, is almost exclusively hardened
and tempered in a vacuum furnace. The range of parts is very large and the compilation
of a treatment batch requires great care. The component dimensions must be comparable
in order to achieve as uniform a temperature field as possible within the charged compo-
nents.
There are two different hardening processes for this tool steel. The current tool
steel standard DIN EN ISO 4957:2018-11 [16] recommends a hardening tempera-
172 6 Heat Treatment Influence Area

Fig. 6.14 Tool made of X153CrMoV12, 1.2379, with crack formation after vacuum hardening
and tempering

ture of 1020 °C ±10 °C and a tempering temperature of 180 °C ±10 °C. Beyond this
recommendation, however, secondary hardening with austenitizing temperatures of
1060–1080 °C and multiple tempering treatments of 520–550 °C has proven to be very
advantageous for subsequent thermal treatments. Tools treated in this way can be coated
and nitrided at temperatures below the tempering temperature. This heat treatment condi-
tion is also advantageous for eroding.
Gas pressure cooling with nitrogen or argon is a relatively mild cooling compared to
liquid quenchants. Since there is also no movement of the components in the hot state in
vacuum single-chamber systems, favorable conditions for crack-free and low-distortion
hardening are given. This is the case even with heavy components with larger cross-sec-
tional differences.
Nevertheless, the tool shown in Fig. 6.14 had developed intense cracking. However,
the cracks did not form during quenching, but during heating for the third tempering
­process. From the temperature records, it was evident that during the first and second
tempering, the tool was charged into a cold oven and heated with the oven via a defined
heating ramp. In contrast, the third tempering took place in an oven already preheated
to tempering temperature. The very different cross-sections were heated at very differ-
ent rates, leading to significant thermal stresses and crack formation at various cross-
sectional positions. After the second tempering, the required hardness of 60 HRC was
already achieved, which is why a reduction of thermal stresses through plastic flow was
not possible.
During gas pressure quenching, larger temperature differences may have occurred, yet
no cracking occurred, as deformable austenite was present during quenching until the
martensite start temperature was reached. Numerous further investigations have shown
that the importance of the heating rate for tempering in terms of the risk of crack forma-
tion is often underestimated.
6.1 Process Parameters 173

Fig. 6.15 Crack formation after quenching and tempering of cylinder block blanks

Example 9—Crack Formation on Rod Sections

Component: Rod sections85 mm ∅ × 65 mm

Material: 42CrMo4+HH, Material No. 1.7225
Error pattern: Crack formation after hardening and tempering
Heat treatment: Quenching and tempering in a bell furnace

The heat-treatable steel 42CrMo4 is a very common material and is used for countless
applications. The quenching and tempering of components made of this material is rather
unproblematic, yet there was crack formation on some cylinder block blanks, Fig. 6.15.
In total, about 100 of these blanks were austenitized in a nitrogen-methanol atmosphere
with a regulated C-level of 0.4% and quenched in oil. Subsequently, they were tempered
for a sufficient length of time at 590 °C under nitrogen.
The investigation had to clarify at which process stage the cracks occurred and
whether there were material-specific reasons, as this quenching and tempering process
had been carried out on identical blanks for a long time without crack formation.
After the obligatory documentation of the delivered damaged parts, a crack was
opened on one blank, Fig. 6.16. The fracture surface, except for the small area of the lab-
oratory-produced residual fracture surface, was covered with a dense oxide layer. Since
only the cooling after tempering took place in air, this finding already indicated that the
crack must have been present before cooling to room temperature.
The fracture surface also showed no oil crack products from the quenching process.
This feature provided another clue as to the time of the crack's formation. Due to the
absence of oil crack products, there was no quenching stress crack. It must therefore
have been a heating stress crack during heating for tempering. This assessment was
reinforced by the heat treatment company’s statement that the quenched blanks were
174 6 Heat Treatment Influence Area

Fig. 6.16 Opened fracture surface with the laboratory residual fracture surface

loaded for tempering in the oven already preheated to 590 °C. The high initial hardness
after quenching, combined with large thermal stresses, was therefore the cause of the
crack. As a corrective measure, the heat treatment company implemented a significantly
reduced heating rate.
Nevertheless, it had to be clarified why cracks had occurred for the first time in this
batch with this process control, while previous batches showed no cracks.
Therefore, a homogeneity assessment of the steel batch used was carried out. The
measurement method followed the procedure described in ASTM standard A534 [17] for
bearing steels. A hardness scan with 20 hardness impressions perpendicular to the seg-
regation lines over a distance of 2 mm was performed with a low test load HV 0.2. The
hardness difference between the highest and lowest hardness value, which was 171 HV
0.2 in this scan, was then determined, Fig. 6.17.
The hardness differences were due to segregation and in the quenched and tempered
state, hardness differences of 171 HV 0.2 mark very pronounced segregations. For the
bearing steels described in ASTM standard A534 [17], a max. hardness difference of 50
HV 0.2 is required. Although the two types of steel are not directly comparable, the order
of magnitude shows that the segregation-related differences in microstructure and hard-
ness are exceptionally large.
Both influencing factors, the too rapid heating during tempering and a segregation-
related increased crack sensitivity, were the cause of the crack formation. Since the
extent of the segregations cannot be determined for each batch and assessed in terms of
a probability of cracking, the measure of lower heating speed during tempering was suit-
able for continuing to carry out crack-free tempering treatments.
Segregations in steel also have a high relevance to damage in other cases yet to be
presented. However, there is no normative description of a measurement method and
limit values.
6.1 Process Parameters 175

800

700
HVmax = 171 HV 0.2
600

500
Hardness [HV 0.2]

400

300

200

100

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Distance [mm]

Fig. 6.17 Hardness scan for determining material inhomogeneity

6.1.2 The Holding Phase

The negative consequences of incomplete Austenite Formation have been described in

the previous chapter and illustrated by examples of errors. This section presents cases
where errors have occurred either at too high a temperature or with too long holding
times.
In hypoeutectoid steels, too high an Austenitizing Temperature or too long a holding
time leads exclusively to Grain Growth. An oversaturation of the austenite with carbon,
on the other hand, is not given. This is, however, a very significant criterion for hypere-
utectoid steels. Even the TTT diagrams, e.g. of the bearing steel 100Cr6, 1.3505, [18]
show a decrease in hardness at usual hardening temperatures with increasing holding
time. The oversaturated austenite then no longer completely transforms into martensite
during quenching, which is why such hardening microstructures contain residual austen-
ite and are reduced in hardness.
In dimensionally stable tools, non-stabilized Residual Austenite is generally undesir-
able, as it can lead to the transformation of the residual austenite into martensite under
mechanical stress, thereby increasing the volume. In the case of tight fits, this can lead to
the destruction of the tool or the component. On the other hand, in martensitically hard-
ened bearing rings, a residual austenite content of 10–15% is deliberately set, which is
however stabilized by a higher tempering temperature and improves toughness.
176 6 Heat Treatment Influence Area

The Overagingis not only limited to temperatures in the austenite region. Long pre-
heating times at 450–500 °C under air lead to a pronounced oxide layer, which no longer
has a firm connection to the steel surface. In a subsequent austenitization under endo-
thermic atmosphere, these layers are indeed reduced again, but the firm connection to the
surface cannot be restored. The former oxide layer then exists as a slightly detachable
double layer. This error can also occur in nitriding or nitrocarburizing treatments [19].
Due to the very high heating rate in inductive heating, the austenitization tempera-
tures must be slightly increased compared to conventional heat treatment. This, how-
ever, leads to a very fine acicular hardening microstructure with precisely set power data.
Overheating up to the melting of the surface is also possible with faulty power data or
insufficient control, and some examples of these errors are presented.
The thermochemical heat treatments, which can also result in overtime with negative
effects, are not presented here, but in a separate chapter.
For heating to form austenite, the standards and material data sheets of the steel man-
ufacturers provide information on recommended temperature ranges. Falling below or
exceeding this temperature range can lead to faulty heat treatment results, to larger dis-
tortions and dimensional changes, or even to crack formation.
A too long holding time during austenitization is not only uneconomical, but can also
lead to an oversaturation of the austenite with carbon, just like increased temperatures,
which causes an increased amount of residual austenite in steels with a higher carbon
content. In the case of higher alloyed steels, for which higher austenitization tempera-
tures are required, there is a risk of coarse grain formation when overtime is used.
An extended holding time during tempering, on the other hand, is less critical, as the
resulting loss of hardness is only slight. The determination of the required tempering
temperature is usually based on the specification of the hardness or tensile strength in
the drawing or work order. For secondary hardening tool steels, it is also important to
specify whether a secondary hardening should be carried out, which requires higher aus-
tenitization and tempering temperatures. Secondary hardening requires multiple temper-
ing and, for the best possible dimensional stability, also deep freezing.
When quenching and tempering corrosion-resistant and hardenable materials, the
determination of the tempering temperature requires special attention. If a tempering
temperature between 500 °C and 600 °C is required for the specified hardness or tensile
strength, this leads to a significant loss of corrosion resistance, Fig. 6.18.
The drawing-based specification of a hardness or tensile strength, which requires a
tempering temperature in the temperature range of 500–600 °C, requires clarification
with the designer or client. Failure to clarify and carry out this tempering treatment not
only carries the risk of component failure, but also a liability risk. The preservation of
the corrosion resistance characteristic for this material does not have to be specified sep-
arately on the drawing or in the heat treatment order. Clarification before carrying out the
heat treatment is recommended in any case.
In numerous heat treatment processes, a preheating is carried out in a separate furnace
for various reasons. This is done for safety reasons in salt bath treatments to ensure that
6.1 Process Parameters 177

Critical
tempering area

Hardness, strength

Bending limit

Corrosion rate

Corrosion breaking load cycle number

(Cr, Fe) 7 C 3

Excretion of: (Cr, Fe) 3 C

(Cr, Fe) 23 C6

100 200 300 400 500 600 700

Tempering temperature [°C]

Fig. 6.18 Influence of the tempering temperature on the mechanical properties and corrosion
resistance of hardenable corrosion-resistant steels [20]

no moisture enters the salt bath. In nitriding and nitrocarburizing, this is associated with
pre-oxidation for surface activation. Geometrically complex components receive preheat-
ing to reduce thermal stresses.
178 6 Heat Treatment Influence Area

The preheating temperatures with oxidizing furnace atmosphere are 300–400 °C. If
the subsequent austenitization is also carried out under an oxidizing furnace atmosphere,
the oxide layer from the preheating stage can be neglected. However, if the austeniti-
zation is carried out in a reducing atmosphere, the oxide layer thickness formed dur-
ing preheating is quite relevant for the further process. At heating temperatures of over
450 °C, oxide layers are formed that no longer adhere to the base material. These very
thick oxide layers are also reduced, but after the heat treatment is completed, they form a
loose, slightly adhering, and thin film on the surface. This phenomenon is known as dou-
ble layer formation in nitriding processes.

6.1.2.1 Damage Examples—Temperature and Time

Example 1—Bursting of a Rolling Bearing Roller under Mechanical Load

Component: Rolling bearing roller

Material: 100CrMnSi6-4, Material No. 1.3520
Error image: Bursting of the rolling bearing rollerunder mechanical load
Heat treatment: Protective gas hardening and tempering

Rolling bearings inherently have only a narrow fit. Therefore, dimensional stability is a
very important property for rolling bearings. This requires that the rolling bearing com-
ponents must either be free of retained austenite or the retained austenite has been stabi-
lized. Non-stabilized retained austenite either transforms to a large extent into martensite
during deep freezing or under mechanical load. This results in volume growth and can
lead to the destruction of the bearing.
The determination of retained austenite in rolling bearing steels is very uncertain,
especially at very low contents, which is why an X-ray measurement is the method of
choice. If a certain retained austenite content has been measured, it can be checked by
deep freezing whether the retained austenite was stabilized or could still transform. Sta-
bilized retained austenite does not transform even after subsequent deep freezing. In the
context of damage investigations on rolling bearings, the X-ray measurement of retained
austenite before and after deep freezing is a standard examination.
This is also how the error image described here was handled. For a large slewing ring
bearing, rolling bearing rollers of various sizes were needed in small quantities. Since
the required dimensions were not available in the usual catalogs of the large rolling bear-
ing manufacturers, a special machine builder was commissioned to manufacture them.
After the mechanical production of the rolling bearing rollers from the rolling bearing
steel 100CrMnSi6-4, 1.3520, the heat treatment was carried out externally in a contract
hardening shop. Finally, the rolling bearing rollers were ground to size and installed in
the slewing ring bearing after delivery.
Already during the functional test of the slewing ring bearing, a serious damage
occurred. Several rolling bearing rollers literally burst, Fig. 6.19a, b. A hardness of 65.3
6.1 Process Parameters 179

Fig. 6.19 Documentation of the burst rolling bearing roller and microstructure images (a) Burst
rolling bearing roller; (b) Fragments of the burst rolling bearing roller; (c) Overview of the harden-
ing microstructure; (d) Martensite and 19% retained austenite as well as undissolved secondary
carbides

HRC was measured on the damaged rolling bearing rollers. This high hardness already
indicated that any existing retained austenite could not have been thermally stabilized.
The used rolling bearing steel showed a very uniform carbide distribution. In the mar-
tensitic matrix, a certain retained austenite content was indeed suspected in the metal-
lographic examination, Fig. 6.19c, d. However, the X-ray measured retained austenite
amount of 19.4% could not be derived from the microstructure image. After a one-hour
deep freezing in liquid nitrogen, the retained austenite content was measured again and
was then 11.0%. In the delivery state, the retained austenite content must have been even
higher, as the measurements were carried out on the already burst roller.
The investigation results thus proved a faulty microstructure condition for this appli-
cation. Retained austenite in this order of magnitude and in an unstabilized state gen-
erally carries the risk of transformation and thus volume enlargement. Hundreds of
measurements on rolling bearing rollers from various manufacturers always showed
retained austenite content, with this value being considered as the detection limit. These
bearing rollers were therefore technically free of retained austenite.
180 6 Heat Treatment Influence Area

Example 2—Bearing Shell Fracture After Very Short Use

Component: Bearing shelland Bearing rollers

Material: 100Cr6, Material No. 1.3505
Error pattern: Fracture of the bearing shells after very short usage duration
Heat treatment: Protective gas hardening and tempering

The microstructure condition in bearing parts with large amounts of non-stabilized resid-
ual austenite is one of the very frequently detected error characteristics. While in the
previously described Example 1 the bearing roller itself was destroyed due to volume
growth, the following example shows a destruction of the bearing shell.
The bearing shell documented in Fig. 6.20a shows a broken out segment. On the run-
ning surface, 6 bearing rollers are still wedged between both flange sides. Both flange
sides were severely damaged after a very short usage duration of approx. 200 h, Fig. 6.20b.
The damage pattern indicated intensive contact with the end faces of the bearing rollers.

Fig. 6.20 Documentation of the broken bearing shell, the compressed bearing roller and the
microstructure formation (a) Broken bearing shell with wedged bearing rollers; (b) Severely dam-
aged inner flange side; (c) Plastic deformation due to compression; (d) Martensite with 12% resid-
ual austenite (measured by X-ray)
6.1 Process Parameters 181

Table 6.5  Results of residual austenite determinations with different measurement methods [21]
Austenitization X-ray measure- Magnetic meas- Magnetic meas- Metallographic
temperature and ment residual urement calibra- urement calibra- estimate in mass
duration austenite content tion 3 in mass % tion 5 in mass % %
in mass %
820 °C 60 min 11.1 ± 3.0 10.8 ± 0.1 11.2 ± 0.1 0–10
840 °C 60 min 14.1 ± 3.0 13.4 ± 0.1 13.4 ± 0.1 0–15
860 °C 60 min 16.2 ± 3.0 15.5 ± 0.1 15.8 ± 0.1 0–15
880 °C 60 min 19.7 ± 3.0 17.7 ± 0.1 17.7 ± 0.1 0–30

The bearing rollers should have a length of 65.00–0.05 mm, which could be inferred
from the drawing. The determination of the roller length on the still available bearing
rollers resulted in values between 65.00 and 65.13 mm. Since it could be assumed that
the bearing rollers had the drawing dimensions at the time of installation, there must
have been an increase in volume during operational load.
The metallographic examination clearly showed that the end face of the bearing roller
had been compressed, Fig. 6.20c. The residual austenite content was measured by X-ray
on the bearing roller and a value of 12% residual austenite was determined. At the time
of bearing assembly, the residual austenite content must therefore have been even higher,
as there had already been a clearly measurable increase in length.
The hardening microstructure did not clearly reveal the residual austenite content of
12% in the metallographic examination, Fig. 6.20d. Especially with tightly toleranced
bearing components, a metallographic determination of residual austenite is fraught with
great uncertainties.
A comparative study of different measurement methods for determining the resid-
ual austenite content was carried out as part of a bachelor’s thesis at the University of
Bremen, Table 6.5. In particular, the metallographic estimate shows very large scatter
widths.

Example 3—Quench Cracking on a Washer

Component: WasherOuter-∅ 78 mm; Hole-∅ 44 mm

Material: C45E, Material No. 1.1191
Error pattern: Quench cracking<Error pattern—Heat treatment Temperature+Time>Quench
cracking</Error pattern—Heat treatment Temperature+Time> after edge
overheating
Heat treatment: Protective gas hardening, tempering and hot-dip galvanizing<Heat
treatment>Protective gas hardening, tempering and hot-dip galvanizing</Heat
treatment>
182 6 Heat Treatment Influence Area

The function of washers is generally known. They absorb the friction and pressure forces
that occur at the nut or screw head during tightening. They also distribute the tightening
forces over a larger area.
For large-volume screw connections, washers with an outer diameter of 78 mm and
a hole diameter of 44 were made from the unalloyed heat-treatable steel C45E, 1.1191.
After mechanical manufacturing, the washers were quenched and tempered. The target
specification was 300 +70 HV. Finally, the discs were hot-dip galvanized at 550 °C to
protect against corrosion. During the final inspection, several cracked washers were dis-
covered, Fig. 6.21a.
In the metallographic examination, local coarse grain formation was observed at the
edges of the washer, Fig. 6.21b, d. Individual grains with a grain size index of G = −1
were present. During hot-dip galvanizing, zinc had also penetrated into the crack open-
ing, Fig. 6.21c.
For the steel C45E, the standard specifies a hardening temperature range of 820–860
°C. The TTAdiagram shows that no coarse grain formation can occur at this temperature
range, Fig. 6.22. It was also noticeable that the coarse grain formation had only occurred

Fig. 6.21 Documentation of crack formation in the washer and microstructure images (a) Crack
formation; (b) Crack progression and local coarse grain formation at the edges; (c) Zinc also on
the crack flanks; (d) Coarse grain formation with grain sizes up to G = −1
6.1 Process Parameters 183

Material

inhomogeneous inhomogeneous
Temperature [°C]

.
Time [sec]

Fig. 6.22 Isothermal time-temperature-austenitizing diagram of the steel C45E—Image source

[15]

at the edges of the washer. From this, it had to be concluded that a significantly higher
furnace temperature with reduced dwell time had been present, leading to overheating of
the edges and the resulting coarse grain formation with crack formation.
A temperature-time documentation was not provided for the investigations.

Example 4—Coarse Lamellar Pearlite in Forged Hand Tools

Component: Forged hand tools

Material: C67, Material No. 1.0603
Error image: Coarse lamellar pearlite after annealing
Heat treatment: soft annealing Annealing

Faulty microstructures can not only be caused by overheating or overtimes in the austen-
ite region. This example describes a too high soft annealing temperature, which led to an
undesired microstructure formation. From hot-rolled bars of the tool steel C67, material
no. 1.0603, with a diameter of 20 mm, tool blanks were drop-forged and subsequently
soft annealed before mechanical processing. The soft annealing treatment was carried out
at a temperature of 710 °C for 6 hours. Subsequently, it was slowly cooled in the furnace.
The small-volume tools were charged in wire mesh baskets as bulk material and three
baskets were stacked on top of each other. The hood annealing furnace was directly gas
184 6 Heat Treatment Influence Area

Fig. 6.23 Microstructure in the delivery state, after drop forging and after annealing (a) Fer-
rite-pearlite microstructure in the delivery state +U; (b) Coarse ferrite-pearlite microstruc-
ture after drop forging; (c) Coarse lamellar pearlite with pearlite fraction PA = 3.9; (d) Globular
soft annealed microstructure with pearlite fraction PA = 3.2

heated and an oxidation-poor annealing treatment was ensured by precise setting of the
lambda value.
The hot-rolled bars were untreated and exhibited a fine-grained ferrite-pearlite micro-
structure without any abnormalities, Fig. 6.23a. After drop forging, the ferrite-pearlite
microstructure was expectedly significantly coarsened, Fig. 6.23b. Completely unex-
pected, however, was the formation of the soft annealed microstructure in the upper
charging basket, Fig. 6.23c. There was no globular formation present and the pearlite
was very coarse. In contrast, the cementite formation in tools from the other two basket
layers was globular and met the target specification, Fig. 6.23d.
The chosen annealing temperature of 710 °C was just below the Ac1 temperature of
726 °C calculated for this melt batch. However, the coarse lamellar pearlite formation
could only have occurred if the Ac1 temperature was significantly exceeded during the
soft annealing treatment. This led to the formation of austenite with complete carbide
dissolution. During the subsequent very slow cooling, the eutectoid transformation of the
austenite occurred with the observed pearlite formation.
6.1 Process Parameters 185

The temperature documentation of the annealing furnace did not show the overtem-
perature. The 710 °C were strictly adhered to and the correct formation of the annealing
microstructures in the two lower baskets confirmed this. From this it was concluded that
there was a significant overtemperature in the upper part of the hood furnace, which was
not detected by the control thermocouple.
Calculation of the Ac1 temperature of the subsequent analysis [22]:

C Si Mn P S Cr Ni Mo V
% % % % % % % % %
0.67 0.21 0.71 0.015 0.006 0.25 0.03 0.01 0.01

Ac 1 = 739−22∗%C+2∗%Si−7∗%Mn+14∗%Cr+13∗%Mo−13∗%Ni+20∗%V

Result: Ac1 = 726 °C

Example 5—Melting on Shafts After Induction Hardening

Component: Shafts ∅ 38 mm
Material: C45, Material No. 1.0503
Error pattern: Melting after induction hardening
Heat treatment: Inductive surface hardening and tempering

Inductive surface hardening is a very common heat treatment process encountered in

practice. The short treatment duration for surface hardening of series parts is particularly
impressive. However, the very rapid heating requires precise power and temperature con-
trol, especially since pre-hardening is rarely carried out for cost reasons.
The continuous time-temperature-austenitization diagram of the C53G steel clearly
shows that significantly higher temperatures are required to form homogeneous austenite
in the normally annealed state than after pre-hardening, Fig. 6.24. With a heating rate
of 300 °C/s, the area of homogeneous austenite of a pre-hardened component is already
reached at 850 °C. If, on the other hand, a ferrite-pearlite microstructure is present,
homogeneous austenite is only reached at an austenitization temperature of 990 °C.
The available power during inductive surface heating is quite capable of melting or
causing surfaces to melt. This creates cavities that predominantly start at triple points
of the grain boundaries, Fig. 6.25a, b. At the same time, this significant overshoot of the
austenitization temperature is associated with coarse grain formation, Fig. 6.25c, d.
In the example shown here, there were individual grains of the grain size class G = 1,
which is extremely unusual for inductive surface hardening. Usually, the very short treat-
ment duration at the correct temperature leads to a very fine acicular martensite. The
coarse grain was not a rejection criterion in this example. The melting, on the other hand,
meant that these shafts had to be discarded. Reworking is not possible with melting.
186 6 Heat Treatment Influence Area

Constant heating rate [°C/sec]

2400 1000 300 100 30 10 3 1 0,22 0,05
1100

Material C35G,1.1183
1050
homogeneous austenite
1000
Ac3 in
(hohom
mo oge
Temperature [°C]

950 ge ne
ne ou
ou s a
s a us
900 us ten
ten ite
i te
)
850 Ac3 Ferrite
(inhomo+ pearlite + au
geneous ste
austenitnite
e)
800 inhomogen
eous aust
enite
Ac1
750

Ferrite + carbide (martensite) or ferrite + pearlite

700
0.1 1 10 100 1000 10000 1E5
Time [sec]

Fig. 6.24 Continuous time-temperature-austenitization diagram of C53G steel—Image source

[15]

Example 6—Crack Formation in the Tooth Root Area of a Rack

Component: Rack
Material: C45E, Material No.: 1.1191
Error pattern: Crack formation in the tooth root area
Heat treatment: Inductive full tooth hardening and Tempering

A similar error, as in the previous Example 5, occurred during the inductive harden-
ing of racks made of the unalloyed heat-treatable steel C45E+N. A full tooth hardening
was specified, the heat treatment pattern of which is defined in DIN ISO 15787 [23],
Fig. 6.26a. After the inductive hardening of a larger number of racks, crack formations
were detected in the tooth root area.
A profile cut through two teeth made the hardness profile visible, Fig. 6.26b. Even
the macroscopic representation revealed the incorrectly formed contour of the hardening
layer. The tooth root areas showed insufficient hardening layers on one side. The heat
treatment pattern was also uneven in the axial direction. In addition, a crack-like defect
was visible on one flank side.
6.1 Process Parameters 187

171.5 µm
130.6 µm

140.4 µm
255.9 µm

145.8 µm

197.7 µm

81.3 µm

103.5 µm 146.1 µm
113.2 µm

Fig. 6.25 Documentation of the melting and coarse grain (a) Melting in the inductively heated
surface layer; (b) Cavities at melted grain boundary triple points; (c) Coarse acicular hardening
microstructure; (d) Individual coarse grains with max. grain size of G = 1

Fig. 6.26 Documentation of target and actual heat treatment patterns (a) Heat treatment pattern of
full tooth hardening [23]; (b) Faulty hardening pattern
188 6 Heat Treatment Influence Area

The insufficient heating caused a mixed microstructure of undissolved ferrite, mar-

tensite, and bainite in the left part of the tooth root, Fig. 6.27a, b. The right part of the
tooth root, on the other hand, showed a martensitic hardening microstructure.
The macroscopically detected crack formation, which had led to this investiga-
tion, turned out to be a melting near the tooth root area, Fig. 6.28a-d. The melted mate-
rial solidified below the melt cavity. There was also an unusually massive scale layer
of approx. 15 μm. The grain boundaries were also melted and the martensite was very
coarse.

Example 7—Difficult to Remove Iron Foil on Small Module Gears

Component: Small module gearswith pitch circle diameter 110 mm

Material: 16MnCr5+FP, Material No.: 1.7131
Error pattern: Difficult to remove thin iron foil
Heat treatment: Case hardening and tempering

The error feature to be treated in this example is very rare after case hardening, but
more often observed during nitrocarburizing. In an internal quench furnace, carburiz-
ing was carried out in the main chamber and the subsequent batch was preheated in the
pre-chamber. The pre-chamber was not temperature controlled and no temperature was
measured. This practice, which had been proven over many years, showed no abnormali-
ties.
A batch of case-hardened small module gear components then stood out during clean-
ing blasting after case hardening and tempering. A thin, foil-like layer was present on
the surface, which could be manually removed, but could not be eliminated by blasting,
Fig. 6.29a .

Fig. 6.27 Documentation of the inductive rack hardening (a) Incomplete and one-sided formation
of the hardening layer; (b) Mixed microstructure in the tooth root area
6.1 Process Parameters 189

Fig. 6.28 Melting and scaling (a) Melting in the tooth root area; (b) Crack formation and massive
scale layer; (c) Melting in the tooth root area; (d) Coarse martensite and grain boundary melting

The metallographic examination carried out in the unetched state revealed a detach-
able layer and a structurally similar layer, however, still adhering to the surface,
Fig. 6.29b. At the surface position, the outer double layer was detached, Fig. 6.29c. The
surface layer showed the typical non-martensitic edge as an unavoidable result of inter-
granular oxidation, Fig. 6.29d.
Since this error pattern was new in the heat treatment operation, the batch history was
followed in detail. It turned out that the preheating duration for this conspicuous batch
was very long, as a special program with a very large carburizing depth was run in the
main chamber. Due to the very long dwell time in the preheating chamber, a high pre-
heating temperature with a very pronounced oxide layer had occurred. The no longer
adhering oxide layer was reduced in the subsequent carburizing process of this batch and
formed a thin, peelable foil.
Since no temperature measurement was installed in the pre-chamber, preheating tests
were carried out in a laboratory oven. This behavior could be reproduced at a preheating
temperature of min. 500 °C.
This certainly very rare error was nevertheless a reason to re-evaluate the preheating
conditions.
190 6 Heat Treatment Influence Area

Fig. 6.29 Documentation of the conspicuous surface (a) Partially detached double layer; (b)
Reduced oxide layer and overlying double layer; (c) Reduced oxide layer after removal of the dou-
ble layer; (d) Non-martensitic surface layer as a result of intergranular oxidation

Example 8—Incomplete Isothermal Transformation of a Spur Gear

Component: Spur GearModule 8

Material: 18CrNiMo7-6, Material No. 1.6587
Error Pattern: Incomplete Isothermal Transformation
Heat Treatment: Case Hardening after Isothermal Transformation

Even with fine-grain stabilized case-hardening steels, there is a risk of coarse grain for-
mation at very large case hardening depths, which require very long carburizing times.
The case hardening after isothermal transformation described in DIN 17022-3:1989-04
[14] leads to grain refinement with sufficiently rapid cooling to the isothermal trans-
formation temperature between 550 °C and 650 °C and holding at this temperature,
Fig. 6.30.
As part of a first sample inspection, a patchy microstructure was discovered in the
case hardening layer during the metallographic examination, Fig. 6.31. At higher magni-
fication, significantly higher residual austenite contents were visible in the bright spots.
6.1 Process Parameters 191

AC3 Core
Temperature

AC1 Edge
Converting Hardening

Tempering
Carburizing
Time

Fig. 6.30 Schematic temperature-time course of case hardening after isothermal transformation

This microstructure was created by an incomplete isothermal transformation. The

bright areas with the higher residual austenite content had not yet been transformed in
the pearlite stage. During the subsequent reheating to the hardening temperature and the
final quenching, a microstructure was present in these areas as in a direct hardening.
The areas that had been isothermally transformed did not dissolve the entire carbon
content during reheating, which is why the residual austenite content was significantly
lower and finely precipitated carbides were present.
The failure to achieve the complete transformation end can be attributed to two possi-
ble causes. This becomes clear when considering the isothermal time-temperature-trans-
formation diagram, Fig. 6.32.

1. The reheating occurs even at the shortest possible transformation duration of 1 ¾ h at

640 °C before this time has elapsed.
2. The optimal transformation temperature is not reached and occurs at 660 °C. At
this temperature, the transformation end only occurs after 4 ½ h.
Reheatingtoo early at the optimal transformation temperature is rather unlikely in most-
cases. On the other hand, too dense charging or insufficient temperatureuniformity of
the furnace can indeed lead to increased transformationtemperatures in the compo-
nent, even though the furnace thermocouple indicatesthe correct temperature.

Another influencing factor on the failure to achieve the transformation end is pro-
nounced segregation in the material. An error example is presented in the material influ-
ence area for this.

Example 9—Overlong Annealing of a Gear

Component: GearPitch circle diameter 36 mm

Material: HS12-1-4, Material No. 1.3302
Error image: Control on the vacuum furnace defective, therefore 24 h annealing duration
Heat treatment: Vacuum hardening
192 6 Heat Treatment Influence Area

Fig. 6.31 Patchy microstructure in the case hardening layer in the area of the tooth flanks; (a)
Noticeable patchy microstructure; (b) Coarse martensite and residual austenite in the bright micro-
structural area; (c)+(d) Ground tooth flank with patchy microstructure without non-martensitic
microstructural seam; (e)+(f) Unground tooth flank with patchy microstructure and non-martensi-
tic microstructural seam

The number of possible errors in heat treatment processes was expanded in this case by
a plant malfunction. Small gears with a pitch circle diameter of 36 mm were made from
the high-speed steel HS12-1-4, 1.3302, and hardened in a vacuum furnace. The chosen
temperature-time sequence had been run for a long time without any disturbances. How-
ever, in this error batch, there was a furnace malfunction and the furnace remained at the
austenitizing temperature of 1200 °C for an entire night.
6.1 Process Parameters 193

1200

1100
18CrNiM7-6, 1.6587, comparison of isothermal transformation at 640 °C and 660 °C
1000

900 Ac3 (820 °C)

4:30 h ati 660 °C
800
Temperature [°C]

700 Ac1 (730 °C)

170
Ferrite Perlite 179 HV
600
Austenite
1:45 h at 640 °C
500
Bainite
400 Ms (395 °C)
308

300 Martensite
200

100

0
0.1 1 10 100 1000 10000 1E5 1E6
Time [sec]

Fig. 6.32 Isothermal TTTdiagram of the case-hardening steel 18CrNiMo7-6 and determination of
thetransformation times at 640 °C and 660 °C

Since a strong grain coarsening was to be assumed, a grain refinement annealing was
carried out at 920 °C and 820 °C with long holding times each. After completion of this
“repair annealing”, hardening and tempering were carried out again according to the
standard program. Gears were sent for sampling as a precaution.
The Snyder-Graff grain number was measured at five different positions and the aver-
age value of 11.2 was within the usual range and could be accepted, Fig. 6.33.
The surface structure, on the other hand, was greatly altered, Fig. 6.34a-d. The bright
microstructural seam was caused by a high amount of residual austenite. In the untem-
pered state, 100% residual austenite was measured by X-ray. After tempering three
times, the residual austenite was not completely transformed and the bright microstruc-
tural seam was still present.
An emission spectroscopic analysis of the surface revealed a nitrogen content of
0.21%. During the entire vacuum treatment, nitrogenation occurred due to dissociated
nitrogen. It could not be determined retrospectively at which treatment step the nitro-
genation occurred.
In the base material, the X-ray measurement of residual austenite yielded a value
below the detection limit of 2%. The surface was not measured again by X-ray, but the
surface hardness only reached a value of 575 HV.
194 6 Heat Treatment Influence Area

Fig. 6.33 Microstructure in the hardened (a) and hardened and tempered state (b)

Fig. 6.34 Greatly altered microstructure throughout the edge layer (a) Microstructure in the tooth
head (not tempered); (b) 100% residual austenite at the surface (not tempered); (c) Microstructure
on the tooth flank (hardened and tempered); (d) Very high residual austenite content at the surface
(hardened and tempered)
6.1 Process Parameters 195

6.1.3 The Quenching Phase

Depending on the objective of the heat treatment, cooling or quenching takes place after
the holding phase. Cooling in still air is typical for normalizing. For higher alloyed case-
hardening steels, e.g. 18CrNiMo7-6, this cooling speed may still be too high and bainite
or martensite formation may occur. These steels are therefore transformed isothermally
in the pearlite stage after rapid cooling to approx. 640 °C. For very large dimensions, a
pre-hardening with high tempering temperature is often alternatively carried out.
Quenching is necessary for the formation of a hardening microstructure. The choice
of quenching medium is determined by the hardenability and component geometry. The
greatest temperature differences in the component occur during quenching. These and
the microstructural transformations cause thermal stresses or transformation stresses in
the component. To simplify understanding of the processes, a rod with a diameter of 100
mm, which was quenched in water after heating to 850 °C, is described as an example.
The thermal stresses occurring during quenching can have different effects. If the
stress amounts do not exceed the elastic range, the thermal stresses are completely
relieved after the cooling is completed and the temperature is fully equalized. If, on the
other hand, the warm yield limit is exceeded, there is a plastic deformation in the compo-
nent, which is perceived as distortion. The occurring tensile or compressive stresses are
always equal in size, or add up to zero.
Case 1: Transformation-free cooling with elastic deformation
When a cylindrical rod with a diameter of 100 mm is quenched in water from 850
°C, temperature differences of max. 525 °C occur between the surface and the core, Fig.
6.35a. These temperature differences cause high thermal stresses. If the steel were able
to withstand these stresses elastically, a stress amount of 625 MPa tensile or compressive
stresses would result, Fig. 6.35b. After complete cooling and temperature equalization in
the component, no stresses would remain.

1000 800
Temperatures
600
Core Tensile stresses in
800
Edge 400
the edge
Temperature [°C]

Difference
Tension [MPa]

200
600
0
400 -200
Compressive stresses
-400
200 in the core
-600

0 -800
0.001 0.01 0.1 1 10 100 0.001 0.01 0.1 1 10
Time [min] Time [sec]

Fig. 6.35 Cooling curves and thermal stresses for surface and core during water cooling of a rod
with 100 mm Ø
196 6 Heat Treatment Influence Area

The question still to be clarified is why tensile and compressive stresses occur during
quenching in the form described, as shown in the graph in Fig. 6.35b. Tensile stresses
occur at the surface and compressive stresses in the core. This can be explained by a sim-
ple model. The rod is conceptually divided into an surface and a core area and the behav-
ior is first considered separately and then combined.
After thorough heating, both the surface and the core have an identical temperature,
Fig. 6.36-1. Since there are no temperature differences, there are also no thermal stresses,
Fig. 6.36-2. Now, when the quenching process is started, the cooling begins at the sur-
face, while the core still has the full initial temperature of 850 °C at this chosen point in
time. A temperature difference between the surface and the core arises, Fig. 6.36-3. For
the component volume, this means that the surface shrinks and the core does not yet,
Fig. 6.36-4. Since the surface and core were only separated in a model-like manner, the
two areas now have to be joined together again. This can only be done if the surface is
stretched by tensile stresses and the core is compressed by compressive stresses.
After reaching room temperature and after complete temperature equalization, there
are no more temperature differences, Fig. 6.36-5. Since a purely elastic behavior was
assumed in a model-like manner, the thermal stresses have been completely reduced after
the temperature equalization, Fig. 6.36-6. This state is hardly achievable with a water
quench, however, it makes the systematics of the occurring stress directions clearly rep-
resentable.
Case 2: Transformation-free cooling with plastic deformation
If the thermal stresses occurring during quenching exceed the warm yield limit, there
is plastic deformation and a reduction in stress, Fig. 6.37a, b. As a result of the plastic
deformation, there is a reversal of stress and after complete cooling, compressive stresses
are present at the surface and tensile stresses in the core.
The development of stresses and stress directions again follows the model previ-
ously described. The initial state at 850 °C is identical to the state shown in Fig. 6.36-1,
-2, which is why the repetitive representation was omitted. Volume changes and ther-
mal stresses, Fig. 6.38-1, -2 also occur in the same direction and size. However, in this

1 2 3 4 5 6

Fig. 6.36 Temperature and volume changes during cooling (transformation-free cooling without
plastic deformation)
6.1 Process Parameters 197

1000 800
Temperatures
600
Core Tensile stresses
800
Edge 400
in the edge
Temperature [°C]

Difference

Tension [MPa]
FS
200
600
0 lg t
400 -200 FS

Compressive
-400
200 stresses in the core
-600

0 -800
0,001 0,01 0,1 1 10 100 0,001 0,01 0,1 1 10
Time [min] Time [sec]

Fig. 6.37 Cooling curves and thermal stresses for surface and core during water cooling of a rod
with 100 mm Ø

1 2 3 4

Fig. 6.38 Temperature and volume changes during cooling (transformation-free cooling with
plastic deformation)

case, the warm yield limit is reached and there is plastic deformation, as was shown in
Fig. 6.37b. The compressive stresses in the core lead to a compression of the core. After
complete cooling, the core area is therefore smaller than the surface area. To compen-
sate for this volume difference, tensile stresses now occur in the core and compressive
stresses at the surface. Compared to case 1, there is therefore a reversal of stress in the
surface and core during plastic deformation during cooling, Fig. 6.38-3, -4.
In transformation-free, e.g. austenitic, materials, this progression of stresses occurs
during cooling. An austenitic temperature dummy body, as is often used in vacuum fur-
naces, shows a clear bulge after numerous quenching cycles.
Case 3: Through-hardening with plastic deformation
The two previously described case examples did not involve transformation processes.
This case example now explains a martensitic hardening across the entire cross-section
while simultaneously evaluating the thermal stresses. The two cooling curves for the
surface and the core in Fig. 6.39a reach the Mstemperature without prior microstructural
transformation. However, martensite formation begins at a later point due to the slower
198 6 Heat Treatment Influence Area

900 800
600 Yield strength
Edge Edge
Core 400
P

Tension [MPa]
600 200
Temperature [°C]

0
B
Edge -200
300 Ms
-400
Core Core
-600
0 -800
1 2 4 8 15 30 60 2 4 8 15 30 60 0,001 0,01 0,1 1
[sec] Time [min] Time [sec]

Fig. 6.39 Thermal and transformation stresses for surface and core with martensite formation
only at the edge

cooling in the core. This results in a double stress reversal during cooling. After cooling,
tensile stresses are present at the surface and compressive stresses in the core, Fig. 6.39b.
The stress development during quenching is also described here in a model-like manner.
In Fig. 6.40-1, -2, the effects of thermal stresses can also be seen initially. Tensile
stresses occur at the surface and compressive stresses in the core. When the Mstempera-
ture is reached at the surface, martensite formation begins (t1 in Fig. 6.39b). The surface
reaches the martensite start point Ms first. The transformation from austenite to martensite
is associated with an increase in volume. To eliminate the conceptual separation of surface
and core, the surface must now be compressed and the core stretched, which is why com-
pressive stresses occur at the surface and tensile stresses in the core, Fig. 6.40-3, -4.
The core reaches the Ms temperature at a later point, at which significant propor-
tions of austenite in the surface have already been transformed into martensite. The mar-
tensite formation now beginning in the core increases the core volume, leading to tensile
stresses at the surface and compressive stresses in the core, Fig. 6.40-5, -6.

1 2 3 4 5 6

Fig. 6.40 Temperature and volume changes during cooling (through-hardening with plastic defor-
mation)
6.1 Process Parameters 199

This stress state after quenching is very prone to cracking, as the surface already has a
tensile preload. The risk of cracking increases even more when very high-strength mate-
rials with possibly complicated geometries are hardened. A timely tempering treatment
is therefore indicated for such steels. Often, the lowest possible temperature is not aimed
for during quenching, but tempering already takes place after reaching 50–60 °C.
Case 4: Surface hardening Case hardening with plastic deformation
The fourth and final case example deals with steels with lower hardenability, which,
even after maximum quenching treatment in water, only transform a few mm at the
surface into martensite. The core undergoes the transformation in the perlite stage and
transforms there close to equilibrium, Fig. 6.41a, while the martensite formation after
reaching the Ms temperature at the surface only begins at a later point.
After the end of the quenching treatment, compressive residual stresses are present at
the surface and tensile residual stresses in the core, Fig. 6.41b.
Since only the surface transforms into martensite, while a ferrite-pearlite microstruc-
ture is formed in the core, Fig. 6.42-3, -4, a larger surface volume is present at the end of
cooling, Fig. 6.42-5, -6. The connection between surface and core can then only occur if
the surface is compressed and the core is stretched. Therefore, in a case hardener, com-
pressive stresses are present at the surface and tensile stresses in the core.

900 800
Core 600 Yield strength
Edge Core
400
Temperature [°C]

P
Tension [MPa]

600 200
0
B
Edge -200
300
Ms -400
Core Edge
-600
0 -800
1 2 4 8 15 30 60 2 4 8 15 30 60 0.001 0.01 0.1 1
[sec] Time [min] Time [sec]

Fig. 6.41 Thermal and transformation stresses for surface and core with martensite formation
across the entire cross-section

1 2 3 4 5 6

Fig. 6.42 Temperature and volume changes during cooling (surface hardening with plastic defor-
mation)
200 6 Heat Treatment Influence Area

This stress distribution is very advantageous for numerous applications, as the risk of
crack formation is low. Insertion tools for pneumatic hammers, for example, are made
from an unalloyed tool steel. At the chisel surface, there is a high hardness and wear
resistance with compressive residual stresses, while in the core there is still sufficient
toughness and the tensile residual stresses do not have a negative impact.
Shell hardeners therefore have a low tendency to crack even with rough water quench-
ing. Often, tempering is even dispensed with in order to maintain maximum hardness.
The four case studies described did not yet include any statement on quenching stress
cracks or hardening cracks. Plastic deformation and the resulting stress reduction require
a deformable microstructure. If the deformability is no longer given due to an already
occurred martensite formation, quenching stress cracks can occur. Both in the micro-
scopic microstructure examination and in the scanning electron microscopic examination
of the crack surface, an intercrystalline crack course is recognizable.
Already in the description of the heating stress cracks during tempering, it became
clear that the intercrystalline crack course alone does not yet allow a clear assignment
to the time of origin. Residues of the quenching agent on the crack surface, such as salt
residues or oil cracked during tempering, are clear characteristics of a quenching stress
crack.
The well-known rule applies to the required quenching speed, to quench a component
only as fast as necessary or to quench as slowly as possible. This rule is very intuitive
for minimizing the stresses that occur during quenching. Increasing the quenching speed
beyond the upper critical cooling speed of martensite formation is no longer associated with
a change in microstructure, but increases the risk of crack formation during quenching.
On the other hand, not reaching the upper critical cooling speed sometimes causes
undesirable microstructure formations. Also of importance are too high cooling speeds,
if, for example, equilibrium-like microstructure transformations are to occur in the pearl-
ite stage.
Insufficient cooling speed can be caused by several factors. These include the harden-
ability of the steel, the dimension of the component, the batch size, the bulk density for
mass parts, the quenching agent, the quenching bath for liquid quenching agents with its
specific design, and the conditions during gas pressure cooling. This list already makes it
clear that numerous error possibilities can occur.
For the assessment of the required quenching speed, the continuous TTT diagram-
sprovide valuable information. However, it must be mentioned as a limitation that the
TTT diagrams are only valid for the specified composition and larger deviations in the
chemical composition can cause significant differences in both directions, [24]. For sin-
gle part quenching, extensive calculation possibilities are available. Using FEM calcula-
tion, the cooling speeds can be calculated quite practically. The Stahlwissen®database
has been using this tool for decades [25].
For the calculation of quenching speeds in complex batches and bulk goods, how-
ever, no calculation models are known, which is why in these cases either experimental
6.1 Process Parameters 201

measurement must be used or the experience of the employees must be drawn upon. In
the case of gas pressure cooling in a vacuum furnace, a direct temperature measurement
in the center of the batch is relatively easy, as the batch does not have to be moved for
quenching in single-chamber systems. For all other quenching systems, the effort for a
temperature measurement during quenching is much greater and sometimes completely
impossible.
Due to the relatively mild quenching effect of nitrogen gas pressure cooling, there is
a risk of pre-eutectoid carbide precipitation in large components, which leads to losses
in toughness. For steels with high residual austenite contents, the cooling speed is also
important at low temperatures. Studies within the T.F.W.W. project 10/2007 have shown
that slow cooling results in residual austenite stabilization [26]. These differences remain
even after the tempering processes, Fig. 6.43. Only deep freezing to −196 °C leads to
both a reduction in the residual austenite content and an equalization of the influence of
the cooling speed, Fig. 6.44.
The cooling rate after solution annealing of precipitation hardening steels must also
be considered. Only a rapid, precipitation-free quenching guarantees a max. hardness in
the final aging.
This also applies to solution annealing treatments of corrosion-resistant steels. Too
slow cooling leads to chromium carbide precipitations at the grain boundaries, which in
turn can lead to a reduction in corrosion resistance up to grain disintegration.

60

Hardened
50 1 x tempered 530 °C
2 x tempered 530 °C
Retained austenite content [%]

40

30

20

10

0
0 100 200 300 400 500 600 700
Cooling time 800 °C > 500 °C [sec]

Fig. 6.43 Residual austenite contents after hardening and tempering depending on the cooling
speed (without deep freezing after quenching)
202 6 Heat Treatment Influence Area

60
Hardened

deep-frozen -196 °C
50
1 x tempered 530 °C 1h

2 x tempered 530 °C 1h

40
Retained austenite [%]

30

20

10

0
0 100 200 300 400 500 600 700
Cooling time 800 °C → 500 °C [sec]

Fig. 6.44 Residual austenite contents after hardening and tempering depending on the cooling
speed (with deep freezing after quenching)

6.1.3.1 Damage Examples—Quenching

Example 1—Crack Formation on a Measuring Pin

Component: Measuring pin

Material: X90CrMoV12, Material No. 1.4112
Error image: Crack formation after hardening and Tempering
Heat treatment: Hardening with oil quenching in the internal quench furnace

The material X90CrMoV18, Material No. 1.4112, has very good hardenability and is
therefore ideally suited for vacuum hardening. This hardening process has also been used
in the past for numerous measuring pins of different dimensions. However, for a current
production series, there was a bottleneck as the available vacuum furnace was occupied
for a long time for other tools. Therefore, it was decided to harden in an internal quench
furnace followed by tempering in a circulating air furnace.
The austenitization was carried out at 1000 °C over the three preheating stages
350 °C, 650 °C, 880 °C. After temperature equalization at 650 °C, the batch was trans-
ferred to the internal quench furnace, which had the temperature of the third preheating
6.1 Process Parameters 203

stage of 880 °C. The furnace was heated up to the hardening temperature of 1000 °C
after the temperature equalization together with the batch.
After a sufficient holding time of 1 h, the batch was quenched in oil, which is the pre-
dominantly used quenching medium for an internal quench furnace. At approx. 80 °C,
the batch was moved into the washing machine and washed at the same temperature. The
tempering in the circulating air furnace was carried out at 150 °C over 3 h after tempera-
ture equalization.
After completion of the entire heat treatment process, crack formation was found on
all measuring bodies treated in the batch, Fig. 6.45. The cracks ran over the entire length
with the starting point at the groove edge. The crack depth reached into the middle of the
component. In the hardening shop, a spot was ground to measure the hardness, which at
58 HRC met the target specification of 57 +2 HRC.

Fig. 6.45 Measuring pin after hardening and tempering with grind for hardness testing in the
hardening shop
204 6 Heat Treatment Influence Area

The chemical composition of the used steel melt corresponded in all points to the tar-
get composition of the material X90CrMoV18, Material No. 1.4112. The X-ray deter-
mined residual austenite content was 6%.
In the metallographic examination, the outer and crack flank surfaces were first evalu-
ated in the unetched state. As expected, the outer surface showed intergranular oxidation,
which is unavoidable in austenitizations in endothermic furnace atmospheres, Fig. 6.47a,
b. On the other hand, there was no intergranular oxidation on the crack flank surface,
which is why a heating stress crack could be ruled out. If a crack had occurred during
heating, the crack flank surface would also have to show intergranular oxidation.
The microstructure formation showed an inconspicuous and low tempered hardening
microstructure with primary and secondary carbides, Fig. 6.46c, d.
The most striking finding was the crack initiation position in connection with the oil
quenching. The sharply executed groove base led to a high notch factor and the oil quench-
ing to a large temperature difference between the surface and the core area. The tempera-
ture differences of both the oil quenching and the gas pressure cooling were calculated via
a FEM simulation and incorporated into the continuous CCT diagram, Fig. 6.47a, b. For

Fig. 6.46 Documentation of the external and crack flank surface as well as the microstructure for-
mation in the base material (a) Intergranular oxidation on the external surface; (b) Intergranular
oxidation-free surface of the crack flank; (c) Overview of the microstructure formation in the base
material; (d) Structureless, low tempered martensite and carbide distribution
6.1 Process Parameters 205

1200

1100 X90CrMoV18, 1.4112, austenitization 1050 °C 15 min, quenching in oil

1000

900
Ac1e (855 °C)
800
Temperature [°C]

Ac1b (810 °C)

Ferrite + carbide
700 100
90
Austenite + carbide Carbide 10 15
600
1
500

400
Tmax: 580 °C
300

200
Ms (150 °C) 554 264
100 Martensite 679 649 580 456 199
0
0.1 1 10 100 1000 10000 1E5 1E6
Time [sec]

1200

1100 X90CrMoV18, 1.4112, austenitization 1050 °C 15 min, gas pressure cooling nitrogen 6 bar

1000

900
Ac1e (855 °C)
800
Temperature [°C]

Ac1b (810 °C)

Ferrite + carbide
700 100
90
Austenite + carbide Carbide 10 15
600

500

400
Tmax : 160 °C
300

200
Ms (150 °C) 554 264
100 Martensite 679 649 580 456 199
0
0.1 1 10 100 1000 10000 1E5 1E6
Time [sec]

Fig. 6.47 Calculation of the max. temperature differences during quenching in oil and during gas
pressure cooling with 6 bar nitrogen
206 6 Heat Treatment Influence Area

the oil quenching, a max. temperature difference of 580 °C was calculated and for the gas
pressure cooling, the max. temperature difference was 160 °C.
In the previously performed vacuum hardening, no cracks had occurred in similar
measuring mandrels with identical groove contour. From this it can be concluded that
the lower temperature difference in the gas pressure cooling did not reach a critical stress
value. The oil hardening carried out for the first time, on the other hand, achieved such
large quenching stresses that cracking occurred.
As a remedial measure, oil quenching was now prohibited on the drawing. A con-
structive and defusing change in the groove design, on the other hand, was rejected.

Example 2—Crack Formation on Rods

Component: Rods with square tips

Material: 36SMnPb14, Material No. 1.0765
Error pattern: Crack formation after Inductive hardening
Heat treatment: Inductive hardening with water quenching + 160 °C 1 h

Due to sulfur contents of 0.10–0.18% with correspondingly increased manganese con-

tent, the free-cutting steel 36SMnPb14 has very good machinability due to a high pro-
portion of manganese sulfides. The steel was obtained as rod material with a dimension
of ∅ 25 mm, and rods with square tips were mechanically manufactured from this. The
chemical composition is compared to the target analysis in Table 6.6.
To increase wear resistance, the square area was inductively hardened and tempered
at 160 °C. A surface hardness of 495 HV 10 was achieved. In the area of the inductively
hardened square, crack formation was discovered on several rods after the heat treatment
was completed. The crack formation could be made visible with a magnetic powder test.
There were three cracks originating from the side surfaces, which met in the center of the
square, Fig. 6.48.
For the microstructure evaluation, a longitudinal and cross section were prepared and
examined metallographically. In the longitudinal section, the crack depth perpendicu-
lar to the square surface could be determined as 1.4 mm. The manganese sulfides were
finely distributed and, according to the Steel-Iron Test Sheet 1572:1971, the manganese
sulfide distribution was rated as 1.2, Fig. 6.49a, b.

Table 6.6  Target and actual analysis of the automatic steel 36SMnPb14

Target analysis C Si Mn P S Cr Ni Pb
36SMnPb14, 1.0765 % % % % % % % %
acc. to DIN EN 0.32 max. 1.30 max. 0.100 – – 0.15
10087
0.39 0.40 1.70 0.060 0.180 – – 0.35
Measurement results 0.33 0.11 1.30 0.028 0.179 0.15 0.11 0.19
All data in mass-%
6.1 Process Parameters 207

Fig. 6.48 Crack formation at the square tip

Fig. 6.49 Crack and microstructure documentation in the area of the damaged square (a) Crack
depth perpendicular to the surface of the square tip; (b) Uniform distribution of the manganese
sulfides; (c) Coarser martensite at the front side and crack initiation; (d) Finer martensite in the
further course of the crack
208 6 Heat Treatment Influence Area

The microstructure formation in the area of crack formation was exemplarily docu-
mented in the cross section, Fig. 6.49c, d. There was a martensitic hardening microstruc-
ture with an intercrystalline crack. Due to the cross orientation, the manganese sulfides
were cut in a point-like manner. No oxidation products and no microstructural changes
were present on the crack flanks. However, it was noted that the martensite needle length
at the front side of the square was somewhat coarser than in the further course of the
crack. This led to the conclusion that a higher austenitization temperature or a different
temperature profile was present at the front side.
Since this free-cutting steel has a very limited hardenability, water was used as a
quenching medium. The combination of the higher temperature and the abrupt quench-
ing medium led to quenching stress cracks.
It should also be noted that the lead in this automatic steel melts at 336 °C and then
solidifies again. Whether this was another influencing factor was not discernible, but it
cannot be fundamentally ruled out.

Example 3—Crack Formation on Stepped Bodies

Component: Stepped bodyfor testing purposes

Material: 30CrNiMo8, Material No. 1.6580
Error pattern: Crack formation after nitrocarburizing and oxidizing
Heat treatment: Salt bath nitrocarburizing and salt bath oxidation

To verify the expected results of nitrocarburizing of axles with different diameters,

stepped bodies with six different diameters from ∅ 100 mm to ∅ 25 mm were made from
the material 30CrNiMo8, 1.6580, for testing purposes. After sufficient preheating in a
circulating air furnace, the nitrocarburizing and oxidizing were carried out in salt baths.
The stepped bodies were quenched in water from the oxidation salt bath.
During a routine magnetic particle inspection, cracks were found in the radii of the
transition to the next dimension stage. Therefore, a metallographic examination was ini-
tiated.
The cutting position for the metallographic examination is marked in Fig. 6.50. The
cracks were only visible in the unetched state and at high magnification, Fig. 6.51a, b.
This was due to the high residual compressive stresses that exist after nitriding treat-
ments. To cause crack formation, tensile stresses must have been present at the time
of crack initiation. This is also to be assumed for the water quenching carried out after
oxidation. The surface cools down much faster than the core, leading to tensile stresses
at the surface. This effect was further amplified by the very different diameters of the
stepped body.
There was a uniformly formed diffusion layer, Fig. 6.51c, and an insufficient com-
pound layer thickness of 4 μm, Fig. 6.51d. This can be corrected by extending the hold-
ing time in the salt bath. On the other hand, crack formation can only be achieved by
reducing the cooling speed, thereby minimizing thermal stresses.
6.1 Process Parameters 209

Cut joint position

Fig. 6.50 Documentation of the nitrocarburized and oxidized stepped bodies

Example 4—Crack Formation on a Pulley

Component: Pulley
Material: S355J2G3, Material No. 1.0570
Error pattern: Crack formation during induction hardening
Heat treatment: Inductive surface hardening

Crack formation was discovered on an inductively hardened pulley during a dye pen-
etrant inspection. To investigate the cause of the crack, a segment of the pulley was
removed, Fig. 6.52. The cracked running surface was cut off and subjected to another
dye penetrant inspection, Fig. 6.53.
The verification of the chemical composition confirmed the target ranges of the speci-
fied material S355J2G3, 1.0570. The aluminum content was slightly below the minimum
content, but still within the permissible range of a piece analysis, Table 6.7.
Metallographic microsections were taken across the cracks. The macroscopic repre-
sentation of a crack, Fig. 6.54a, already showed that in the hardened surface layer, the
microstructural linearity was no longer present only at the immediate surface. Predomi-
nantly, the linear microstructure of the original structure was still present.
210 6 Heat Treatment Influence Area

Fig. 6.51 Cracks in the radius area and formation of the nitriding layer (a) Crack depth 31 μm;
(b) Crack depth 42 μm; (c) Bonding layer and diffusion zone; (d) Bonding layer with 4 μm

Fig. 6.52 Pulley segment with marking of the crack position

6.1 Process Parameters 211

Fig. 6.53 Documentation of the crack and the dye penetrant inspection

Table 6.7  Target and actual analysis of the steel S355J2G3

Target C Si Mn P S Cr Ni Mo Al Cu
analysis % % % % % % % % % %
S355J2G3, max. max. max. max. max. max. max. max. min. –
1.0570
acc. to DIN 0.22 0.55 1.60 0.035 0.035 0.30 0.30 0.08 0.020 –
EN
10250-2
Permissible ± 0.02 + 0.04 ± 0.06 + 0.005 + 0.005 + 0.05 + 0.05 + 0.03 -0.005 –
deviation
acc. to DIN
EN 10250-2
Measure- 0.15 0.26 1.24 0.006 0.007 0.17 0.03 0.002 0.018 0.04
ment results
All values in mass-%

The intergranular crack followed the high hardness microstructural lines, Fig. 6.54b,
d. The linear original microstructure in Fig. 6.54c was typical for this unalloyed con-
struction steel.
The material S355J2G3 is not a typical steel for inductive surface hardening. The low
carbon content results in a high Ac3temperature and thus a high hardening temperature.
Accordingly, the massive martensite was coarsened, Fig. 6.54e, f.
There were quenching stress cracks present, which were caused by the simultaneous
interaction of several factors.

1. The low carbon content required a high austenitizing temperature, resulting in a cer-
tain grain coarsening with a reduction in toughness.
2. Due to the low carbon content, the ferrite lines were wider than the pearlite lines. The
carbon had to diffuse over greater distances after the formation of austenite, which
made the formation of a homogeneous austenite more difficult.
212 6 Heat Treatment Influence Area

Fig. 6.54 Microstructure documentation at various positions (a) Crack depth 3.2 mm; (b) Inter-
granular crack progression; (c) Linear ferrite-pearlite structure; (d) Crack progression in the hard
structure; (e) Massive martensite in the hardening layer; (f) Coarsened massive martensite

3. The unalloyed construction steel had a low hardenability, which is why it had to be
quenched abruptly.
4. The higher hardening temperature also caused a larger temperature difference
between the component to be quenched and the quenching medium, and thus larger
thermal stresses.
6.1 Process Parameters 213

Example 5—Premature Failure of a Spherical Roller Bearing

Component: Spherical roller bearing ∅ 560 mm outer diameter

Material: 100Cr6, Material No. 1.3505
Error image: Premature failure
Heat treatment: Hardening and tempering, process parameters were not known

A spherical roller bearing was designed for the drive of a schnitzel press with a lifespan
of 30,000 hours. After only 4,500 operating hours, the bearing was completely destroyed
due to bearing shell peeling of the bearing inner ring, Fig. 6.55. The degree of damage
was much more pronounced on one running surface. For further investigations, three cut
positions were taken, Fig. 6.56.
From the microstructure images, it was evident that both raceway surfaces did not
have a completely martensitic microstructure. There were mixed microstructures with
high proportions of pearlite and bainite. The central web, on the other hand, showed a
higher martensite content and lower bainite, Fig. 6.57, 6.58and 6.59.
There was intense decarburization at the undercut. Although this area was not relevant
to the load, it indicated an austenitization without protective gas , Fig. 6.60.
Hardness profiles were created from both running surfaces, starting from the raceway
surface, Figs. 6.61and 6.62. The hardness directly below the running surface was com-
pletely insufficient on both sides, which was to be expected due to the mixed microstruc-
tures shown. The minimum hardness for rolling bearing components of 58 HRC was
significantly undercut with 46-52 HRC (estimate by conversion).
From the microstructure, it could also be inferred that there must have been insuf-
ficient cooling speed during hardening. The mechanical processing of rolling bearing
steels is carried out in the +AC state. However, remnants of the soft annealing micro-
structure were no longer present, so it can be assumed that the austenitization tempera-
ture was sufficiently high. Overheating of the bearing inner ring was also ruled out as a

Fig. 6.55 Different degrees of damage on the two running surfaces

214 6 Heat Treatment Influence Area

Fig. 6.56 Marking of the cut positions

Fig. 6.57 Microstructure formation on the more severely damaged running surface

Fig. 6.58 Microstructure formation on the central web

6.1 Process Parameters 215

Fig. 6.59 Microstructure formation on the more severely damaged running surface

Fig. 6.60 Microstructure formation at the undercut

cause of the error, as no friction martensite and no high-tempered martensite were vis-
ible. Instead, significant amounts of pearlite and martensite were shown. Only in the
microstructure of the central web was the martensite content higher, but there was also
significant bainite formation here.
Since it was a purchased part, the manufacturer of the drive for the schnitzel press
could not provide any information on the heat treatment of the bearing inner ring.

Example 6—Insufficient Hardness on Guide Rails

Component: Guide railcold formed from strip steel

Material: C60E +Cr, Material No. 1.1221
Error pattern: Target hardness not achieved
Heat treatment: Bainitizing(isothermal transformation in the bainite stage)
216 6 Heat Treatment Influence Area

800

700

600

500
Hardness [HV 1]

400

300

200

100

0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0
Edge distance [mm]

Fig. 6.61 Hardness profile on the heavily damaged running surface

800

700

600

500
Hardness [HV 1]

400

300

200

100

0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0
Edge distance [mm]

Fig. 6.62 Hardness profile on the less damaged running surface

6.1 Process Parameters 217

By isothermally transforming in the lower bainite stage (bainitizing), components

achieve both high hardness and toughness. At the same hardness, the toughness is higher
compared to tempered martensite [27]. This favorable combination of properties has led
to numerous component applications. In particular, components that are subjected to
shock-like loads gain higher fracture safety through bainitizing.
However, several boundary conditions must be met for bainitizing. The cooling from
the austenitizing temperature to the isothermal transformation temperature must be so
rapid that a perlite transformation is avoided. With the carbon steel C60E used here, this
is only possible with thin dimensions and a small amount of chromium was added to
improve hardenability.
When quenching, it is also necessary to avoid falling below the martensite start tem-
perature, as even partial martensite formation results in the loss of higher toughness.
After bainitizing, no tempering takes place. The highest hardness values are achieved at
transformation temperatures just above Ms, which emphasizes the special temperature
control.
The aforementioned requirements for quenching can practically only be realized with
thin-walled components in salt baths. The necessary temperature profile is schematically
drawn into an isothermal time-temperature diagram, Fig. 6.63.
Thin-walled guide rails made of the steel C60E +Cr were isothermally transformed
in an external hardening shop after cold forming. The target hardness was specified as

AC3

AC1
Core

Edge
Temperature [°C]

End
Start
1% 99 %

t 99 % Conversion time
Holding time (log t) [min]

Fig. 6.63 Surface and core cooling curves and isothermal transformation (schematic)
218 6 Heat Treatment Influence Area

40 +5 HRC. The hardness measurement in the hardening shop on a few parts resulted in
hardness values of 43.5 to 45 HRC, thus meeting the target specification.
After the guide rails were returned, a random check of the hardness was carried out.
A guide rail with a hardness of only 34 HRC was found. As a result, a larger quantity
was tested and numerous components with insufficient hardness were discovered and the
entire treatment batch was blocked.
Since there were no changes in temperature and throughput during the entire heat
treatment, the hardening shop initiated an investigation to find the cause and shared the
process data.
The components were austenitized at 890 °C for 10 min. in a conveyor belt furnace.
The temperature was confirmed by drag element measurements. The quenching took
place in a salt bath at 330 °C and the holding time for the isothermal transformation was
also 10 min.
In the first step of the investigation, the chemical composition was checked, Table 6.8.
From the results, it could be inferred that both the i.-O. component and the n.-i.-O. com-
ponent originated from the same melt batch. Furthermore, it was noticed that the chro-
mium content was below the lower target value limit. The also hardenability-increasing
alloy element manganese was within the target value range, but at the lower limit.
The hardenability for this batch composition was calculated according to the formulas
from Stahl-Eisen-Test Sheet 1664:2004-06 [28] and compared to the scatter band from
DIN EN ISO 683-1:2018-09 [29], Fig. 6.64. Accordingly, the hardenability was at the
lowest permissible limit. A positive influence of the chromium analysis supplement was
not recognizable.
The microstructure examination of one i.-O.-part and one n.-i.-O.-part led to further
insights. The i.-O.-component was within the target value range in terms of hardness,
but the microstructure showed about 5% perlite in addition to bainite, Fig. 6.65a. Undis-
solved carbides were not recognizable in this microstructure. The n.-i.-O.-component
had a completely perlite microstructure and there were still numerous undissolved car-
bides present, Fig. 6.65b.

Table 6.8  Target and actual analyses of an i.-O. and an n.-i.-O. component

Target analysis C Si Mn P S Cr Ni Mo Al
C60E + Cr % % % % % % % % %
0.57 0.15 0.35 max. max. 0.20 0.03 max. 0.008
0.63 0.30 0.70 0.025 0.004 0.40 0.40 0.10 0.04
i.O. component 0.57 0.17 0.39 0.012 0.003 0.13 0.06 0.01 0.016
n.i.O. component 0.56 0.017 0.40 0.011 0.003 0.13 0.05 0.01 0.015
All data in mass-%
6.1 Process Parameters 219

Material C60E 1.1221 | Austenitizing temperature 830 °C

70

65

60

55

50
Hardness [HRC]

45

40

35

30

25

20
0 5 10 15 20 25 30 35 40 45 50
Face distance [mm]

Fig. 6.64 Hardenability calculated according to SEP 1664 [28] and target value range from DIN
EN ISO 683-1:2018-09 [29]

Fig. 6.65 Microstructure documentation of the i.O.-component (a) and the n.i.O.-component (b)

Several causes were responsible for this error pattern:

1. A material with too low hardenability was used for this application. The supplemen-
tary delivery specification of 0.20–0.40% Cr was not met.
2. The heating in the conveyor furnace was uneven. All carbides were dissolved in the
i.-O.-component, while numerous undissolved carbides were still present in the n.-i.-
O.-component. These parts did not have homogeneous austenite.
220 6 Heat Treatment Influence Area

3. The process was not sufficiently validated. With targeted sampling, the too low hard-
ness at unfavorably positioned locations would have been noticed and defective parts
would not have been delivered.

Nitrate-containingquenching salt baths can be accelerated in their quenching effect by a

smalladdition of water, which can bring additional process reliability tobainitizing.

Example 7—Strongly Varying Hardnesses on Rivets

Component: Rivets
Material: X6NiCrTiMoVB25-15-2, Material No. 1.4980
Error pattern: Strongly varying hardness
Heat treatment: Solution annealing

The described cases from the area of martensitic and bainitic hardening are expanded in
this example to the area of precipitation hardening. The steel X6NiCrTiMoVB25-15-2,
1.4980, used for rivets, not only has good corrosion resistance, but also high heat resist-
ance.
Precipitation hardening steels must first be solution annealed. For this steel, a solu-
tion annealing temperature between 900 and 980 °C is required. The solution annealed
state achieved at this temperature must be quenched without precipitation. At room tem-
perature, an oversaturated austenitic microstructure is then present. The hardening is
achieved by a long-term annealing at 720 ± 14 °C.
After proper precipitation hardening, this steel achieves a typical hardness of 35-38
HRC. The rivets delivered for examination had not reached this hardness, which is why
an investigation should be carried out. The solution annealing of a 300 kg bulk charge
was carried out in a vacuum furnace with nitrogen gas pressure cooling.
The hardness check confirmed the insufficient hardness values with hardness values of
20 to 28 HRC. The material composition fully complied with the standard specification.
The metallographic examination was first carried out on the surface of a heat-treated
rivet, Fig. 6.66a. On the surface, there was a darkly etching seam extending approxi-
mately 25 μm. The microstructure on the surface before heat treatment did not show this
seam, Fig. 6.66b.
To identify this seam, EDX measurements were carried out and the results showed
that the darkly etching area was nitrided and had nitrogen contents of up to about 5%,
Fig. 6.67.
Both the insufficient hardness and the nitrided surface pointed to the same heat treat-
ment error. Nitriding during gas pressure cooling occurs when molecular nitrogen disso-
ciates sufficiently at high temperatures. Atomic nitrogen is then absorbed by the surface.
The nitriding was only possible due to an insufficient cooling rate in the 300 kg bulk
charge. The rivets inside were cooled very slowly, but were immediately contacted with
6.1 Process Parameters 221

Fig. 6.66 Microstructures before and after heat treatment (a) Microstructure on the surface after
heat treatment; (b) Microstructure on the surface before heat treatment

Nitrogen contents of 3.3 % to 4.7 %

Spectrum 21 Spectrum 22
Spectrum 23
+ Spectrum 24
+
+
Spectrum 26
+

Spectrum 27 Spectrum 28

Nitrogen content < 1 %

Fig. 6.67 EDX measurements in the darkly etching area

nitrogen after the start of cooling. At the same time, the slow cooling initially caused
precipitation with coarse particle size. The final hardening was thus ineffective.
Solution annealing in a vacuum furnace with nitrogen gas pressure cooling is com-
pletely unsuitable, especially for the high batch weight. Quenching should be done in a
liquid quenching medium. Only then is the solution annealed state preserved and harden-
ing can be successfully carried out.
222 6 Heat Treatment Influence Area

6.1.4 Furnace Atmospheres

Various methods are used in heat treatment to protect against oxidation and decarburiza-
tion. The systems and media used for this can both provide pure protection and enable
targeted changes in the chemical composition in the surface layer.
Even in vacuum heat treatment, protective gases are used for the majority of the
process time. Only before the start of the process is a vacuum created in the cold state,
then convection-supported heating usually takes place under nitrogen. At a temperature
of 800–850 °C, this nitrogen is pumped out again and heating to higher temperatures
takes place under vacuum. During quenching, a gas with high pressure is used again. The
quenching gas can be nitrogen or helium, in a few cases argon is also used. Hydrogen
would be very suitable as a quenching gas, but the use of this gas requires a very high
level of safety equipment.
The pumping out of nitrogen at temperatures of 800–850 °C is necessary because at
higher temperatures a dissociation of molecular nitrogen to atomic nitrogen begins. This
would lead to unwanted nitriding.
However, vacuum furnaces are also used today for thermochemical heat treatments.
Corrosion-resistant steels can be specifically nitrided at high temperatures and vacuum
carburizing using acetylene with subsequent gas pressure cooling is already widely used.
Another possibility to protect components from oxidation and possibly also to treat
them thermochemically is given by the salt bath methods. Although this heat treatment
technique has lost its overall importance, there are still numerous systems for neutral
hardening, case hardening and nitrocarburizing in use.
The use of granules for carburizing is now only carried out in individual cases and on
a laboratory scale. The grain size stability test described in DIN EN ISO 683-3 [30] for
example, describes carburizing according to McQuaid-Ehn in a carburizing powder.
Plasma-assisted processes occupy a certain special position. The gases used in plasma
nitriding or plasma nitrocarburizing become reactive gases on the component surface
only through the formation of plasma.
The heat treatment in air should not go unmentioned. The oxygen in the air causes
oxidation or scale formation, and from temperatures of around 600 °C, decarburization
also occurs. These two reactions are known and expected. Less well known are changes
in the surface composition due to oxidation-stable accompanying elements. The two ele-
ments copper and nickel are oxidation-stable, which is why the selective oxidation of
iron leads to an enrichment of copper and nickel contents on the surface with subsequent
grain boundary diffusion. This undesirable surface reaction occurs even when the con-
tents of copper and nickel do not exceed the limit values for unintentional trace elements.
Steels from electric steel production with 100% scrap input are generally more sus-
ceptible, as these two elements cannot be filtered out during steel production. Steels from
the blast furnace route and subsequent converter production are usually an order of mag-
nitude lower in copper and nickel content. Of course, this does not include the deliber-
ately nickel-alloyed steels.
6.1 Process Parameters 223

For a description of the diverse process sequences of various heat treatments, refer-
ence is made to the extensive literature, for example [1–5].
The potential errors in thermochemical processes will be discussed in a separate chapter.

6.1.4.1 Damage Examples—Atmosphere

Example 1—Excessive Elasticity in Fabrics

Component: Fabricwith 100 μm wire diameter

Material: X2CrNi19-11, Material No. 1.4306
Error Image: Excessive Elasticityafter recrystallization annealing
Heat Treatment: Recrystallization Annealing in Vacuum Furnace

When weaving metal wires from the austenitic material X2CrNi19-11, 1.4306, there is a
certain cold hardening, which should be solved by a recrystallization annealing at 1050
°C. The recrystallization temperature is identical to the solution annealing temperature
for austenitic steels. Low recrystallization temperatures are not applicable, as this would
lead to the formation of chromium carbides, which would result in a significant decrease
in corrosion resistance. Grain decay is even to be feared.
The fabrics were wound up and the wire rolls had a diameter of about 250 mm. To
prevent chromium carbide formation during cooling, a quenching treatment with 5 bar
nitrogen was required.
After completion of the recrystallization annealing, a technological process evalua-
tion was carried out using narrow fabric strips, which were manually bent over a mandrel
with a defined diameter. One side of the fabric strip was fixed, the other was released
again after bending. Fabric strips from the outer windings led to the expected small
springback angles, as a slight plastic deformation also occurred during bending. How-
ever, if fabric strips from the middle of the roll were tested, a significantly larger spring-
back angle resulted, indicating increased elasticity.
The metallographic examination of the conspicuous fabric strips revealed the cause.
The fabric showed a lamellar microstructure and a check of the nitrogen content revealed
a value of 1.65%, while the carbon content was 0.02%, Fig. 6.68. These were therefore
chromium nitrides.
The inner windings cooled down significantly slower, but were in contact with nitro-
gen from the beginning of the cooling. Since the molecular nitrogen dissociates to a cer-
tain extent at the high annealing temperatures, nitriding occurred. The outer windings, on
the other hand, cooled down very quickly to a temperature that no longer led to nitrogen
dissociation and the associated nitriding.
Avoiding this unwanted nitriding can only be achieved by using argon or helium as
quenching gas.
224 6 Heat Treatment Influence Area

Fig. 6.68 Chromium nitrides in pearlitic formation

Example 2—Poor Machinability of Sheets

Component: Sheetswith 2.5 mm sheet thickness

Material: X2CrNiMoN17-13-3, Material No.: 1.4429
Error pattern: Poor machinability in the center of the sheets
Heat treatment: Recrystallization annealing in a vacuum furnace

A similar problem as in the first example also arose after a vacuum annealing treat-
ment with nitrogen gas quenching. Sheets made of the corrosion-resistant steel X2CrNi-
MoN17-13-3, 1.4429, were to be recrystallization annealed after cold forming. Due to
the thin sheet thickness, packages of 50 sheets each were assembled and fixed with wire,
Fig. 6.69. Here too, rapid cooling was required, which is why nitrogen gas quenching
was carried out at 5 bar.
After the recrystallization annealing was completed, some sheets in the area in the
center were very difficult to machine. Other sheets from the same treatment batch, on the
other hand, could be machined without any problems.

Fig. 6.69 Schematic representation of the charging of thin sheets

6.1 Process Parameters 225

In parallel to this vacuum treatment, recrystallization annealing was carried out on

identical sheets in a second company. These sheets could be machined without any prob-
lems.
The examination of both the difficult and easy to machine sheet surfaces was carried
out using GDOES, Fig. 6.70. In the case of nitrogen gas quenching, there was also a sig-
nificant nitriding with nitride formation, which was associated with an increase in hard-
ness. The local limitation resulted from the slow cooling in the core area of the densely
packed sheet package. The gaps between the sheets were sufficiently large for contact
with nitrogen, the surface temperature due to the slow cooling speed was still so high
that nitrogen dissociation with subsequent nitriding occurred.
In the second heat treatment company, the cooling was carried out under argon and
thus there was no change in the surface composition.
Targeted investigations on samples of the hot work steel X40CrMoV5-1 showed a
clear dependence of the nitrogen uptake on the cooling speed in GDOES measurements,
Fig. 6.71.
As a technological experiment, hardness testing on the surface with different test
loads is also possible. Since the applied surface causes an increase in hardness, a higher
hardness is measured with a small test load of HV 1 than with the larger test load HV 10.
In the fastest cooling variant with λ = 0.52, there was no application and no difference
in hardness was measured between the different test loads, Fig. 6.72.

1.2

1.0
Unannealed sheet metal
Annealing under argon
Annealing under nitrogen
Nitrogen concentration [%]

0.8

0.6

0.4

0.2

0
0 5 10 15 20 25 30 35 40 45 50 55
Removal depth [µm]

Fig. 6.70 Results of the GDOES measurements

226 6 Heat Treatment Influence Area

1.00

0.80
= 6.60
= 3.30
= 0.52
Nitrogen content [%]

0.60

0.40

0.20

0
0.0 5.0 10.0 15.0 20.0 25.0 30.0
Distance from the surface [µm]

Fig. 6.71 Influence of the cooling speed on the nitrogen uptake

660

640
HV 1
HV 3
HV 10
620
Vickers hardness

600

580

560

540
= 6.60 = 3.30 = 0.52

Fig. 6.72 Hardness tests with different test loads to prove an application
6.1 Process Parameters 227

Example 3—Corrosion Formation on Bipolar Forceps

Component: Bipolar forceps

Material: X20Cr13, Material No. 1.4021
Error pattern: Corrosion formation after a few cleaning cycles
Heat treatment: 1050 °C → Cooling under Ammonia cracking gas

The corrosion-resistant steel X20Cr13 can be hardened in a vacuum furnace due to its
sufficient hardenability, but this requires appropriate throughput quantities for economic
reasons. If this is not the case, smaller belt furnaces are also used, which are operated
with a protective gas from split ammonia. The liquid ammonia is split into hydrogen and
nitrogen at high temperatures and a catalyst in an ammonia splitter. The ideal decay reac-
tion is:
2NH3 → N2 + 3H2
If the ammonia splitter is not operated optimally or if unrecognized malfunctions occur,
unsplit ammonia also enters the furnace chamber, which leads to an application. The
material X20Cr13 has good corrosion resistance in the hardened state, as the minimum
chromium content of 10.5% is available for the formation of a passive layer. However,
the absorption of nitrogen forms chromium nitrides, which bind a large part of the avail-
able chromium. The minimum content is then undercut in the martensitic base matrix
and sufficient corrosion resistance is no longer given.
The described error occurred with bipolar forceps that were hardened in a protective
gas furnace operated with ammonia cracking gas. After a few sterilizations, the forceps
tip was completely corroded, Fig. 6.73.

Fig. 6.73 Corroded forceps tips

228 6 Heat Treatment Influence Area

A metallographic examination proved that a nitrogen pearlite seam of approx. 11.2

μm had occurred on the surface, Fig. 6.74. As a result, large parts of the chromium were
bound to chromium nitrides, which led to the loss of corrosion resistance.

Example 4—Hole Breakthrough in a Heating Muffle

Component: 9340 mm long heating muffle of a sintering plant

Material: GX40NiCrSi38-19+Nb, Material No. 1.4865 +Nb
Error Image: Hole breakthrough of the 15 mm thick wall
Heat Treatment: Sintering at 1100 °C under nitrogen/hydrogen protective gas

This case of damage does not concern the heat-treated material, but the heating muffle
itself. In a sintering plant, there was a hole breakthrough on the underside of the 9340
mm long heating muffle and tearing on the upper side of the 15 mm thick wall, Fig. 6.75.
The muffle was made from the heat-resistant cast material GX40NiCrSi38-19 +Nb,
1.4865+Nb. The sintering plant was initially operated at temperatures below 1000 °C
with a protective gas of 90% nitrogen and 10% hydrogen. After an increase in the sinter-
ing temperature to 1100 °C, there was a significant reduction in service life and the wall
breakthrough.
In the micrograph at the breakthrough position of the muffle underside, a large num-
ber of precipitations were visible, Fig. 6.76. A similar picture emerged in the area of the
tear or crack flank on the muffle top, Fig. 6.77.
In the EDX examination of the various precipitations, niobium carbides, chromium
carbides, and chromium nitrides were detected. The spectrum of a chromium nitride is
documented in Fig. 6.78. From the atomic % ratio of 29% nitrogen and 56% chromium,

Fig. 6.74 Seam of nitrogen pearlite on the surface

6.1 Process Parameters 229

Fig. 6.75 Damage spots on the underside and top of the muffle (a) Breakthrough on the underside
of the muffle; (b) Tearing on the top of the muffle

Fig. 6.76 Microsection position at the breakthrough site of the muffle underside (a) Overview of
the breakthrough site; (b) Large number of precipitations

Fig. 6.77 Microsection position at the tear of the muffle top (a) Overview of the muffle inside;
(b) Precipitations at the crack flank
230 6 Heat Treatment Influence Area

Spectrum 1

Spectrum 1

Electron image Scale range 542 cts Cursor: 0.000

Fig. 6.78 EDX spectrum of a chromium nitride (a) Marking of the EDX measurement position;
(b) Chromium nitride

it was evident that Cr2N precipitations were formed, which lead to precipitation harden-
ing in austenitic steels [31].
The sintering temperature of 1100 °C and the nitrogen content of 90% led to a par-
tial dissociation of molecular nitrogen. In atomic form, the surface absorbs the nitrogen,
forming chromium nitrides Cr2N. This process removes a large amount of chromium
from the matrix, leading to a reduction or even loss of heat resistance. As a result, there
was a removal of the surface and finally a wall breakthrough.

6.2 Thermochemical Heat Treatments

While thermal processes do not intend to change the surface composition, this is the
objective in thermochemical processes. Carbon, nitrogen, and boron are among the ele-
ments most frequently transferred to the steel surface. Carburizing, nitriding, nitrocarbur-
izing, and boronizing are the corresponding thermochemical heat treatment processes.
The mentioned elements diffuse into the surface in these processes, unlike coating pro-
cesses where layers grow.
In thermochemical heat treatment processes, unavoidable side reactions also occur,
which have a negative impact on the component properties [32]. Therefore, the elements
oxygen and hydrogen must be included in the considerations.
The case hardeningconsists of the individual steps of heating, carburizing, quench-
ing, and tempering. In salt bath case hardening, carbonitriding always occurs because the
carbon donor is cyanide (CN). Carbonitriding in endothermic protective gas atmospheres
requires a targeted addition of ammonia. In the process-determining water gas reaction,
in addition to the transfer of carbon, oxygen and hydrogen are also absorbed. The result-
ing errors and damages are described in detail in the damage examples.
6.2 Thermochemical Heat Treatments 231

The carburizingin vacuum furnaces is also one of the established thermochemical heat
treatment processes today. While earlier attempts with propane as a carburizing gas failed,
the use of acetylene as a carbon donor was the breakthrough for this process. The process
technology differs significantly from the protective gas processes, as atmosphere control via
oxygen probes is not possible [33, 34]. Although the absence of oxygen prevents intergran-
ular oxidation during carburizing in vacuum plants, manganese effusion can occur [35].
The classic gas nitridingis carried out in ammonia/nitrogen gas atmospheres. For
nitrocarburizing, a carbon-releasing gas must be added. In the salt bath process, only
nitrocarburizing is possible again, as carbon and nitrogen are transferred simultane-
ously. In plasma nitriding and nitrocarburizing processes, the plasma causes a splitting
of molecular nitrogen, which is why molecular nitrogen can be used for plasma nitrid-
ing. For nitrocarburizing processes, a carbon-releasing gas must also be added here. In
addition to the process-typical error possibilities, a frequently encountered error in the
pre-treatment occurs in the gas and plasma processes. Passive layers on the component
surface hinder or completely prevent the absorption of nitrogen, resulting in soft surface
areas. Examples of this type of error are therefore described in the chapter “Manufactur-
ing before heat treatment”.
A certain special position among the thermochemical processes is taken by boroniz-
ing. The treatment is usually carried out at temperatures in the austenite range using
boron-releasing substances. The best-known process is powder boronizing. This forms a
very hard and wear-resistant iron boride layer.
Detailed descriptions of all thermochemical heat treatment processes are contained in
[1–5].

6.2.1 Damage Examples—Case Hardening

Example 1—Soft Spots on a Drive Wheel

Component: Drive wheelwith a pitch circle diameter of 300 mm, module 4

Material: 18CrNiMo7-6, Material No. 1.6587
Error pattern: Soft spots after case hardening and tempering
Heat treatment: Direct hardening, process parameters were not communicated

The completely insufficient and highly fluctuating surface hardness of 39–52 HRC
prompted the heat treater of this drive wheel to conduct an independent investigation, as
he suspected insufficient hardenability or steel mix-up. After the obligatory initial docu-
mentation, the investigation began with the determination of the chemical composition.
From the analysis results, Table 6.9, the hardenability was calculated according to the for-
mulas from SEP 1664 [28] and graphically compared to the standard deviation, Fig. 6.79.
232 6 Heat Treatment Influence Area

Table 6.9  Target and actual analysis of the case-hardening steel 18CrNoMo7-6

Target analysis C Si Mn P S Cr Ni Mo Al N
18CrNiMo7-6, % % % % % % % % % %
1.6587 0.15 max. 0.50 max. max. 1.50 1.40 0.25 – –
acc. to DIN EN
10084:2008-06 0.21 0.40 0.90 0.025 0.035 1.80 1.70 0.35 – –

Measurement 0.17 0.27 0.52 0.008 0.027 1.67 1.43 0.31 0.030 0.006
results

Material 18CrNiMo7-6 1.6587

70

65

60

55
Hardness [HRC]

50

45 +HH
+H
40 +HH
+H
35 +HL
+H
+HL
30

25

20
0 5 10 15 20 25 30 35 40 45 50
Face distance [mm]

Fig. 6.79 Analysis results and calculation of hardenability according to SEP 1664 [28]

The calculated hardenability was within the scatter band +HH in the lower third.
Given the small component size for this case-hardening steel, no obvious cause of error
could be derived from the hardenability.
The subsequent metallographic examination revealed a highly carburized hardening
microstructure in the case hardening layer, Fig. 6.80a-c. At the transition from the tooth
head to the tooth flank, there were massive and numerous bone carbides and a very large
amount of residual austenite. The lack of intergranular oxidation on both the tooth head
and the tooth flank indicated a grinding process after case hardening. In the area of the
pitch circle, no bone carbides were present anymore, but the residual austenite content
was still very high. In the non-carburized tooth root area, an inconspicuous tempered
6.2 Thermochemical Heat Treatments 233

Fig. 6.80 Microstructure documentation in the tooth profile (a) microstructure overview at the
tooth head; (b) bone carbides and high residual austenite content; (c) microstructure formation in
the pitch circle; (d) massive martensite in the tooth root

massive martensite was present, Fig. 6.80d. In addition to the metallographic examina-
tions, the residual austenite content was determined to be 74% by X-ray diffraction.
Further investigations were waived based on the findings. There was a strong over-
carburization. The bone carbides were formed when the solubility limit of austenite for
carbon was exceeded. This indicated that a very high C-level was maintained for too
long. It is likely that the high C-level was not reduced at all during the boost phase. The
originally suspected low hardenability by the client was clearly refuted.

Example 2—Edge Chipping on a Running Plate

Component: Running plate ∅ 110 mm, Pin ∅ 20 mm

Material: 20MnCr5, Material No. 1.7147
Error pattern: Edge chipping during grinding after case hardening
Heat treatment: Direct hardening
234 6 Heat Treatment Influence Area

After case hardening in an internal quench furnace, the running plates made of case-
hardening steel 20MnCr5, 1.7147, were damaged by chipping during grinding, Fig. 6.81.
In the microstructure of the case hardening layer, a high residual austenite content and
network-like bone carbides were characteristics of over-carburization, Fig. 6.82. The
fracture line followed the embrittling bone carbides.
The measured surface carbon content of 1.05% confirmed the significant over-carbur-
ization. Bone carbides are formed when the solubility of austenite for carbon is exceeded
during the carburizing phase. This is always a sign of a C-level that has not been reduced
or not sufficiently reduced before reaching the carbide limit.

Example 3—Bubble Formation on a Cold-Formed Cup

Component: Cold-formed cups

Material: C15E2C, Material No. 1.1132
Error pattern: Bubble formation on the surface
Heat treatment: Carburizing, case hardening and tempering

The hardenability of unalloyed case-hardening steels can be increased by adding ammo-

nia during carburizing. This carburizing then leads to a uniform surface hardness without
soft spots, even with oil quenching.
Cold-formed cups made of the unalloyed case-hardening steel C15E2C, 1.1132, were
carburized in a conveyor belt furnace and subsequently quenched in oil. The entire heat
treatment line was linked and some cups were only removed for checking the case hard-
ening depth after tempering. During visual inspection, an anomaly was noticed, which
can be seen in Fig. 6.84. This was not an isolated case, all ten samples submitted for
examination showed similar bubble formation.
A cut was made at the position, as documented in Fig. 6.83, and initially documented
in the unetched and etched state, Fig. 6.84. The entire surface showed a pearl necklace-

Fig. 6.81 Chipping during grinding of a case-hardened running plate

6.2 Thermochemical Heat Treatments 235

Fig. 6.82 Intergranular crack propagation and hardening microstructure with bone carbides and
high residual austenite content

Fig. 6.83 Noticeable bubbles on the inner surface of the cups and documentation of the cut posi-
tion

like cavity formation below the surface and in places there were bulges, which were
macroscopically perceived as bubbles.
This error pattern is typical for a carburizing treatment with too much ammonia. The
splitting of ammonia on the surface not only produces atomic nitrogen, but also atomic
hydrogen. Both elements are absorbed by the surface. The excess amounts of nitrogen
and hydrogen cannot be dissolved in the lattice and therefore accumulate below the sur-
face. The recombination to molecular hydrogen leads to a massive volume increase with
a strong increase in internal pressure, and finally to bubble formation. This is a compara-
ble process to pickling bubble formation.
Usual amounts of ammonia in carburizing are 3–5%. After the investigations were
completed, a defective solenoid valve was discovered on the conveyor system, which had
led to an uncontrolled addition of ammonia.
236 6 Heat Treatment Influence Area

Fig. 6.84 Pearl necklace-like cavity formation under the surface

Example 4—Numerous Ultrasonic Indications on a Spur Gear

Component: Spur GearModule 24, Tip Circle Diameter 1576 mm

Material: 18CrNiMo7-6, Material No. 1.6587
Error Image: Numerous Ultrasonic Indicationsup to KSR 3.9 mm
Heat Treatment: Case Hardening and Tempering
6.2 Thermochemical Heat Treatments 237

After case hardening and tempering or before flank grinding, an ultrasonic test was per-
formed on the spur gear, module 24, made of material 18CrNiMo7-6, 1.6587. A total
of 70 error indications with error sizes of up to KSR = 3.9 mm were detected, which
extended over the entire cross-section. The spur gear had a tip circle diameter of 1576
mm and a total weight of approx. 4 t.
To achieve the required case hardening depth of 5.5 mm, a carburizing duration at 940
°C of 145 h was necessary. During carburizing in endothermic carburizing atmospheres,
there is always a significant hydrogen uptake [36]. Through a timely and sufficiently
long tempering duration, the absorbed hydrogen can effuse again after direct hardening.
Due to the very long carburizing duration, a single hardening with the aim of grain
refinement was carried out on this spur gear. After carburizing, the spur gear was there-
fore not quenched, but cooled to room temperature and after a lay time of 8 working
days, it was reheated and hardened and tempered again. A hardening temperature of 840
°C was chosen for the single hardening. The quenching was carried out in a quenching
oil. Finally, it was tempered at 180 °C for 12 h (Fig. 6.85).
To investigate the UT indications, a segment was removed from the faulty spur gear
on a band saw. Heating during cutting was thus avoided, but the surface hardness of
approx. 60 HRC required the use of a carbide-tipped saw blade.
The initial analysis determination with an optical emission spectrometer showed with-
out exception standard-compliant values for the hardenability class +HH. Also, the total
hydrogen content measured by melt extraction showed no abnormalities with 0.6 ppm
and the case hardening depth also corresponded to the drawing specification.
In the next step, a fracture surface was created on the segment, which already pro-
vided a hint of flake formation, Fig. 6.86.
The flake cracks were visible in the micrograph at numerous positions, Fig. 6.87. The
cracks were predominantly present in the positively segregated areas. The reason for this
lies in the different hydrogen solubility of the austenite and the transformed microstruc-
ture. Austenite can dissolve higher hydrogen contents than all other microstructures that

Fig. 6.85 Spur gear upon delivery for examination

238 6 Heat Treatment Influence Area

Fig. 6.86 Flake formation visible in the fracture surface

Fig. 6.87 Crack formation in the positively segregated areas

form during cooling. Since the austenite transformation begins in negatively segregated
areas, the hydrogen is pushed into the remaining austenite, leading to a significant con-
centration increase in these areas.
In order for this increased concentration not to lead to cracking due to the recombina-
tion of hydrogen, a timely reheating must take place for the single hardening. A lay time
of eight working days was significantly too long and therefore caused the flake formation
due to the recombining hydrogen. The subsequent simple hardening with tempering was
successful in terms of hydrogen effusion, as confirmed by the hydrogen measurements,
however, the cracks were already present.
6.2 Thermochemical Heat Treatments 239

Example 5—Crack Formation on an Unloaded Spur Gear

Component: Spur gearModule 4.5, pitch circle diameter 350 mm

Material: 18CrNiMo7-6, Material No. 1.6587
Error pattern: Crack formation in the unloaded state
Heat treatment: Case hardening and tempering

A total of 80 straight-toothed spur gears made of the case-hardening steel 18CrNiMo7-6,

1.6587, with module 4.5 were sent to an external hardening shop for case hardening.
A case hardening depth of 1.20 +0.30 mm and a surface hardness of 700 +40 HV were
specified. The tolerance specifications were unrealistically narrow, but this did not result
in a cause of damage.
The return of the spur gears took place after the case hardening and tempering treat-
ment before the weekend and the gears were stored in a warehouse. The final production
was planned for the following week. On the first working day of the following week, the
first spur gear was removed from the warehouse. Even with the naked eye, massive crack
formation was recognizable, Fig. 6.88. The inspection of the other spur gears revealed
that all 80 spur gears had cracked over the weekend.
A delayed crack formation after case hardening without external loads points to a
hydrogen-induced brittle fracture. However, it was unclear why there was a 100% failure
in this batch, as identical spur gears had been produced and case hardened without any
problems for years.
The examination of the chemical composition in the base material revealed no devia-
tions from the target specifications of the gear manufacturer's order specification. A mar-
ginal carbon content of 0.90% was measured on the carburized surface.
The damage hypothesis of the hydrogen-induced brittle fracture was clearly con-
firmed by the scanning electron microscopic fracture surface examination. In the
non-carburized base material, there was an intergranular fracture with gaping grain
boundaries, Fig. 6.89.

Fig. 6.88 Spur gear with crack formations

240 6 Heat Treatment Influence Area

Fig. 6.89 Intergranular fracture surface with gaping grain boundaries

The metallographic examination revealed a high residual austenite content and car-
bide formation in the case hardening layer, Fig. 6.90. The non-martensitic microstruc-
ture at the surface is unavoidable during carburization in an endothermic atmosphere. In
the partial circle of the flank, the case hardening depth was determined to be 1.64 mm,
Fig. 6.91. This value was above the target requirement, but the grinding allowance must
also be taken into account. A cause of damage could not be derived from this anyway.
Since the metallographically determined residual austenite content was very high,
a X-ray diffraction residual austenite measurement was carried out. Here, the austen-
ite content of 26% was less surprising, rather a slight shoulder formation was observed
in the 200-reflex of the martensite, Fig. 6.92a. This shoulder formation occurs because
the martensite is tetragonally distorted in the quenched and not tempered state. The test
piece was tempered at 180 °C for 1 hour after the measurement and measured again. The
shoulder formation did not occur again, Fig. 6.92b.
The process data of the case hardening were not provided by the executing heat treat-
ment company. The initially formulated damage hypothesis was confirmed, but the cause
can only be described as an assumption.

Fig. 6.90 Martensite, high residual austenite content, non-martensitic seam at the surface and car-
bides in the case hardening layer
6.2 Thermochemical Heat Treatments 241

800

700
CHD550 1.46 mm
600

500
Hardness [HV 1]

400

300

200

100

0
0.0 0.5 1.0 1.5 2.0
Edge distance [mm]

Fig. 6.91 Determination of the case hardening depth and marking of the required surface hard-
ness

1. The carburization ended with a too high surface carbon content of 0.90%.
2. The too high surface carbon content led to a high residual austenite content and to
carbide formation.
3. The high residual austenite content resulted in a too low surface hardness. At a sur-
face distance of 0.1 mm, the hardness was 592 HV 1, which is 108 HV units below
the target requirement. According to ISO 6336-5 [37], a hardness drop at this test
point of max. 40 HV units is permissible.
4. The shoulder formation in the X-ray diffraction residual austenite measurement indi-
cated a tetragonal distortion of the martensite. After a tempering treatment of 180 °C
for 1 hour, this distortion was no longer present.

Points 1 to 4 were supported by objectivemeasurement results. Based on these facts,

it was assumed that after quenching, theinitial hardness was measured, which was too
low. To not further reduce the surfacehardness, the tempering temperature was reduced.
As a result, the effusion of thehydrogen absorbed during carburization was no longer
given, which ultimately led tothe hydrogen-induced brittle fracture. The tensile stresses
required for this damagemechanism resulted from the residual stresses of the case-hard-
ened spur gears.
242 6 Heat Treatment Influence Area

11000

10000

9000

8000
Shoulder formation on
200-reflex of the Martensit
7000
Lin (Counts)

6000

5000

4000

3000

2000

1000

0
26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
2-theta scale

8000

7000

6000
No more shoulder formation
after tempering 180°C 1h
5000
Lin (Counts)

4000

3000

2000

1000

0
26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
2-theta scale

Fig. 6.92 X-ray diffraction diagrams in the delivery state (top) and after tempering at 180 °C for
1 hour (bottom)
6.2 Thermochemical Heat Treatments 243

Example 6—Unusual Microstructure in a Test Coupon

Component: Pinionwith module 100; tip circle diameter 1040 mm

Material: 18CrNiMo7-6, material no. 1.6587
Error pattern: Unusual microstructure in the co-traveling sample
Heat treatment: Case hardening and tempering

A not quite everyday gear part with the module 100, which was made from the case-
hardening steel 18CrNiMo7-6, 1.6587, was delivered for examination, Fig. 6.93. This
pinion was a sample hardening and before reassignment of a larger series, the target
specifications should be checked and evaluated as part of a detailed initial sample exami-
nation. A microstructure examination had already been carried out on a test coupon made
of the same type of case-hardening steel, and the microstructure was described as unu-
sual, which is why investigations were commissioned on the pinion itself.
For the examination, a tooth was sawn out and a test coupon Ø 60 mm was disassem-
bled for mechanical testing, Fig. 6.94. The microstructure formation in the pitch circle
of the tooth flank was free of bone carbides below the tooth flank surface at a depth of
approx. 125 μm, Fig. 6.95. This was followed by a layer about 400 μm thick with inten-
sive formation of bone carbides. From this it could be concluded that during carburiza-
tion the carbide limit was clearly exceeded and in the subsequent diffusion phase only in
a narrow area the carbides were dissolved again by a correspondingly low C level.

Fig. 6.93 Gear delivered for examination—Module 100

244 6 Heat Treatment Influence Area

Fig. 6.94 Samples taken for examination

Fig. 6.95 Microstructure in the case hardening layer at the pitch circle

The grain boundaries formerly covered with carbides were still recognizable and
formed a “tiger skin” pattern, Fig. 6.96. In the area of intergranular oxidation, there was
a non-martensitic microstructure surface.
In the area of the 30° tangent, due to insufficient cooling speed, there was intense
pearlite formation, Fig. 6.97.
From the examination results, it was evident that the sample hardening showed errors
both in the atmosphere control and during quenching. While it is common to start the
carburizing process with a higher C-level to reduce process times, the C-level must be
reduced before reaching the carbide limit to prevent carbide formation. In carburizing
processes that are monitored and controlled via simulation calculations, the C-level is
automatically lowered before reaching the carbide limit.
When using the case-hardening steel 18CrNiMo7-6, it should be noted that the alloy
factor increases the effective C-level and shifts the carbide limit to lower values com-
pared to an unalloyed steel.
6.2 Thermochemical Heat Treatments 245

Fig. 6.96 Intergranular oxidation and “tiger skin” pattern of the dissolved carbides

Fig. 6.97 Microstructure in the case hardening layer at the 30° tangent

The high proportion of pearlite at the 30° tangent indicated an insufficient quench-
ing speed during sample hardening. Investigations on previously produced series parts of
identical size showed no pearlite formations.

6.2.2 Damage Examples—Nitriding/Nitrocarburizing

Example 1—Unknown Microstructural Phase in Cold Formed Components

Component: Cold formed components

Material: DD13, Material No. 1.0335
Error pattern: Unknown structural phase in the covered area
Heat treatment: Plasma nitrocarburizing
246 6 Heat Treatment Influence Area

On numerous components, only certain functional surfaces are desired to be case-hard-

ened or nitrided/nitrocarburized. To realize this requirement, the remaining surfaces must
be protected by suitable methods. A very commonly used method is the application of a
protective paste, which is supposed to prevent carburization or nitriding.
Different hardening protection agents are used for the different heat treatments. For
individual parts, this is usually done manually with a brush. To prevent the protective
effect from allowing carbon or nitrogen to penetrate, careful work is required. The sur-
faces to be protected must be cleaned. Grease residues under the protective paste would
evaporate at high temperatures and form bubbles or cracks. Sometimes several coats are
necessary if very long treatment times have to be realized. Sufficient drying times must
be observed between the individual coats and before charging. Detailed information is
available from the manufacturers of hardening protection agents.
For geometrically very simple components, a cover cap, e.g., made of copper, can
also be used in plasma-assisted nitriding/nitrocarburizing processes. The cover cap pre-
vents the formation of the plasma and thus also the nitrogen uptake. However, it does not
prevent the penetration of a carburizing gas, which also has a carburizing effect without
plasma formation.
The nitrocarburized functional surface showed a uniformly formed compound layer
with a typical pore rim, Fig. 6.98a. After contrasting with 3% HNO3, the rod-shaped iron
nitrides in the diffusion layer were also visible. The carbide etching agent sodium pic-
rate stains carbides brown, Fig. 6.98b. In the compound layer, the carbon content is dis-
solved, which is why no coloring occurs there.
Deposits were visible on the surface to be protected after contrasting in 3% HNO3,
Fig. 6.99a. These deposits were interpreted by the manufacturer of the sheet metal parts
as a compound layer, which led to the rejection of the entire batch, as a surface free of
compound layer was required at this position.

Fig. 6.98 Formation of the compound layer and representation after contrasting with two differ-
ent etching agents (a) Functional surface with compound and diffusion layer (3% HNO3 etching);
(b) Functional surface with compound and diffusion layer (sodium picrate etching)
6.2 Thermochemical Heat Treatments 247

Fig. 6.99 Microstructure on the surface covered with a copper cap (a) Covered area (3%
HNO3etching); (b) Covered area (sodium picrate etching)

Here too, the contrast with the carbide etching agent sodium picrate shows that these
are not nitrides, but carbides, Fig. 6.99b. The additional check using energy-dispersive
X-ray analysis impressively confirmed the metallographic result, Fig. 6.100.
The protective cap therefore protected against nitrogen uptake, but not against car-
burization. Direct contact with the carbon-donating gas was sufficient for carburization,
which led to a slight carbon uptake. Since there is no austenite formation during nitro-
carburizing and ferrite can practically not dissolve carbon, carbide formation occurred.
However, since iron carbides, like the compound layer, have a high hardness, the treat-
ment batch was not accepted even after the clarifying investigation.

Spectrum 3
Spectrum C Fe

Spectrum 1 13.62 86.38

Spectrum 1 Spectrum 2
Spectrum 2 13.39 86.61

Spectrum 3 – 100.00

Fig. 6.100 Results of EDX measurements in the carbides

248 6 Heat Treatment Influence Area

Example 2—Insufficient Surface Hardness on Toothed Drive Shafts

Component: Toothed drive shafts ∅ 8 mm × 110 mm

Material: 16MnCr5, Material No. 1.7131
Error pattern: Insufficient surface hardness
Heat treatment: Gas nitrocarburizing

Nitrocarburizing in gas, plasma or salt bath usually aims at the formation of a compound
layer to increase adhesive wear resistance. If no large surface pressures occur, a temper-
ing treatment is often omitted before nitrocarburizing. However, for these drive shafts,
the case-hardening steel 16MnCr5 was gas nitrocarburized in the quenched and tempered
state.
A surface hardness of 600 +100 HV 1 was specified. The nitrocarburizing treatment
was carried out at 580 °C for 8 h. In the subsequent hardness test, the surface hardness
values were slightly below the target requirement. The heat treatment company then
decided on an intermediate annealing at 750 °C and the drive shafts were nitrocarburized
again at 580 °C for 8 h. Since the surface hardness after the double nitrocarburizing was
even lower than before, the heat treatment company decided to conduct an investigation.
The verification of the chemical composition confirmed the case-hardening steel
16MnCr5 with a central analysis position, Table 6.10.
The hardness in the core area was 131 HV 0.5 and a limit hardness of 180 HV 0.5
was set for the determination of the nitriding hardness depth, Table 6.11.
The compound layer was very extensive at approx. 25 μm, but there was also a
porosity of approx. 90%, Fig. 6.101. There was intensive carbonitride formation in the
diffusion layer.

Table 6.10  Target and actual analysis of the case-hardening steel 16MnCr5

Target analysis C Si Mn P S Cr Ni Mo Al
16MnCr5, 1.7131 % % % % % % % % %
acc. to DIN EN 0.14 max. 1.00 max. max. 0.80 – – –
683-3
0.19 0.40 1.30 0.025 0.035 1.10 – – –
Measurement results 0.17 0.18 1.12 0.013 0.022 0.99 0.02 <0.01 0.033
All data in mass-%

Table 6.11  Results of the Vickers hardness measurement (core)

Vickers hardness values [HV 0.5] Average value
1st value 2nd value 3rd value 4th value 5th value [HV 0.5]
Measurement 133 134 124 130 134 131
results
6.2 Thermochemical Heat Treatments 249

Fig. 6.101 Documentation of the compound layer and the diffusion layer (a) Highly porous com-
pound layer; (b) Diffusion zone with carbonitrides at the grain boundaries

In the center of the component, a soft annealing microstructure with remnants of the
original quenched and tempered microstructure was shown, Fig. 6.102.
The hardness profile makes it clear that the second nitrocarburizing treatment led to
an even lower surface hardness, Fig. 6.103.
The max. achievable hardness after nitriding/nitrocarburizing is determined by
the amount of nitride-forming alloy elements. In the case of the case-hardening steel
16MnCr5 used here, this is mainly the element chromium. However, the element chro-
mium is also a strong carbide former. If a high tempering temperature was chosen before
nitriding/nitrocarburizing, a considerable amount of the chromium is already bound in
carbides. The insufficient chromium content in the matrix was therefore responsible for
not reaching the target hardness after the first treatment. The intermediate annealing at
750 °C made the result even worse, as it led to the formation of austenite and coarse
carbonitrides formed during slow cooling, which did not lead to a sufficient increase in
hardness.

Fig. 6.102 Documentation of the microstructure in the center of the component

250 6 Heat Treatment Influence Area

400

350
NHD180 0.65 mm
300

250
Hardness [HV 0.5]

200

150

100

50

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Edge distance [mm]

Fig. 6.103 Hardness profile for determining the nitriding hardness depth

References

1. Zoch, H.-W. und Spur, G.: Handbuch Wärmebehandeln und Beschichten. Carl Hanser Verlag
2015. ISBN 978-3-446-43003-7
2. Läpple, V.: Wärmebehandlung des Stahls. Europa-Verlag 11. Auflage 2014 ISBN 978-3-8085-
1330-9
3. Liedtke, D. und 4 Mitautoren: Wärmebehandlung von Eisenwerkstoffen I. expert-Verlag 2017.
ISBN 978-3-8169-3401-1
4. Eckstein, H.-J.: Wärmebehandlung von Stahl, VEB Deutscher Verlag für Grundstoffindustrie,
Leipzig 1971
5. Pitch, W.: Grundlagen der Wärmebehandlung von Stahl. Verlag Stahleisen m.b.H., Düsseldorf
1976
6. Ordinanz, W.: Dauer der Wärmebehandlung von Härtegut. Werkstatt und Betrieb 95 (1962) H.
5, S. 257–259
7. Eckstein, H.-J.: Technologie der Wärmebehandlung von Stahl, VEB Deutscher Verlag für
Grundstoffindustrie, Leipzig 1976
8. Gießmann, Horst: Wärmebehandlung von Verzahnungsteilen, expert-verlag, Renningen 2017
9. DIN EN ISO 898-2: Mechanische Eigenschaften von Verbindungselementen aus Kohlenstoff-
stahl und legiertem Stahl—Teil 2: Muttern mit festgelegten Festigkeitsklassen—Regelgewinde
und Feingewinde
10. DIN EN ISO 683-2:2019-09: Für eine Wärmebehandlung bestimmte Stähle, legierte Stähle
und Automatenstähle—Teil 2: Legierte Vergütungsstähle
References 251

11. DIN EN 10269:2006-07: Stähle und Nickellegierungen für Befestigungselemente für den Ein-
satz bei erhöhten und/oder tiefen Temperaturen
12. DIN EN 10269:2014-02: Stähle und Nickellegierungen für Befestigungselemente für den Ein-
satz bei erhöhten und/oder tiefen Temperaturen
13. DIN EN ISO 18265: Metallische Werkstoffe—Umwertung von Härtewerten
14. DIN 17022-3:1999-04: Wärmebehandlung von Eisenwerkstoffen; Verfahren der Wärmebehan-
dlung; Einsatzhärten
15. Hougardy, H-P.: Umwandlung und Gefüge unlegierter Stähle. Verlag Stahleisen mbH. 1990.
ISBN 3-514-00423-4
16. DIN EN ISO 4957:2018-11: Werkzeugstähle
17. ASTM A534-17: Standard Specification for Carburizing Steels for Anti-Friction Bearing
18. Orlich, J., Rose, A. und Wiest, P.: Atlas zur Wärmebehandlung de Stähle Band 3. Verlag Stahl-
eisen m.b.H. 1973. ISBN 3-514-00133-2
19. Sommer, P. und Ph. Sommer: Doppelschichtbildung beim Nitrieren/Nitrocarburieren.
Abschlussbericht des Projekts 15 des Technologie Forums Werkstoff und Wärme. 2016-01
20. Oppenheim, R.: DEW-Technische Berichte 2 (1962) S. 87/92
21. Schulz, F.: Vergleich verschiedener Methoden zur Bestimmung des Restaustenitgehalts in
gehärteten Proben. Bachelorarbeit Universität Bremen 2013
22. Spur, G. und Th. Stöfele: Handbuch der Fertigungstechnik Band 4/2 Wärmebehandeln. Carl
Hanser Verlag München 1987 ISBN 3-446-14954-6
23. DIN ISO 15787:2018-08: Technische Produktinformation—Wärmebehandelte Teile aus
Eisenwerkstoffen—Darstellung und Angaben
24. Sommer, P., P. Kisters, und C. Sorg: Die Härtbarkeit des Einsatzstahls 18CrNiMo7-6. DER
WÄRMEBEHANDLUNGSMARKT 23 (2016) Heft 3 S. 5–9
25. Eschert, K-P.: 25 Jahre Datenbank StahlWissen®- von den Anfängen bis heute. DER
WÄRMEBEHANDLUNGSMARKT 19 (2016) Heft 3 S. 5–13
26. Sommer, P. und J. Sommer: Einfluss der Abkühlungsgeschwindigkeit auf die Stabilisierung
des Restaustenits. Projekt 10/2007 des Technologie Forums Werkstoff & Wärme
27. Imelmann, W.; Köstlin, K.; Seghezzi, H. D.: Die Zwischenstufenumwandlung als fertigung-
stechnisches Problem. HTM Härterei-Techn. Mitt. 27 (1972) 1 S. 1–6
28. SEP 1664:2004-06: Ermittlung von Formeln durch multiple Regression zur Berechnung der
Härtbarkeit im Stirnabschreckversuch aus der chemischen Zusammensetzung von Stählen.
Beiblatt zum SEP 1664: Ermittlung von Formeln durch multiple Regression zur Berechnung
der Härtbarkeit im Stirnabschreckversuch aus der chemischen Zusammensetzung von Stählen
29. DIN EN ISO 683-1:2018-09: Für eine Wärmebehandlung bestimmte Stähle, legierte Stähle
und Automatenstähle—Teil 1: Unlegierte Vergütungsstähle
30. DIN EN ISO 683-3:2019-04: Für eine Wärmebehandlung bestimmte Stähle, legierte Stähle
und Automatenstähle—Teil 3: Einsatzstähle
31. Jiang, N.Y., Liu, F.S., 2015. Aging Precipitation Evolving Process and its Effects on Mechani-
cal Properties of 0Cr21Ni6Mn9N Austenitic Stainless Steel. MSF. https://doi.org/10.4028/
www.scientific.net/msf.816.255
32. Clausen, B.; F. Hoffmann, H.-W. Zoch, und P. Mayr: Randschichtschädigung beim Härten und
Einsatzhärten von Stählen. HTM Z. Werkst. Wärmebeh. Fertigung 60 (2005) 1, S. 12–18
33. Löser, K., K. Ritter und B. Gornick: Vakuum-Mehrzweckkammerofen zum Vakuumaufkohlen
und Hochdruck-Gasabschrecken. ewi—elektrowärme international—Ausgabe 03 2008
34. Löser, K., V. Heuer, und G. Schmitt: Auswahl geeigneter Abschreckparameter für die Gasab-
schreckung von Bauteilen aus verschiedenen Einsatzstählen. HTM Z. Werkst. Wärmebeh. Fer-
tigung 60 (2005) 3, S. 142–149
252 6 Heat Treatment Influence Area

35. Clausen, B.; Laue, S.; Burtchen, M.; Hoffmann, F.; Mayr, P.: Randschichtschädigung infolge
Niederdruckaufkohlung. HTM Z. Werkst. Wärmebeh. Fertigung 58 (2003) S. 13–19
36. Sommer, P.: Failures Caused by Hydrogen Absorption during Heat Treatment Processes.
Materials Testing 56 (2014) 5, S. 450–455
37. ISO 6336-5:2016-08: Calculation of load capacity of spur and helical gears—Part 5: Strength
and quality of materials
Manufacturing Influence Area After Heat
Treatment 7

Contents

7.1 Damage Examples—Manufacturing After Heat Treatment. . . . . . . . . . . . . . . . . . . . . . . . . 255

References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274

The processing of hardened components with a hardness of at least 47 HRC is referred to

as hard machining [1]. The most well-known hard machining methods are grinding, hard
turning, hard drilling, hard milling and wire and sinker EDM. Detailed explanations of
the cutting hard machining methods can also be found in [1]. Wire and sinker EDM are
very often used in toolmaking [2].
The various methods for surface strengthening, e.g. shot peening, are primarily used
to improve the properties of the surface layer and are not formative hard machining in
the true sense.
An overview of the diverse and possible manufacturing processes after heat treatment
is listed in Fig. 7.1.
Hard machining after heat treatment is a cost-intensive manufacturing step and also
not risk-free. All methods have in common that high machining forces must be applied,
which lead to intense heat development on the surfaces to be machined. Only by suffi-
cient cooling during machining can unwanted microstructural changes on the component
surface be avoided. Otherwise, there can be new hardening layers and crack formations.
With large-volume cutouts by wire EDM, the residual stress state is greatly changed,
which can lead to a significant shape change.

© The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, 253
part of Springer Nature 2024
P. Sommer, Mistakes Before, During and After Heat Treatment of Steel,
https://doi.org/10.1007/978-3-658-44160-9_7
254 7 Manufacturing Influence Area After Heat Treatment

Manufacturing process after heat treatment

Mechanical hard treatment
Eroding Grinding
Wire EDM Hard turning
Die sinking Hard drilling
Hard milling
Thermal shaping
Laser cutting Surface hardening
Plasma cutting Cleaning blasting
Shot blasting
Straighten
Mechanical Engraving
Flame Electro-engraving
Laser engraving

Fig. 7.1 Possible manufacturing processes after heat treatment

In the case of thermochemical heat treatment processes, it is particularly important

to determine a machining allowance, as the surface layer is changed in its properties. The
machining allowance is in many cases based on experience and is oriented towards the
expected dimensional and shape changes during heat treatment. If the dimensional and shape
change is symmetrical, a symmetrical hard machining can also take place. The situation
becomes somewhat more difficult if there has been an asymmetrical distortion during heat
treatment and dimensions or properties of functional layers can no longer be maintained due
to machining. This applies in particular to case-hardened components, as an asymmetrical
removal can lead to an uneven case hardening depth and failure under operating conditions.
However, property improvements are also achieved in hard machining. After case
hardening of alloyed case-hardening steels in endothermic atmospheres, there is always
an unavoidable intergranular oxidation. As a result of the intergranular oxidation, a soft
non-martensitic microstructure seam is formed, which has a comparable depth to the
depth of the intergranular oxidation. This soft surface layer can be ground off on tooth
flanks, but this is practically impossible in the tooth root area. A targeted cleaning blast
leads to an increase in the service life under cyclic loads.
Since high temperature loads occur in all hard machining processes, new hardening
layers with friction martensite and higher tempered layers behind the friction martensite
(dark etching areas) can occur if the cooling is insufficient or the machining parameters
are incorrect. Grinding burn is one of the most well-known manufacturing defects on
case-hardened or surface-hardened components.
The range of possible treatment errors in hard machining is as large as in the other
treatment steps, as the following examples show.
7.1 Damage Examples—Manufacturing After Heat Treatment 255

7.1 Damage Examples—Manufacturing After Heat Treatment

Example 1—Stamp Fracture on a Tablet Press Tool After Short Use

Component: Tablet press tool

Material: X100CrMoV5, Material No.: 1.2363
Error image: Stamp fracture after short usage
Heat treatment: Vacuum hardening and tempering

The damage case to be described below is a classic in the area of manufacturing after
heat treatment. A tablet press tool made of cold work steel X100CrMoV5, material no.
1.2363, broke vertically to the longitudinal axis after a very short period of use. On the
fracture surface, a small fatigue fracture surface with about 5% area share and a corre-
spondingly large residual fracture were present, Fig. 7.2.
The tool was marked with a laser engraving after hardening and tempering. This led
to local melting and the formation of untempered martensite and high proportions of
residual austenite, Fig. 7.3a, b. A fatigue crack started at this position. The high hard-
ness and the associated low toughness led to a large-scale residual violent fracture after
a small spread. The violent fracture surface was very fine-grained over the entire cross-
section, indicating good material homogeneity and appropriate heat treatment.
The microstructure after hardening and tempering was 60.4 HRC, within the target
specification of 59 +4 HRC. The material showed a dimension-typical linearity with few
primary carbides, Fig. 7.3c, d.

Fig. 7.2 Small fatigue fracture surface and large residual fracture
256 7 Manufacturing Influence Area After Heat Treatment

Fig. 7.3 Micristructure in the area of the laser engraving and the base material (a) Melting due to
a laser engraving; (b) Side view of the melting zone; (c) Microstructure in the base material; (d)
Fine acicular tempered martensite and line-arranged primary carbides

Example 2—Crack Formation on an Inner Lamella After Short Use

Component: Inner lamella Ø 325 mm

Material: C75, Material No. 1.0605
Error image: Crack formation after short usage
Heat treatment: Hardening and tempering

After a few days in operation, an inner lamella in a drive was broken in several places
and further cracks were present, Fig. 7.4. The inner lamella was purchased and, accord-
ing to the manufacturer, made from the spring steel C75, 1.0605. The test certificate indi-
cated the condition as “hardened and tempered” and the hardness was given as 44 HRC.
No further details about the manufacturing process were available.
The material composition and the hardness on the flat sides of the inner lamella were
checked and the information from the test certificate was confirmed. The microstruc-
ture in the base material also showed a proper microstructure of tempered martensite,
Fig. 7.5.
7.1 Damage Examples—Manufacturing After Heat Treatment 257

Fig. 7.4 Fragments of the broken inner lamella and further crack formations

Fig. 7.5 Proper quenching and tempering microstructure in the base material of the inner lamella

On the surface, however, there was a contour-close new hardening layer, which was
cracked, Fig. 7.6. In the martensitic new hardening layer with a thickness of approx.
150 μm, a hardness of 835 HV was present. From this microstructure, it was evident that
the gearing was made by laser cutting and was not post-treated. The hardness of 835 HV
corresponded to the max. quenching hardness of the used spring steel. The crack and
fracture formation always started in the area of the highest bending stress. Due to the
very high brittleness, even at low stress, cracking occurred.

Example 3—Breakage of a Chain Pin

Component: Chain pin Ø 30 mm

Material: 23CrMoB3-3, Material No. 1.7271
Error image: Pin breakage
Heat treatment: Quenching and tempering Tempering
258 7 Manufacturing Influence Area After Heat Treatment

Fig. 7.6 Contour-accurate new hardening layer on the surface with crack formation

Fig. 7.7 Fatigue fracture surface with small residual force fracture surface and cracks on the pin
surface

The boron-alloyed steel 23CrMoB3-3 used here is not included in DIN EN ISO 683-
2:2018-09 [3], but it is listed in the Steel-Iron List [4] under the material number 1.7271.
Chain pins were made from this steel and case-hardened. After more than two years of
service life, a breakage of the chain pin occurred. The fracture surface showed a large
7.1 Damage Examples—Manufacturing After Heat Treatment 259

fatigue fracture portion and correspondingly small residual force fracture, Fig. 7.7. The
shiny areas on the fatigue fracture surface were caused by the relative movement of both
fatigue fracture halves.
Numerous network-like cracks were visible on the pin surface.
The crack network on the bolt surface already suggested grinding burn damage. This was
confirmed by metallographic examination, Fig. 7.8. Friction martensite layers had formed
all around. Behind this re-hardening layer, the martensitic microstructure etched darkly, as
the heat exposure from the grinding process caused a higher tempering temperature.
Numerous cracks had formed, one of which spread as a fatigue fracture. Due to the
very small residual fracture surface, a low nominal load could be inferred.

Example 4—Corrosions on a Bearing Inner Ring

Component: Bearing inner ring ∅ 220 mm

Material: 100Cr6, Material No. 1.3505
Error image: Corrosions
Heat treatment: Hardening and tempering

Fig. 7.8 Re-hardening layers and darkly etching microstructure behind the friction martensite
260 7 Manufacturing Influence Area After Heat Treatment

Rolling bearing components require a very good macroscopic and microscopic degree
of cleanliness. While the determination of the microscopic degree of cleanliness is only
possible through destructive metallographic tests, ultrasonic testing is used to determine
the macroscopic degree of cleanliness. To increase the detection sensitivity, ultrasonic
tests can be carried out using immersion technique [5, 6]. In particular, bearing rings are
very suitable for this due to their rotationally symmetrical geometry.
The surface of a bearing inner ring with ∅ 220 mm tested in UT immersion technique
became noticeable due to a corrosion attack during the test, Fig. 7.9. Numerous corro-
sion scars formed, Fig. 7.10.
The composition of the corrosion products was determined by energy dispersive
X-ray analysis (EDX), Fig. 7.11. Traces of chlorine and sulfur were also detectable in
the corrosion products. For comparison, the raceway surface was also measured, where
neither chlorine nor sulfur could be detected.

Fig. 7.9 Damaged running surface of a new bearing inner ring part

Fig. 7.10 Corrosion scars on the raceway surface

7.1 Damage Examples—Manufacturing After Heat Treatment 261

Spectrum 1
Spectrum 2

Spectrum 4
Spectrum 3

O Si S Cl Ca Cr Fe

Spectrum 1 18.54 0.58 0.53 0.28 - 5.07 Rest

Spectrum 2 24.23 0.41 0.45 - - 3.30 Rest

Spectrum 3 32.15 0.49 0.47 0.24 0.27 5.61 Rest

Spectrum 4 - 0.43 - - - 1.42 Rest

Fig. 7.11 EDX measurements of the corrosion products and the bearing surface

Based on this finding, the immersion technique fluid was evaporated on a pure
iron carrier material and the residues were also measured using EDX. Here too, small
amounts of sulfur and chlorine were determined.
Furthermore, comparative corrosion tests were carried out with the immersion tech-
nique fluid and pure tap water on samples of the bearing inner ring. The samples in the
immersion technique fluid showed clear signs of corrosion after just 15 minutes, while
the samples in pure tap water were still corrosion-free after this test duration. After a
30-minute test duration, the samples in tap water also showed initial corrosions.
The origin of the low sulfur and chlorine content could not be determined.

Example 5—Shrink Disc Fracture

Component: Shrink disc ∅ 820 mm

Material: 50CrMo4, Material No. 1.7228
Error image: Failure due to fracture
Heat treatment: Tempering
262 7 Manufacturing Influence Area After Heat Treatment

Shrink discs are used as shaft-hub connections. After tempering, oil holes were drilled in
a radial direction. Starting from an oil hole, several small-scale fatigue cracks occurred,
resulting in a very large residual force fracture, Fig. 7.12.
During drilling, intense friction martensite formation occurred, Fig. 7.13. Starting
from this brittle new hardening layer, numerous cracks formed, which grew under cyclic
loading. Due to the inherently high tension of a shrink disc, a very large residual force
fracture occurred.

Example 6—Crack Formations on the Surface of a Piston Rod

Component: Piston rod

Material: 42CrMo4, Material No. 1.7225
Error image: Crack formations on the surface
Heat treatment: Quenching and Tempering, Inductive hardening and Chrome plating
­tempering,

Fig. 7.12 Fatigue cracks on the oil hole surface

Fig. 7.13 Friction martensite layer on the oil hole surface

7.1 Damage Examples—Manufacturing After Heat Treatment 263

During the magnetic particle inspection of a finished piston rod, crack formations were
discovered, Fig. 7.14. The forged piston rod was quenched and tempered, inductively
hardened, and ground before chrome plating. The crack formation on the surface was
network-like, which already indicated grinding burn cracks.
The metallographic examination on a microsection perpendicular to the crack forma-
tion confirmed the grinding burn crack formation, Fig. 7.15. The inductively hardened
surface was annealed higher due to the excessive heating during grinding, which led to
a dark etchability. A friction martensite layer was no longer present. The heat input from
grinding has led to a hardness reduction of 200 HV, Fig. 7.16.

Example 7—Crack Formations on the Tooth Flanks of a Spur Gear

Component: Spur gear Module 10

Material: 18CrNiMo7-6, Material No. 1.6587
Error image: Crack formations on the tooth flanks
Heat treatment: Case hardening CHD550 = 1.5 +0.5 mm

Fig. 7.14 Piston rod and visible cracks after the magnetic particle inspection

Fig. 7.15 Shell-shaped crack and dark etching surface

264 7 Manufacturing Influence Area After Heat Treatment

800

700

600

500
Hardness [HV 0.2]

400

300

200

100

0
0.0 1.0 2.0 3.0 4.0 5.0
Edge distance [mm]

Fig. 7.16 Hardness profile perpendicular to the ground surface

The grinding burn test according to ISO 14104:2017-04 [7] is one of the standard
test methods after the completion of flank grinding. During the grinding burn test,
cracks were discovered on some tooth flanks of a case-hardened spur gear, module 10,
Fig. 7.17.
The metallographic examination clearly revealed the cause of the crack. The dark
etching areas beneath the tooth flank surface demonstrated an excessively high heat input
during grinding, Fig. 7.18. Crack formation occurs due to grinding-induced tensile resid-
ual stresses. In [8], the residual stress profiles for four different situations are explained,
Fig. 7.19.
Proper grinding processing leads to high compressive residual stresses on the surface.
Already the formation of tempering zones, as can be seen in Fig. 7.18, causes tensile
residual stresses on the surface. If the tensile stresses exceed the separation strength,
crack formation occurs.
If there are new hardening zones, the maximum of the tensile residual stresses is
slightly below the surface, and in the case of multi-stage grinding, there are indeed low
compressive residual stresses on the surface, but tensile residual stresses are also present
at a shallow depth. In both of these cases, cracking can occur.
7.1 Damage Examples—Manufacturing After Heat Treatment 265

Fig. 7.17 Spur gear and cracks on the tooth flank

Fig. 7.18 Grinding burn without and with crack formation

Example 8—Breakage of a Ship’s Crankshaft

Component: Ship's crankshaft

Material: 42CrMo4, Material No. 1.7225
Error image: Total failure due to the break of the crankshaft
Heat treatment: Quenching and tempering Tempering to a hardness of 280—320 HBW

A ship owner had arranged for the ship's crankshaft of his cargo ship to be overhauled
after 13 years and approximately 87,000 operating hours. After completion and reinstal-
lation, there was a total failure after only 894 further operating hours, as the crankshaft
had broken in the area of the fourth crankpin. For transport, the two fracture halves had
been fixed with a massive metal plate, Fig. 7.20.
Since both fracture halves had been reassembled, the crack progression was very easy
to see, Fig. 7.21.
266 7 Manufacturing Influence Area After Heat Treatment

Train ground, no grinding burn ground, tempering zone

Train
Residual stresses

Residual stresses
Depth Depth
Print

Print
ground, new hardening zone Multi-stage grinding,
with surrounding tempering zone concealed grinding burn
Train

Train
Residual stresses

Residual stresses
Depth Depth
Print

Print

Fig. 7.19 Schematic representation of residual stress profiles [8]

Fig. 7.20 Fracture halves of the overhauled ship’s crankshaft fixed with a metal plate

The opened fracture surface revealed a large fatigue fracture surface with an origin
at the radius and a small residual force fracture surface, Fig. 7.22. For microstructure
examination, a microsection was taken, the location of which is also shown in Fig. 7.22.
Further microsections were taken, which are not listed in this damage description.
The metallographic examination revealed a massive change in the radius surface,
Fig. 7.23. Up to a depth of 135 μm, the surface was thermally damaged. There was an
extensive friction martensite layer and a darkly etching microstructure seam.
In the quenching and tempering microstructure of the base material, a segregation-
related microstructure linearity with pearlite formation was present in negatively segre-
gated streaks, Fig. 7.24.
7.1 Damage Examples—Manufacturing After Heat Treatment 267

Fig. 7.21 Crack progression in the area of the fourth crankpin (a) Progression into the compo-
nent; (b) Progression on the surface

Ground joint position

Fatigue fracture Residual force fracture

Fig. 7.22 Fatigue fracture surface with origin at the radius and residual fracture surface

In the hardness progression perpendicular to the radius surface, a max. hardness of

505 HV was measured, Fig. 7.25.
The strongly developed friction martensite layer with the dark etching layer led to
tensile stresses that were sufficiently high to initiate a crack (see also Fig. 7.19). In a
joint final discussion of all participants, it was clarified that all radii had been dry ground
and polished. The inspection of the other radii with a UCI hardness tester revealed simi-
lar hardness values around 500 HV 1. Accordingly, all radii were at risk of cracking,
with the crack formation occurring at the position with the greatest bending stress.
268 7 Manufacturing Influence Area After Heat Treatment

135.2 µm
79.7 µm

Fig. 7.23 New hardening layer and darkly etching edge below the radius surface (a) Transition
from the radius surface to the fatigue fracture surface; (b) Microstructure formation slightly below
the crack starting position

Fig. 7.24 Segregation-related microstructure linearity with pearlite formation in negatively segre-
gated lines (a) Lamellar microstructure in the base material; (b) Upper bainite and pearlite in the
quenched and tempred microstructure

The same hardness test with the UCI hardness tester was carried out on two further
crankshafts that had not yet been reinstalled. The same manufacturing error was present
on these crankshafts as well.

Example 9—Breakage of Heads on Thread-Cutting Screws

Component: Thread-cutting screws M4 × 14

Material: C15, Material No. 1.0401
Error image: Breakage of screw heads 2 days after assembly
Heat treatment: Case hardening, tempering and galvanic coating
7.1 Damage Examples—Manufacturing After Heat Treatment 269

900

800

700

600
Hardness [HV 1]

500

400

300

200

100

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Edge distance [mm]

Fig. 7.25 Hardness progression perpendicular to the radius surface

The possible uses of thread-cutting screws are very diverse. According to DIN 7500-
1:2009-06 [9], this type of screw is defined as a screw that forms its own counter-
thread by cutting when screwed into a pre-drilled core hole, e.g. with the help of cutting
grooves, Fig. 7.26. For this purpose, screws were made from the unalloyed case-harden-
ing steel C15, 1.0401, by cold forming and subsequently carbonitrided, case-hardened
and tempered with the addition of ammonia. The tempering temperature was specified as
280 °C for 1 hour. For corrosion protection, the screws were galvanically zinc-plated and
finally tempered.
The individual manufacturing steps were communicated and are listed below:

Screw size: M4 × 14
Material: C15, 1.0401
Carburizing: 910 °C 40 min
C-level: 1.0% + 5% ammonia
Quenching: High-performance quenching oil
Annealing: 280 °C 1 h
Pickling: 3 × 7.5 min in 12% hydrochloric acid
Galvanizing: 1–2 μm zinc layer
Tempering: 210 °C 8 h
Final galvanizing: 10–12 μm zinc layer
270 7 Manufacturing Influence Area After Heat Treatment

Fig. 7.26 Operating principle of thread-cutting screws (Image source: https://www.bossard.com/

eshop)

The screws were used to secure power connection boxes in household appliances (refrig-
erators, dishwashers, and washing machines). A few hours after assembly, several screw
heads had broken off and there was a risk that the power connection box could fall
onto the housing floor and lead to electric shocks during domestic use. As a precaution,
screws from this batch were therefore dismantled on 33,000 household appliances and
replaced with a new checked delivery.
The delayed breakage a few hours after assembly and the examination of the screw
fracture surfaces clearly indicated a hydrogen-induced brittle fracture, Fig. 7.27. The
review of the surface hardness revealed two very different hardness levels with 580 HV
0.3 and 860 HV 0.3 respectively.
As there were contradictory statements about the origin of the hydrogen, the individ-
ual manufacturing steps were retraced and checked after completion. The total hydrogen
contents were measured by melt extraction before and after each manufacturing step. In
addition, stress tests were carried out according to DIN EN ISO 15330:2000-01 [10],
Fig. 7.28. For this, the screws were stressed with 90% of the fracture torque and checked
for head fractures every 24 hours. After 96 h, the stress test ended and by loosening the
screws it was checked whether there were fractures in the thread area.
In the initial state, the total hydrogen content was below 0.5 ppm, which is a typical
value for hot-rolled wires. After carburizing and case hardening, the value increased to
2.6 ppm, and reduced to 0.7 ppm after annealing at 280 °C for 1 h, Fig. 7.29. The stress
test on about 120 screws in the non-annealed state led to almost 100% stress failures,
while no stress failures occurred in the annealed screws with the same test quantity.
All results of the tension tests are listed in Table 7.1. Taking into account the surface
hardness measured on the damaged parts, two causes could be identified. Screws with
a surface hardness of 860 HV 0.3 could neither be tempered nor annealed. This batch
left the hardening shop without annealing and was also not properly tempered. The sur-
7.1 Damage Examples—Manufacturing After Heat Treatment 271

Fig. 7.27 Intergranular fracture surface of the hydrogen-induced brittle fracture

Fig. 7.28 Stress test according to DIN EN ISO 15330-01

< 0.5 ppm

before case
hardening

2.6 ppm
after 910 °C 40 min oil

0.7 ppm Case

hardening
+ 280 °C 1h

0.0 0.5 1.0 1.5 2.0 2.5 3.0

Hydrogen content [ppm]

Fig. 7.29 Hydrogen content before and after case hardening

272 7 Manufacturing Influence Area After Heat Treatment

Table 7.1  Results of the tension tests of different treatment conditions

Treatment condition Number of tension failures
Case hardening without annealing Nearly 100%
Case hardening with annealing None
Case hardening without annealing + galvanizing without tempering 100%
Case hardening without annealing + galvanizing with tempering Few
Case hardening with annealing + galvanizing without tempering Numerous
Case hardening with annealing + galvanizing with tempering None

face hardness of 580 HV 0.3 corresponded to the hardness after annealing at 280 °C for
1 hour. Since there were still tension failures, there must also have been improper tem-
per treatment here.

Example 10—Crack Formation on a Mold Insert After Short Use

Component: Mold insert

Material: Special hot work steel with 3.5% Mo and 2.0% W
Error pattern: Crack formation after short service life
Heat treatment: Vacuum hardening and annealing

A mold insert made of a Mo-W alloyed special hot work steel had failed after a very
short service life due to multiple radial and axial crack formations, Fig. 7.30. It was
known that a vacuum hardening with annealing to a hardness of 50 HRC was carried out
for production. The process data of the heat treatment were not communicated. After the
quenching and tempering, the delicate shaping was done by sinker EDM.

Fig. 7.30 Mold insert with crack formations

7.1 Damage Examples—Manufacturing After Heat Treatment 273

In a first investigation section, the axial crack was opened, Fig. 7.31. On the opened
crack surface, it could be seen that the crack ran from inside to outside and the beach-
marks showed a cyclic crack growth.
The determination of the chemical composition confirmed the information from
the supplying plant and the checked hardness exactly matched the target hardness of
50 HRC.
In the metallographic examination, the microscopic purity level was first determined
according to ISO 4967 Method A, Table 7.2. Although there were no specifications, the
results indicated a very high purity of the used steel melt.
The quenched and tempered microstructure in the base material also provided no
clues to a possible cause of damage, Fig. 7.32. The tempered martensite was slightly
coarsened, but there were no pre-eutectoid carbide precipitations at the grain boundaries.
Therefore, a sufficiently high toughness was to be assumed.
The cause of the damage was evident on the eroded surfaces, Fig. 7.33. There were
new hardening layers and remnants of re-solidified melts present. Cracks formed on
this faulty surface, which propagated into the base material. During sinker erosion,
such layers always occur, however, the discharge energy must be reduced from the
roughing voltage to the finishing voltage as it approaches the final contour [2]. Even
if the finishing voltage has prevented a new hardening layer, a stress relief annealing
30–50 °C below the tempering temperature is recommended. However, this stress relief
annealing can only be successfully applied if no crack formation has occurred during
erosion.

Fig. 7.31 Opened crack with beachmarks

Table 7.2  Results of the purity grade evaluation according to ISO 4967 Method A
A B C D DS
fine thick fine thick fine thick fine thick –
0 0 0 0 0 0 0 0.5 0
274 7 Manufacturing Influence Area After Heat Treatment

Fig. 7.32 Quenched and tempered microstructure with slightly coarsened tempered martensite

Fig. 7.33 Eroded surface with non-removed melt and new hardening layers

References

1. Tönshoff H. K.; Denkena, B.: Spanen Grundlagen. Springer-Verlag, 2. Erweiterte Auflage;

2004
2. Wendl, F.; Wupper, K.-D.: Einfluß der Abschmelzrate beim Elektroerodieren auf Gefügeaus-
bildung und Zähigkeit einiger Werkzeugstähle. HTM Härterei-Techn. Mitt. 44 (1989) S. 197–
204
3. DIN EN ISO 683-2:2018-09: Für eine Wärmebehandlung bestimmte Stähle, legierte Stähle
und Automatenstähle—Teil 2: Legierte Vergütungsstähle
4. Register Europäischer Stähle (Stahl-Eisen-Liste) www.stahldat.de
5. Buckstegge, J.: Erfahrungen beim Einsatz der Ultraschall-Tauchtechnik zur Bestimmung des
Makroreinheitsgrades von Wälzlagerstählen; HTM Härterei-Techn. Mitt. 54 (1999) 4 S. 250–
258
6. Hering, J.: Ultraschall-geprüfte Stahlreinheit—Verfahren für die Praxis, Ergebnisse aus der
Praxis; HTM Härterei-Techn. Mitt. 54 (1999) 4 S. 259–265
7. ISO 14104:2017-04: Gears—Surface temper etch inspection after grinding, chemical method
8. Seidel M.W.: Schleifbrand und dessen Prüfung. Hanser-Verlag 2020
References 275

9. DIN 7500-1:2009-06: Gewindefurchende Schrauben für Metrisches ISO-Gewinde –Teil 1:

Technische Lieferbedingungen für einsatzgehärtete und angelassene Schrauben
10. DIN EN ISO 15330:2000-01: Verbindungselemente—Verspannungsversuch zur Entdeckung
von Wasserstoffversprödung—Verfahren mit parallelen Auflageflächen
Operating Conditions Influence Area
8

Contents

8.1 Damage Examples—Operating Conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282

References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312

The potential uses of heat-treated components are virtually unlimited in number. So are
the possible errors that can occur. The error types listed in Fig. 8.1 provide an initial
overview.

Overload
In general, damage can occur if the load conditions that the designer has based on a com-
ponent are exceeded. The result of this are violent fractures. Depending on the material
condition and loading speed, brittle fractures or deformation fractures occur. In the case
of deformation fractures, an overload can sometimes be identified by a constriction or
plastic deformation. Fig. 8.2 shows a crossbeam from a metal shelf that was overloaded.
The fracture surface shows the typical shear dimples of a deformation fracture.

In the case of high-strength material conditions, however, there are fractures with-
out deformation. In Fig. 8.3 the left part of the image shows a brittle torsional fracture
of an extruder screw. This fracture occurred when a metallic foreign body got into the
extruder, leading to a spontaneous blockage. In contrast to the brittle fracture of an
extruder screw, the right part of Fig. 8.3 shows a tough torsional fracture. The plastic
deformations on the spline are also clearly visible.

© The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, 277
part of Springer Nature 2024
P. Sommer, Mistakes Before, During and After Heat Treatment of Steel,
https://doi.org/10.1007/978-3-658-44160-9_8
278 8 Operating Conditions Influence Area

Area of influence Operating conditions

Overload
Corrosion Static

Scarring dynamic

Cross-section loss Single event

Temperature
Wear and tear Overtemperature
Lack of lubricant Low temperatures

Current passage
Operating error

Fig. 8.1 Listing of some causes of errors during use

Fig. 8.2 Overload fracture of a crossbeam from a metal shelf and shear dimples fracture surface

Fig. 8.3 Brittle torsional fracture (left); Tough torsional fracture (right)
8 Operating Conditions Influence Area 279

The causes of possible overload situations are very diverse and in most cases caused
by disruptions of normal plant operation. These processes are rarely admitted by the
respective plant operators. Often, overload situations are not recognized during opera-
tion. It does not always have to lead to a spontaneous violent fracture. A single overload
impulse can lead to a first crack, which then propagates under nominal load conditions
and finally ends as a fatigue fracture with a small residual fracture portion, Fig. 8.4.

Corrosion
Most steels used in mechanical engineering are unalloyed or low-alloyed and do not
have corrosion resistance, which is why environmental influences are of great impor-
tance. In a corrosive environment, this leads to a corrosion attack with sometimes seri-
ous consequences. Corrosion scars cause a significant notch effect, which leads to a
significant reduction in dynamic strength properties. The example of the bending fatigue
strength of quenched and tempered steels shows the influence of corrosion by pipe and
salt water, Fig. 8.5. The damaging effect of corrosion also increases with the increase in
tensile strength.

A heat-treatable steel with 1000 MPa tensile strength therefore loses 70% of the bend-
ing fatigue strength compared to the undamaged surface. As can be seen from Fig. 8.5,
the damaging effect of corrosion is even higher than the damage caused by mechanical
notches.

Fig. 8.4 Fatigue fracture with a residual fracture portion of approx. 50% (left); Fatigue fracture
with a residual fracture portion <5% (right)
280 8 Operating Conditions Influence Area

0
Reduction in flexural fatigue strength [%]

20

40

60

Polished surface
Ground surface
Roughened surface
80
with ring-shaped pointed notch
with rolled skin (lower limit)
Corrosion due to tap water
Corrosion due to seawater
100
400 600 800 1000 1200 1400
Tensile strength [MPa]

Fig. 8.5 Influence of surface quality on the bending fatigue strength of quenched and tempered
steels [1]

Temperature
Almost all mechanical properties of steels are temperature dependent. While hardness
and tensile strength decrease at elevated operating temperatures, a decrease in toughness
must be expected at low temperatures.

For areas of application at higher temperatures, the highest possible hot hardness is
required. If the permissible stress temperature is exceeded, the tool or component loses
its shape stability, which can lead to failure. However, it is not only the decrease in hard-
ness and tensile strength that is relevant, with increasing temperature the reactivity of the
steel surface with the surrounding atmosphere also increases. The change in the surface
through reactions or reaction products can even be the dominant damage mechanism.
This applies, for example, to “metal dusting”. The mechanical properties of a heat-resist-
ant steel are by no means exhausted at 600 to 700°C. However, contact with carburizing
gases can lead to complete failure.
Insufficient preheating of temperature-stressed components is another possible cause
of damage. During forging, injection molding or glass forming, there is a brief high con-
tact temperature on the surface. Since the temperature difference within a component is
directly proportional to the thermal stresses that develop, preheating can reduce the tem-
perature difference and thus the stress caused by thermal stresses. If preheating is omit-
ted or insufficient, this can lead to crack formation.
8 Operating Conditions Influence Area 281

Formation
temperature
Austenite grain size
Notched impact strength

Carbide size
Tempering Carbide precipitates at
temperature and the grain boundaries
time

Test temperature

Fig. 8.6 Schematic notch impact toughness-temperature curve and influences of microstructure
formation

The change in toughness with falling temperature is typically described by notch

impact toughness-temperature curves, Fig. 8.6. Within a narrow temperature range,
the fracture behavior changes from a tough to a brittle fracture. The test temperature at
which equal proportions of tough and brittle fracture are achieved is referred to as the
transition temperature.
The influence of the microstructure formation on the position of the transition temper-
ature is schematically described in Fig. 8.6. The material selection and the heat treatment
condition thus determine the lower permissible operating temperature. If this is undercut,
brittle fractures can occur.

Wear
Wear is according to [2] “the propagating material loss from the surface of a solid body,
caused by mechanical causes, i.e. contact and relative movement of a solid, liquid or
gaseous counter body”. A moving component is therefore always subject to wear and
must therefore have sufficient wear resistance. However, since wear is a system property,
the operating conditions must be equally included in the wear assessment. An optimal
surface condition, for example of a case-hardened gear, fails after a short period of use
282 8 Operating Conditions Influence Area

if there is a lack of lubricant or if the lubricant is contaminated. The same applies in

reverse if optimal lubricant conditions exist and the surface condition is deficient. These
errors have been described in the previous chapters.

Other Operating conditions

The list of possible errors in the usage phase of components and tools is by no means
completed with the four influencing factors described above. However, these four influ-
encing factors already cover a large area of investigated damage cases. The specific
damage cases described below do not only come from the areas of overload, corrosion,
temperature and wear, but also deal with, for example, damage to bearings caused by
electric currents and wrong decisions in improving operating conditions. It must also be
considered that faulty operating conditions cannot always be proven. This means that
assumptions must also be made in the damage assessment.

8.1 Damage Examples—Operating Conditions

Example 1—Breakage of a Load Hook

Component: Double hook according to DIN 15402:1982-11

Material: P355N, Material No. 1.0562
Error image: Breakage of the load hook
Heat treatment: Tempering

Fig. 8.7 Documentation of the broken load hook

8.1 Damage Examples—Operating Conditions 283

A double hook according to DIN 15402:1982-11 [3] of a 12.5 t crane system broke while
lifting rods with a total weight of 10 t, Fig. 8.7. The breakage occurred directly during
lifting, which is why the rods were not moved.
From the marking, it could be seen that it was a load hook of strength class P with
the load hook number 5 according to DIN 15402. According to DIN 15400 [4], these
load hooks are made from the material StE355 (current designation: P355N), material
no. 1.0562.
The crane hook broke in the area of the threaded shaft. The fracture surface showed
two fatigue fracture surfaces offset by 180° with the typical macroscopic beachmarks
and the very flat fracture structure, Fig. 8.8.
The crack propagation occurred from both fatigue cracks towards the core of the
threaded shaft. In the center of the fracture surface was the residual force fracture with
a strongly fissured fracture structure. The residual force fracture portion was about 50%.
Numerous cracks were visible on the surface. The present fracture characteristics are
typical for a fatigue fracture under double-sided bending alternating stress with high
notch effect on the entire circumference [5].
Upon closer inspection of the thread, it was found that several more cracks were
already present in the thread turns behind the fracture surface, Fig. 8.9.
A very homogeneous microstructure was present without any recognizable segre-
gations, Fig. 8.10a. Behind the fatigue fracture surface, another crack initiation could
be seen, Fig. 8.10b. The thread surface showed minor cold deformation features from
mechanical processing. The microstructure of ferrite and high-tempered bainite was very
fine-grained, Fig. 8.11.

Fig. 8.8 Fracture surface and model image from a double-sided bending
284 8 Operating Conditions Influence Area

Fig. 8.9 Further cracks also in other thread turns

Fig. 8.10 Microstructure formation in the area of the fatigue fracture (a) transition to the fatigue
fracture surface with minor cold deformation on the surface; (b) thread with another crack initia-
tion behind the fatigue fracture surface

Fig. 8.11 Very fine-grained microstructure of ferrite and high-tempered bainite

8.1 Damage Examples—Operating Conditions 285

The chemical composition and mechanical properties met the specifications of DIN
15400, which is why a presentation of these results was omitted.
From the investigations carried out, it could be inferred that the load hook was so pre-
damaged by a fatigue fracture that the residual fracture occurred when lifting a load of
10 t. The material used was standard-compliant and showed good purity and homogene-
ity as well as a very fine-grained quenched and tempered microstructure.
Hints about the cause of the crack initiation were not discernible from the inves-
tigation results, especially since the crane hook had been in use for several years and
was regularly checked for cracks. It must therefore be assumed that the cracks can be
attributed to one or more previous overload events. The time of crack initiation could be
localized to the period since the last crack inspection. The case of damage is a typical
example of the fact that assumptions had to be made, as a direct proof of overload, which
must have caused the damage, was not possible.

Example 2—Violent fracture of a connecting element

Component: Connecting element of a construction crane

Material: St52-3 (S355J2G3), Material No. 1.0570
Error image: Violent fracture
Heat treatment: Normalizing

This damage is one of the few cases where both the manufacturer and the operator have
assessed the cause of the damage in the same way. On January 18, 2007, the cyclone
Kyrill swept across Northern Europe. Wind speeds of up to 225 km/h were measured.
On this day, a construction crane fell onto a building of the University in Utrecht and
destroyed large parts of the building, Fig. 8.12. Fortunately, no one was injured, as the
building was empty at the time of the crane accident.
Due to the high wind load, a break occurred in a concreted connection element,
Fig. 8.13. In the component drawing for this part, the earlier steel designation St52-3 was
still used, which was replaced by the designation S355J2G3 in the current standard. The
use of earlier steel designations in a drawing or in a heat treatment order is not an iso-
lated case. If there is a need for clarification, this should be checked by consultation.
It was a pure force fracture. All parties involved were aware that the strong wind had
caused this force fracture. An examination of the broken connection element was never-
theless initiated as a precaution. The specified material and the required tensile strength
were confirmed. Although the undershooting of the notch impact toughness was not
responsible for this damage event, the manufacturing process was nevertheless checked.
286 8 Operating Conditions Influence Area

Fig. 8.12 Crane accident during the storm Kyrill on 01/18/2007

Fig. 8.13 Broken connection element

8.1 Damage Examples—Operating Conditions 287

Fig. 8.14 Ready-to-use support roller of the hot rolling mill

Example 3—Break of a roll pin

Component: Support roller back-up roller of a heavy plate hot

rolling mill
Material: CrMo-alloyed special quenched and tempered steel
Error image: Break of the roll pin
Heat treatment: Quenching and tempering Tempering

In a heavy plate hot rolling mill, a pin break had occurred on a back-up roller. The back-
up roller had a total length of 8250 mm, a ball diameter of 2550 mm and a pin diameter
of 1525 mm, Fig. 8.14. The total weight was 240 t. It is understandable that with these
dimensions and weights, transport and disassembly were not possible, which is why this
damage had to be inspected on site. The back-up roller was made from a forged CrMo-
alloyed quenched and tempered steel, hardened by spray cooling and subsequently tem-
pered. The surface hardness was 245 HBW.
The broken back-up roller was installed in the rolling mill depending on the rolling
program and removed again after completion of the rolling lot. Until the break of the roll
pin, the back-up roller was used in a total of 10 operating periods.
288 8 Operating Conditions Influence Area

Fig. 8.15 Fracture surface on the roll pin

The fracture surface with a diameter of 1550 mm and a very large residual fracture
surface made it clear which stress peaks occurred when piercing, especially of thermo-
mechanically rolled steel grades, Fig. 8.15. The break occurred in the radius from the
transition of the pin to the ball. Noteworthy were the numerous fatigue crack initiations,
which stood out darkly from the large residual fracture surface with a firmly adhering
corrosion layer.
The rolling operation could be maintained as several replacement back-up rollers
were available. This also made the operating method possible from personal observation.
When rolling sheets at typical rolling temperatures of approx. 1200 °C, scale formations
occur on the surface. To prevent this scale from being rolled into the surface, descaling is
carried out by high-pressure water jets before entering the roll gap. In the process, steam
also gets into the radius area through the sliding bearing seals.
The radius to the roller ball was therefore not only subjected to bending stress, but
also to a corrosion attack. The resulting corrosion scars led to an increase in notch effect
and ultimately to a crack. A spear-shaped crack could be made visible on a separated
segment during the later metallographic examination, Fig. 8.16. The crack was filled
with corrosion products.
The determination of the composition of the corrosion products in the crack using
energy-dispersive X-ray analysis provided no indication of a corrosion-enhancing ele-
ment. The rust contained about 3% chromium and an increased carbon content from the
lubricants, in addition to oxygen and iron.
8.1 Damage Examples—Operating Conditions 289

Fig. 8.16 Deep seat-shaped corrosion scars in the area of crack formation (a) Spear-shaped crack
unetched; (b) Spear-shaped crack after etching in 3% Nital

A quenched and tempered steel does not have corrosion resistance, which is why it
is necessary to cover the pin radius with corrosion protection paint. The broken back-up
roller did not have this corrosion protection.
From many damage case investigations of other components, it could be inferred that
the negative influence of corrosion scars on the bending fatigue strength is underesti-
mated. With the present tensile strength level of about 800 MPa, corrosion by tap water
leads to a reduction in bending fatigue strength by 60% (see also Fig. 8.5).

Example 4—Strong roughening of the surface on a molding tool

Component: Mold for glass bottle production

Material: EN-GJL-200, Material No. 5.1300
Error image: Strong roughening of the surface after only 24 h of production use
Heat treatment: Ferritizing annealing and stress-relief annealing

Mold halves made of the unalloyed, lamellar gray cast iron EN-GJL-200, 5.1300,
according to DIN EN 1561:2012-01 [6] were produced for the production of glass bot-
tles. The surface was polished before use. Already after one day of production, the sur-
face in contact with the glass melt became rough and therefore had to be repolished,
Fig. 8.17. This process repeated itself on the following days, which is why the mold
halves were complained about at the mold manufacturer. The reason for the complaint
was quite unspecific, as a poor casting quality was generally criticized.
Upon request, it was reported that a release agent is sprayed into the mold every 30
minutes of production time to prevent the glass from sticking. There was no information
about the composition of the release agent.
The examination of the cast material described in the standard did not allow any con-
clusions to be drawn about the described error pattern. A typical ferritic microstructure
290 8 Operating Conditions Influence Area

Fig. 8.17 Surface structure of the mold halves for glass production (a) Roughened surface; (b)
Repolished surface

with lamellar graphite was present. Therefore, the surfaces were examined using energy-
dispersive X-ray analysis. In Fig. 8.18a, b, the EDX spectra of the base material and
the rough surface are compared. At various positions of the rough surface, Fig. 8.18c,
the element contents were quantified and the ranges of the elements were compiled in
Fig. 8.18d.
In addition to the elements iron, silicon, the corrosive elements sulfur and chlorine
were detected in significant amounts. The high oxygen content comes from the corro-
sion products and the high carbon content from the carrier material of the release agent.
Unfortunately, the release agent was not made available for counter measurement, but it
can be assumed with high probability that the elements sulfur and chlorine came from
the release agent and had a corrosive effect.

Example 5—Corrosion on an exhaust pipe after a short operating period

Component: Exhaust pipe

Material: X2CrNiMo17-12-2, Material No. 1.4404
Error pattern: Corrosion after a short operating period
Heat treatment: Solution annealing

In 1912, the birth of stainless steels occurred and the material X12CrNi18-8 with the
memorable material number 1.4300 was developed at the Krupp cast steel factory (later
Friedrich Krupp AG and today ThyssenKrupp AG) in Essen. The currently standardized
base steel for all stainless steels is X5CrNi18-10, 1.4301. Starting from this basic type,
many other steel types with even greater corrosion resistance have been developed. The
material used here, X2CrNiMo17-12-2, 1.4404, has increased resistance to pitting due to
its low carbon content and molybdenum content.
8.1 Damage Examples—Operating Conditions 291

Carbon: 20 - 40 %
Oxygen: 13 - 30 %
Silicon: 0.9 - 2.1 %
Sulphur: 0.3 - 2.3 %
Chlorine: 0.3 - 2.0 %
Traces of sodium, manganese
Rest iron

Fig. 8.18 EDX analysis of the roughened surface (a) EDX spectrum of the base material; (b)
EDX spectrum of the rough surface; (c) Topography of the rough surface; (d) Summary of the
analysis results

Nevertheless, after a short operating period, the exhaust pipe of a pellet low-tempera-
ture heating system was damaged by very strong corrosion, raising doubts about whether
the correct material in the correct heat treatment condition was used for the pipeline,
Fig. 8.19.
The presence of a passive layer as a prerequisite for a corrosion protection effect is
well known today. The chromium content in the matrix required for passive layer forma-
tion must not be reduced by the formation of chromium carbides. Therefore, to avoid
carbide formation, a solution annealing treatment at 1000–1050 °C with rapid cooling is
necessary. This pretreatment and the adherence to the low carbon content were doubted
in the case of damage described here and were complained about to the supplier.
The examination of the chemical composition confirmed the specified material
X2CrNiMo17-12-2, 1.4404, Table 8.1. The nickel content was slightly undershot, but it
was still within the permissible limit deviation for piece analyses.
The metallographic examination revealed severe pitting damage, Fig. 8.20. This cor-
rosion pattern alone indicated that grain boundary carbides were not responsible for the
corrosion damage, as this would have resulted in intergranular corrosion.
292 8 Operating Conditions Influence Area

Fig. 8.19 Exhaust pipe damaged by corrosion

The investigations were continued in the scanning electron microscope using energy-
dispersive X-ray analysis. No chromium carbide formation was detectable at the grain
boundaries, Fig. 8.21.
The analysis of the corrosion deposits at several positions on the inside of the pipe
revealed a chlorine content of 10–12%, Fig. 8.22. Due to the low-temperature heating,
the dew point of the exhaust gas was undercut on the long exhaust pipe, which is why the
chlorine-containing exhaust gases became moist and even with this high-alloy corrosion-
resistant steel, severe pitting damage occurred. Naturally, the question immediately arose
as to how chlorine-containing exhaust gases could be produced when burning wood pel-
lets.
It was therefore only logical to take a closer look at the wood pellets. Even a vis-
ual inspection revealed through the colors that pellets with PVC residues were used,
Fig. 8.23. EDX measurements at the colored spots confirmed the suspicion.
The entire exhaust pipe had to be replaced and from now on pellets with a certificate
of quality were ordered, which guaranteed PVC-free.

Example 6—Cracking in a Molding Tool

Component: Mold for glass bottle production

Material: EN-GJL-200, Material No. 5.1300
Error pattern: Cracking
Heat treatment: Ferritizing annealing and stress-relief annealing
 8.1

Table 8.1  Analysis measurement results and comparison with target analysis

Target analysis C Si Mn P S Cr Ni Mo Al
X2CrN- % % % % % % % % %
iMo17-12-2, Max. Max. Max. Max. Max. 16.50 10.00 2.00 –
1.4404
acc. to DIN EN 0.03 1.00 2.00 0.045 0.015 18.50 13.00 2.50 –
10088-2
Permissible devia- +0.005 +0.05 +0.005 +0.003 –
Damage Examples—Operating Conditions

±0.04 ±0.20 ±0.15 ±0.10

tion
Measurement 0.03 0.42 0.93 0.027 0.008 17.15 9.88 2.14 0.021
results
Trace elements Ti Cu W Co V Nb Sn B N
% % % % % % % % %
Measurement 0.007 0.32 0.03 0.168 0.095 0.022 0.012 0.0019 0.080
results
< Nwg. below detection limit. All information in mass- %
293
294 8 Operating Conditions Influence Area

Fig. 8.20 Pitting originating from the inside of the pipe

Spectrum Si Cr Mn Fe Ni Mo

Spectrum 3 Spectrum 1 0.75 16.4 1.19 67.8 10.5 3.30

Spectrum 1
Spectrum 2 0.86 17.7 1.29 67.9 9.8 2.38

Spectrum 2

Spectrum 4 Spectrum 3 0.69 17.4 0.98 68.6 10.7 1.46

Spectrum 4 0.56 16.6 1.10 69.6 9.3 2.72

Fig. 8.21 EDX measurements at grain boundaries and in the matrix

Spectrum

Spectrum 3 . . . . . .
Spectrum 3

Without
. . . . . .
picture

Without
. . . . . .
picture

Without
picture . . . . . .

Fig. 8.22 EDX measurements of the corrosion deposits on the inside of the pipe
8.1 Damage Examples—Operating Conditions 295

Fig. 8.23 Wood pellets with colored PVC residues

Fig. 8.24 Microstructure formation after repair welding near the first crack (a) High pearlite con-
tent near the weld seam; (b) Martensite and residual austenite in the HAZ with a hardness of 858
HV 1

Similar to the previous Example 4, molds made of the unalloyed, lamellar gray cast
iron EN-GJL-200, 5.1300, according to DIN EN 1561 [6] were used here. Already on the
first day of use, cracks had occurred in several molds, which were repaired by welding.
The repair had to be carried out quickly to maintain production. However, after another
crack, it was decided to complain about the molds.
The mold manufacturer stated that to reduce the temperature gradient and thus the
thermal stresses, the molds must be preheated to 300–400 °C before use in the glass
machine, which was explicitly communicated in the delivery documents.
296 8 Operating Conditions Influence Area

The investigations to describe the casting quality provided no clues to the cause of
the damage. The microstructure and hardness were typical of the component and free of
complaints. The evaluation of the welding repair of the first crack carried out by the glass
bottle manufacturer, on the other hand, provided important clues to the cracking.
During welding, some of the graphite was dissolved in the melt and due to the rapid
cooling, large amounts of pearlite were formed, Fig. 8.24a. In the heat-affected zone,
there was even martensite formation with residual austenite, Fig. 8.24b. The hardness in
this hardening structure was determined to be 858 HV 1.
This high hardness was to be expected with the described repair welding. However,
this hardness value proved that the prescribed preheating before use of the molds was
omitted. After preheating at 300–400 °C, the hardness should have been significantly
lower.
Thus, it was determined that the stress-reducing preheating was not carried out, which
greatly increased the thermal stresses and ultimately led to cracking. Formally, this only
clarified the probable cause of the second crack, but it can be assumed that the first crack
also had the same cause, as the molds had already failed on the first day of use due to
cracking.

Example 7—Pitted Damage on a Chain Link

Component: Round link chain

Material: X6CrNiTi17-12-2, Material No. 1.4571
Error pattern: Pitted damage
Heat treatment: Solution annealing

Not only the components treated in a heat treatment furnace can show defects or lead to
damage. The heat treatment conditions can damage components of the furnace system
itself. In Fig. 8.25, a round link chain from a protective gas furnace with pitted damage

Fig. 8.25 Round link chain

with pitted damage
8.1 Damage Examples—Operating Conditions 297

is documented. This error pattern was discovered during a plant inspection and although
the cross-sectional reduction was still small, the chain was replaced as a precaution.
The material for the round link chain was specified as steel X6CrNiMoTi17-12-2,
material no. 1.4571, which was also confirmed by the OES analysis. Although it is not a
classic heat-resistant steel, this material is frequently used in industrial furnace construc-
tion up to temperatures of 800–900 °C. The titanium stabilization allows safe welding.
The chain was not directly exposed to the high temperature in the furnace chamber,
but was used as a suspension for an intermediate door. However, it was surrounded by
the endothermic carburizing atmosphere. The exact temperature and atmospheric condi-
tions on the chain were not known. However, the error pattern indicated a typical tem-
perature of 550–650 °C. When chromium-nickel steels are exposed to the carburizing
atmosphere in this temperature range, the carbon activity reaches values of aC ≥1. The
formation of chromium carbides leads to high-temperature corrosion, which is referred
to as “metal dusting”. [7] The surface attack on the round link chain was locally formed
as pitting. However, there can also be a widespread attack, as the wall breakthrough of
a annealing muffle in Fig. 8.26 shows.
The surface decay is caused by strong carbide formation and subsequent oxidation,
Fig. 8.27. There are practically no countermeasures, which is why damaged components

Fig. 8.26 Annealing muffle destroyed by “metal dusting” made of material X1NiCrMoCu25-20-5
298 8 Operating Conditions Influence Area

Fig. 8.27 Typical manifestations of surface damage due to “metal dusting”

must be replaced after a relatively short operating period. Nickel-based alloys are more
resistant to “metal dusting”, but they are very expensive.

Example 8—Crack formation on the surface of a pipe connector

Component: Pipe connector

Material: 42CrMo4, Material No. 1.7225
Error pattern: Crack formation on the surface
Heat treatment: Nitriding after quenching and tempering Quenching and nitriding

Pipe connectors can be used to couple pipes and thus extend them to any length. In the
case described here, abrasive material flowed through the pipe, which is why the pipe
connector was made of a heat-treatable steel and nitrided after quenching and tempering.
During assembly, massive crack formation occurred on the surface, Fig. 8.28. Even from
a macroscopic view, it was noticeable that the two openings had different diameters.

Fig. 8.28 Pipe connector with cracked surface (a) Cracked surface; (b) Differences in diameter
8.1 Damage Examples—Operating Conditions 299

Fig. 8.29 Differently pronounced network of cracks on the surface (a) Pronounced network of
cracks; (b) Minor crack formation

Fig. 8.30 Microstructural formation after quenching and tempering and nitriding (a) Quenched
and tempered microstructure in the base material; (b) Compound layer with pores

Upon further examination, it was noticed that not only a single massive crack had
occurred, but a network of cracks was present on the surface at the opening side with the
larger diameter, Fig. 8.29.
On the opposite opening of the pipe connector, the surface was free of cracks. In this
area, all drawing specifications were initially checked.
The material composition corresponded to the standard specification and was incon-
spicuous. The metallographic examination showed a flawless quenched and tempered
microstructure and a uniformly formed compound layer, Fig. 8.30. The pore seam in
the compound layer with a proportion of about 40% corresponded to the expected result
after gas nitriding. The nitriding hardness depth of NHD350 = 0.35 mm was determined
according to DIN 50190-3:1979-03 using a hardness gradient curve, Fig. 8.31. All
microstructural formations and hardness results were in accordance with the drawing,
appropriate, and fully met the target specifications.
300 8 Operating Conditions Influence Area

700

600

NHD350 0.35 mm
500
Hardness [HV 0.5]

400

300

200

100

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Edge distance [mm]

Fig. 8.31 Determination of nitriding hardness depth according to DIN 50190-3:1979-03

Fig. 8.32 Crack formations up to the end of the diffusion layer (a) Crack formation on the outer
shell; (b) Crack formation on the inner surface

The externally visible cracks were examined metallographically for their extent, Fig.
8.32. As expected, numerous cracks were present, running perpendicular to the surface
and all within the diffusion zone. The cracks were present both on the outer shell and on
the inner surface.
From the investigation results, it was concluded that there must have been a deliberate
or unconscious plastic deformation during assembly. However, due to the high surface
hardness, crack-free deformation was not possible. The mechanical screwing in of the
8.1 Damage Examples—Operating Conditions 301

Fig. 8.33 New hardening layer in the area of the internal thread

pipe to be connected resulted in a misalignment. The friction forces that occurred were
so great that a new hardening layer with a hardness of almost 900 HV 0.2 was formed
and the shell surface was torn open, Fig. 8.33. At the same time, this load led to crack
formation in large areas of the surface.

Example 9—Premature failure of a rolling bearing

Component: Rolling bearing

Material: Various rolling bearing steels according to ISO 683-17:2015-02
Error pattern: Premature failure
Heat treatment: Martensitic hardening and tempering or bainitizing

The large number of rolling bearing damages is summarized here. All bearings failed
significantly before the calculated service life and the causes are only partially known so
far. Bearing components with fatigue damage often show pittings and possibly also “but-
terflies” at the end of the calculated service life, Fig. 8.34.
Bearing components with a usage corresponding to the calculated service life rarely
come for damage investigation, as there is no damage in the actual sense and the calcu-
lated service life is often even exceeded. In contrast, damages caused by “White Etching
Cracks” or “White Etching Areas” are much more often the subject of extensive damage
case investigations. This is also due to the fact that there is still no uniform picture of the
causing factors and these damages occur after a fraction (5–15%) of the calculated ser-
vice life.
The damage pattern is characterized by a microstructural transformation occurring
beneath the tread, which, unlike the basic microstructure, does not etch and is therefore
recognizable as a white microstructural component, Fig. 8.35. In these “White Etching
Areas”, cracks form as the stress continues, which develop up to the tread surface and
eventually lead to breakouts.
302 8 Operating Conditions Influence Area

Fig. 8.34 Damage to bearing components after classic fatigue (a) Pitting formation on the run-
ning surface; (b) “Butterfly” on a manganese sulfide

Fig. 8.35 ”White Etching Areas” and “White Etching Cracks” on a bearing inner ring

In Fig. 8.36a, initial, locally limited crack formations were discovered during a mag-
netic powder inspection. Isolated breakouts, also locally limited, represent the next stage
of damage, Fig. 8.36b. Partial and complete damage to the raceway characterize the fail-
ure of a bearing, Fig. 8.36c, d.
A comparable error formation was present in both martensitic and bainitic hardened
bearings. The burnishing of bearings did not prevent this error pattern either.
Although rolling bearings from wind turbines are already highly affected due to the
large number of installations, this damage pattern also occurred in many other areas of
application.
From laboratory and test stand experiments, it is known that after loading bearing
inner rings with hydrogen, this error pattern can be reproducibly generated, [8]. It is
unclear whether and by which reactions hydrogen can be absorbed from the lubricant
8.1 Damage Examples—Operating Conditions 303

Fig. 8.36 Different stages of damage (a) First signs of cracks in the MT test; (b) first breakout
points; (c) Advanced damage; (d) Damage to the entire raceway

under operating conditions. Extensive investigations with lubricants and lubricant addi-
tives are currently being carried out for this purpose.
In another series of tests on a bearing test stand, structural stresses were artificially
introduced into the bearing seat by waviness. These tests also led to “White Etching
Cracks”. A comprehensive literature evaluation on the damage mechanism “White Etch-
ing Cracks” can be found in [9].

Example 10—Irregularities on the flanks of a spur gear shaft

Component: Spur gear shaft

Material: 18CrNiMo7-6, Material No. 1.6587
Error pattern: Irregularities on various tooth flank surfaces
Heat treatment: Nitriding after quenching and Quenching and nitriding
tempering
304 8 Operating Conditions Influence Area

Fig. 8.37 Line-like surface irregularity on several tooth flanks

On a spur gear shaft, line-like surface irregularities were detected after a short operating
period (a few days), the origin of which was initially unknown, which is why an inves-
tigation was initiated, Fig. 8.37. The spur gear shaft was made from the case-hardening
steel 18CrNiMo7-6, 1.6587, and had been case-hardened.
A topography profile was created in the area of the linear anomalies using a laser
scanning microscope, Fig. 8.38. The surface in the area of the anomalies showed numer-
ous small and finely distributed craters. The maximum height difference was approxi-
mately 7.2 μm. On the intact surface, machining-induced grooves and partially smeared
surface structures could be seen.
A detailed overview of the surface topography was provided by the scanning electron
microscopic examination, Fig. 8.39. The surface showed numerous melting craters and
melt beads. The surface damage was therefore caused by a current flow, which resulted
in electrical discharges with local melting. The high flash temperatures during an electri-
cal discharge also alter the lubricant. This leads to oil cracking products, which cause the
lubricant to age faster.
The melting was also clearly visible in the metallographic examination, Fig. 8.40. A
new hardening layer about 10 μm thick had formed on the surface. The case hardening
8.1 Damage Examples—Operating Conditions 305

Fig. 8.38 Three-dimensional topography profile and documentation of a line profile

layer was perfectly formed with fine acicular martensite and minor residual austenite,
Fig. 8.41.

Example 11—Crack formation on a ship crankshaft after short operating time

Component: Ship crankshaft

Material: 42CrMo4, Material No. 1.7225
Error pattern: Crack formation after only 22000 operating hours
Heat treatment: Tempering

The ship crankshaft submitted for damage analysis had failed after only 22000 operat-
ing hours due to crack formation on a main bearing journal and the adjacent crankpin.
Starting from the worn surface, a fatigue fracture had formed, Fig. 8.42. The crankshaft
306 8 Operating Conditions Influence Area

Fig. 8.39 Melting craters and melt beads in the area of the linear anomaly

was made of the heat-treatable steel 42CrMo4, 1.7225, and had been quenched and tem-
pered. The bearing journals were subsequently induction hardened.
During the customer-side disassembly, it was already determined that the connect-
ing rod bearings showed no signs of wear. In contrast, the two main bearings 2 and 3
were completely destroyed. The engine block and the bearing caps were also damaged
and showed signs of high temperature stress due to the blue temper color. Burnt oil was
noticeable on the bearing surface.
In the overview shot, the strong friction marks on the bearing journal surface were
noticeable. The adjacent bearing journal also showed distinctive signs of wear, Fig. 8.43.
For the metallographic examination, a microsection was taken from the heavily worn
bearing journal area. On the surface, a friction martensite layer about 125 μm thick had
formed as a result of insufficient lubrication. Numerous cracks had also formed on this
surface, away from the starting point of the fatigue fracture surface, Fig. 8.44.
The hardness profile perpendicular to the bearing journal surface could be divided
into the following areas, see Fig. 8.45.

1 = New hardening area,

2 = Thermally damaged inductive hardening layer,
3 = Hardness of the base material.
8.1 Damage Examples—Operating Conditions 307

Fig. 8.40 Documentation of the intact surface of the load flank

Fig. 8.41 New hardening layer in the area of damage caused by current flow

The description of the base material was also omitted here, as the material showed no
defects and was not responsible for the damage pattern. There was a severe lack of lubri-
cation at both bearing positions.
308 8 Operating Conditions Influence Area

Fig. 8.42 Documentation of the crankshaft, the crack position, and the crack surface

Fig. 8.43 Wear marks on the adjacent bearing journal

8.1 Damage Examples—Operating Conditions 309

Fig. 8.44 New hardening layers on the bearing journal surface with intensive crack formation

1 2 3
900

800

700

600
Hardness [HV 1]

500

400

300

200

100

0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
Edge distance [mm]

Fig. 8.45 Hardness profile perpendicular to the bearing journal surface 1 = New hardening layer;
2 = Thermally altered inductive hardening layer 3 = Hardness of the base material
310 8 Operating Conditions Influence Area

Fig. 8.46 Damage to the raceways (a) Inner ring; (b) Outer ring

Example 12—Damage to the raceway of a ball bearing

Component: Ball bearing

Material: 100CrMnSi6-4, Material No. 1.3520
Error pattern: Damage to the raceways
Heat treatment: Hardening and tempering

During an inspection after 3500 operating hours, comparable damage to the raceways of
both the inner and outer rings was found on a ball bearing, Fig. 8.46. The damage was
present at two positions offset by 180°. The balls also showed similar damage.
The damages were axially oriented, but they also showed an expansion in the cir-
cumferential direction. The individual friction traces were present next to each other at
a dense, very regular distance. The burnishing layer was completely rubbed off in these
scratch marks.
In the area of the friction traces, new hardening layers had formed on all three bear-
ing components, Fig. 8.47, 8.48 and 8.49. In places, there were also breakouts. The low
annealed martensitic hardening microstructure was without abnormalities. The carbide
distribution was very uniform.
The damage pattern was due to standstill markings (False Brinelling). This damage
occurs when systems are exposed to irregular vibrations at a standstill over a longer
period of time and there is no longer sufficient lubrication. Standstill markings not only
occur on bearings, but are also observed on tooth flanks. In the advanced stage, this dam-
age mechanism leads to surface disruption up to a total failure [10].
8.1 Damage Examples—Operating Conditions 311

Fig. 8.47 New hardening layers on the outer ring in the area of the friction traces

Fig. 8.48 New hardening layers on the inner ring in the area of the friction traces

Fig. 8.49 New hardening layers on the ball surface in the area of the friction traces
312 8 Operating Conditions Influence Area

References

1. Amareller, S.: Stahl und Eisen 84 (1964) S. 926/31

2. Gräfen, H.: Lexikon Werkstofftechnik, VDI-Verlag 1991
3. DIN 15402:1982-11: Lasthaken für Hebezeuge. Doppelhaken Rohteile
4. DIN 15400:1990-06: Lasthaken für Hebezeuge. Mechanische Eigenschaften, Werkstoffe,
Tragfähigkeiten und vorhandene Spannungen
5. Pohl, E.J.: The Face of metallic Fractures. Published by the Münchener Rückversicherungs-
Gesellschaft, München 1964
6. DIN EN 1561:2012-01: Gießereiwesen—Gusseisen mit Lamellengraphit
7. Mehner, A.; Hoja, T.; Hoffmann, F.; Zoch, H.-W.; Kleingries, M.; Ackermann, H.; Köhne, H.;
Lucka, K.: Metal dusting of high temperature Cr-Ni steels and Ni-based alloys. HTM J. Heat
Treatm. Mat. 65 (2010) 1 S. 30–36
8. Vegter, R.H.; Slycke J.T.: The Role of Hydrogen on Rolling Contact Fatigue Response of
Rolling Element Bearings. Journal of ASTM international Vol 7. No. 2, S. 1–12
9. Vegter, R.H.; Stadler, K.; Vaes, D.: White Etching Cracks—Lagerrisse. https://www.elektro.
net/93376/white-etching-cracks-lagerrisse
10. Deters, L.; Schadow, C.:False Brinelling-Wälzlagerschäden und ihre Ursachen VDI-Fachkon-
ferenz Schadensmechanismen an Lagern. Aachen 2014