See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/275386718

Leak-Before-Break: What Does It Really Mean?

Article  in  Journal of Pressure Vessel Technology, Transactions of the ASME · August 2000

DOI: 10.1115/1.556183

CITATIONS READS
37 9,115

1 author:

G. Wilkowski
Engineering Mechanics Corporation of Columbus
252 PUBLICATIONS   1,884 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Improved model for predicting thermal aging on toughness of CASS. View project

PRCI Project MAT-8-3 View project

All content following this page was uploaded by G. Wilkowski on 15 September 2021.

The user has requested enhancement of the downloaded file.

 Leak-Before-Break
What Does It Really Mean?
Leak-before-break (LBB) is a term that has been used for decades in reference to
Gery Wilkowski a methodology that means that a leak will be discovered prior to a fracture
President occurring in service. LBB has been applied to missile casings, gas and oil
Engineering Mechanics pipelines, pressure vessels, nuclear piping, etc.,. LBB also has several technical
Corporation of Columbus definitions. For instance, LBB can occur for an axial flaw in a pipe where the
3518 Riverside Dr. – Suite 202 penetration of the wall thickness will result in a stable axial through-wall crack.
Columbus, OH 43221 This is LBB under load-controlled conditions. LBB could also occur for a
[email protected] circumferential crack in a pipe with high thermal expansion stresses. This might
be LBB under compliant displacement-controlled conditions. Finally, LBB might
occur when the flaw is stable under normal operating conditions and remains
stable when there is a sudden dynamic event (i.e., seismic loading). This might
be a time-dependant inertial LBB analysis. These analyses are deterministic,
and could be extended to probabilistic evaluations as well. This paper describes
some of the technical LBB approaches, applications, and significance of the
methodology used in the applications.

Introduction In 1965, Kobayashi[2] made an improvement to the

This paper summarizes: Irwin LBB model by making an improvement to the
• The history of LBB, which originated from axial surface flaw stress intensity factor expression, see
flawed cylinder applications, Figure 2.
• LBB applications in different industries, or assuming R/B>>1, and for short flaw lengths MF=1,
• LBB analyses under different loading conditions,
and
• Recent considerations for a new LBB Regulatory
Guide for the USNRC.

History of LBB
Perhaps one of the earliest technical approaches for
leak-before-break (LBB) was one published by
Irwin [1] in 1961. This was for the application of an
axial flaw in a pressure vessel using linear elastic
fracture mechanics for missile applications. LBB was then
postulated to occur if the length of the flaw was less
than twice the thickness of the cylinder, see Figure 1.

Figure 2 LEFM LBB model with Kobayashi

surface flaw improvement

As can be seen in the equations in Figure 2, there are

three factors not described in the sketch. These are R,
MK, and MF. The MK term is the Kobayashi stress
intensity magnification factor accounting for the
proximity of front free surface. The MF term is a
bulging stress magnification factor from Folias for
axial through-wall flaws[3]. Finally R is the pipe
radius. The terms KIc and KIct refer to the toughness in
Figure 1 LEFM LBB approach by Irwin[1] the axial and radial directions, respectively. It can be
seen how the upper equation in Figure 2 simplifies to
the lower equation. Hence this expression
From LEFM analysis, Irwin showed that the crack- incorporated free surface effects, bulging effects, and
driving force would be greater in the radial direction toughness differences in the through-wall crack versus
than in the axial direction as long as the axial crack surface crack growth directions.
length was less than twice the cylinder thickness. A different approach for LBB was also being pursued
at that time. This was the Pellini Failure Analysis

1
Diagram or FAD[4], see Figure 3. [Note, the FAD
(failure assessment diagram) is a similar sounding
term used in the R6 type analyses[5], but is quite
different from the Pellini approach.] Pellini and co-
workers developed many terms to describe fracture
behavior as a function of a reference temperature. For
instance, they coined the terms:
• NDT = nil ductility temperature
• CAT = crack arrest temperature
• FTE = fracture transition elastic
• FTP = Fracture transition plastic

The Pellini approach was primarily developed for

ship applications, but he also suggested applying it to
piping flaws as well. One of the difficulties in
applying the Pellini approach to LBB was that there
was no way to include bulging effects such as
determined by Folias.

Top - Service failure at NDT

Middle - arrested hydrostatic burst (FTE + 20F)
Bottom - Pneumatic burst at FTP
Figure 4 Examples of failure modes at
Pellini reference temperatures
2 2
πKc /(8cσf ) = ln{sec[πMT σh/(2σf)]}

Using the empirical relation 12Cv/Ac=Kc2/E with Cv

Figure 3 Pellini Failure Analysis Diagram
being the Charpy upper shelf energy in ft-lb, and Ac
being the cross-sectional area of the Charpy specimen
(0.124 in2 for a full-size specimen) the above equation
The Pellini method, however, was quite useful in
yields
differentiating between catastrophic brittle fracture
versus ductile tearing, see Figure 4. 2
12πCvE/(8cσf Ac) = ln{sec[πMT σh/(2σf)]}
In the 1960’s, Battelle had started to develop LBB
methodologies for axial flaws in gas pipelines, that As the toughness approaches infinity, then the above
was also extended to nuclear piping, LNG piping, and equation simplifies to the following limit-load
chemical plant piping[6]. This methodology compared solution.
the failure pressures of axial through-wall flaws to MT σh = σf
axial surface flaws. where,
MT = Axial TWC bulging factor from
The Battelle methodology involved non-linear Folias, Erdogan, Sanders, etc.
fracture mechanics analysis by including a Dugdale
2
plastic zone, and defining plane stress toughness, Kc, MT = Kshell/Kplate = f(c /Rt)
empirically by relating it to the Charpy upper shelf
energy. This worked well, as long as the failure mode 2
was ductile tearing from the surface or through-wall MT = [1+ 1.61(c /Rt)]0.5
flaws. The relationship for an axial through-wall
crack (TWC) is as follows. The relationship for axial surface cracks (SC) was
2 2
πKc /(8cσf ) = ln{sec[πMpσh/(2σf)]}

2
 or, using the same empirical Charpy energy relation • Gas cylinder applications (steel and
aluminum) - chemical and automotive [15]
2
12πCvE/(8cσf Ac) = ln{sec[πMpσh/(2σf)]} • Composite pipe - natural gas and oil
pipelines
As the toughness approaches infinity, then the above • Offshore pipe laying[16] - natural gas and oil
equation simplifies to the below limit-load solution pipelines

Mpσh = σf Different Industries and Applications

LBB procedures and analyses can change from
Where, Mp = Axial surface flaw bulging factor industry to industry based on different levels of risk
Mp = [1-a/(tMT )]/(1-a/t), Maxey empirical and types of loading that might occur for the different
industrial applications. Two very different approaches
surface-crack bulging factor
are summarized below.

In the natural gas pipeline industry, there are very

long-distance transmission pipelines in remote areas.
Typically these are buried pipes where axial flaws are
more of a concern since longitudinal stresses are
frequently compressive. Leakage will occur at
Failure Break normal operating service loads (pressures) and there
stress/ are seldom any pressure excursions above the
flow TWC curve
operating pressures. For such remote pipelines, a
stress large leakage could be tolerated from risk viewpoint.
Leak Hence a tolerable leakage might be up to 30-percent of
the cross-sectional opening under normal operating
pressures.

In the nuclear industry, LBB has been applied to

piping for the purpose of eliminating equipment that is
used for restraining pipe whipping from a postulated
2c/(Rt)0.5 pipe rupture[17]. The concern in this application is
Figure 5 LBB analysis from Battelle relationships with above ground plant-piping systems where
using limit-load equations circumferential flaws are historically more prevalent
than axial flaws. In this LBB approach, it is desirable
The Battelle axial flaw equations have been to detect small amounts of leakage at normal operating
implemented into many standards, i.e., nuclear piping conditions so that the leakage size flaw (with some
applications in Appendix C and H of Section XI of the safety factor) would be stable at transient (typically
seismic) stresses. Hence, the flaw orientation in these
ASME Boiler and Pressure Vessel code [7], and ANSI
B31G for corrosion flaws in gas transmission piping[8]. analyses is circumferential, and pressure stresses as
well as many other stress component contribute to the
LBB analysis. Those stresses need to consider normal
Many other subsequent LBB studies were conducted operating stresses for leakage detection, and transient
after the Battelle work in 1960’s to 1970’s. Some of stresses for crack stability analysis. It is also essential
these were1: that there not be any subcritical crack growth
• MPA-Stuttgart Phenomenological Burst mechanism that could cause long surface flaw to
Test Programs – nuclear piping[9] occur. Such long surface flaws could lead to failure
• Reactor pressure vessels (ORNL)[10] - under the transient loads without any leakage warning.
nuclear If there were a mechanism that could cause long
• Steam generator tubes – nuclear[11] surface flaws, then one would have to invoke an
• CANDU zirconium pressure tubes[12] - augmented inspection process for LBB to work. For
nuclear example, ultrasonic testing might be deemed adequate
• Zirconium cladding over uranium fuel tubes to making sure the flaw lengths would be less than a
- nuclear desired value.
• NASA and aerospace applications to cabin
chambers – aerospace[13] LBB under Different Loading Conditions
• Cryogenic pipeline applications [14] - refinery In many structural applications, there can be a
& LNG variety of sources of stresses or loads. Crack stability
• Storage tanks - refinery & LNG behavior can depend significantly on the type of
loading. The major categories are;

1
References are given for some cases, the author can be
contaced for additional information.

3
 • Load-controlled stresses
• Pure displacement-controlled stresses
• Displacement-controlled stresses in
structures with significant compliance so
there is still a large amount of stored elastic
energy, and
• Time-dependant stresses.

These aspects are differentiated below.

Load-Controlled LBB
Examples of pure load-controlled stresses occur from
pressure or dead-weight loads. In the case of pressure
loads, if the fluid can decompress quickly (i.e., water
at ambient temperature), then even the load-controlled
pressure stresses may behave like a displacement-
controlled stress. A gas-pressurized line might be a Figure 7 Circumferential blunt flaw LBB criterion[13]
good example of a true load-controlled stress.
Examples of failures from these cases are illustrated in Displacement-Controlled LBB (Local secondary
the middle versus bottom pictures in Figure 4. stresses)
Displacement-controlled stresses could be classified
as local displacement-controlled stresses (i.e., weld
Dead-weight loads are typically considered as true
residual stresses or through-thickness temperature
load-controlled stresses. Pipe hangers or other
gradients), or global displacement controlled stresses
supports, however, may limit dead-weight loads
where the compliance of the structure can add to
physically. In such cases, the pipe may experience
unstable behavior (i.e., restraint of thermal expansion
load-controlled stress until the displacements reach a
stresses in a pipe system). Local displacement-
limiting value.
controlled stresses, which involves energy stored over
a short gage section, seldom contribute to fracture
An axial flaw leak-before-break analysis was shown
unless the material is extremely brittle. Weld residual
in Figure 5. Similarly, circumferential surface flaws
stresses may be important for subcritical cracking (i.e.,
may also behave in a LBB manner under load-
stress-corrosion cracking) and for crack-opening
controlled stresses. This is illustrated in Figures 6 and
displacement under elastic loading for leakage
7, which were developed to assess the maximum girth
considerations, but seldom contribute to fracture for
weld repair that could be made on an offshore lay
ductile materials used in most pressure vessel and
barge [16]. In this case, the girth weld may have to be
piping applications.
repaired at a location on the barge where the dead-
weight bending loads on the pipe are significant.
Displacement-Controlled LBB (Global secondary
stresses)
Global displacement-controlled stresses can
contribute to ductile fracture even of extremely tough
materials such as TP304 stainless steel. To address
this aspect, the J-integral/tearing modulus
methodology was developed. The J/T analysis
approach is illustrated in Figure 8. Figure 9 shows a
circumferentially through-wall cracked TP304 pipe
experiment that was used to validate this analysis
procedure. In this experiment, the helical spring
represented the stored elastic energy from thermal
expansion stresses for a pipe length of 28 feet[18].

The J/T stability analysis procedure predicts when

instability might start. From experimental evidence, it
is possible for the crack to jump only a small length in
certain conditions. To assess such displacement-
controlled instability conditions, an energy-balance
Figure 6 Experiment with 70-percent deep was developed in Reference 19. Figure 10 shows test
circumferential groove in a pipe under bending results of two identical surface-flawed pipe
which exhibited load-controlled LBB behavior experiments with different pipe lengths. The top case
shows a limited instability (smaller load drop) than the
bottom case.

4
 Limited instability

displacement
Figure 8 J/T stability analysis procedure

Spring to
provide higher
compliance
Figure 10 Experimental records of limited
instability and total instability of
circumferential surface flaws in TP304 pipe

Test pipe
with
circumferential
TWC

Figure 9 Circumferential through-wall-cracked

TP304 pipe experiment under compliant
displacement-controlled loading to validate J/T
analysis [15]

Combined Load-Controlled and Displacement-

Controlled LBB
In addition to allowing for the estimation of how far
Figure 11 Energy Balance Analysis [15] showing
a crack jump might be, the energy-balance analysis
how surface to through-wall crack transition is
method also allows for several other key aspects. The
predicted, as well as stability under combined load-
transition of a surface crack to a through-wall crack,
controlled and displacement-controlled stresses
and the magnitude of unstable through-wall crack
growth can be assessed by the energy-balance
Time -Dependant Stress LBB
approach, see Figure 11. Additionally, it is possible to
Although Figure 11 shows a schematic of conducting
assess the stability for combined load-controlled a combined load-controlled and displacement-
stresses and displacement-controlled stresses, see controlled stresses, real piping system stresses for a
Figure 11. In Figure 11, the resulting crack would
nuclear plant are more involved. Including the effects
jump to point “I”. The crack would be completely
of time-dependant stress components like inertial and
unstable if the load-controlled stresses were equal to seismic anchor motion stresses in all the previously
P 1, but would be stable if the load-controlled stress mentioned analyses is typically done by assuming that
was P2.
those stress components do not vary with time.

5
 A more detail analysis procedure to account for time-
dependant stress variations was established in
Reference 20. In this analysis procedure, a special
cracked-pipe element was used to represent the global
moment-rotation behavior due to the crack, see Figure
12. This element can be adjusted to account for
constant pressure axial forces, and then a dynamic
pipe analysis can be conducted. A significant
experimental effort was undertaken as part of a
program called the International Piping Integrity
Research Group (IPIRG) program to assess
circumferentially cracked-pipe systems under seismic
loading at LWR temperatures[21]. Figure 13 is a
schematic of a 406-mm (16-inch) diameter pipe
system. Figure 14 shows a comparison of
experimental and predicted moment versus time
behavior. From such dynamic analysis, the effects of
the crack plasticity on damping and changing the
system response can be determined. In typical nuclear
pipe LBB analyses, a response-spectrum analysis is
used to determine the seismic loads. This is an
uncracked pipe elastic stress analysis. These elastic Figure 13 Illustration of pipe system used in
stresses are then used with elastic-plastic fracture IPIRG program to assess crack stability under
mechanics, which gives a conservative estimation of seismic loading at LWR temperatures
the actual crack-driving force. This dynamic non-
linear analysis procedure allows this inherent
conservatism to be quantified.

Figure 14 Comparison of cracked-pipe element

analysis and pipe system experimental data from
the IPIRG program[21]

Recent Considerations for a New LBB

Regulatory Guide for the USNRC
The U.S. Nuclear Regulatory commission has had a
Draft Standard Review Plan (3.6.3) in place since
1986[17]. The analysis capabilities and knowledge of
material property behavior under cyclic dynamic
loading corresponding to seismic loads has increased
since then[22]. Hence, the NRC plans to create a
Regulatory Guide for LBB that would repla ce the
Figure 12 Schematic of how to create a cracked
Draft Standard Review Plan. The regulatory guide
pipe element for dynamic LBB analysis
would be more specific than the standard review plan
on how the LBB analysis should be conducted.

6
 Draft Standard Review Plan 3.6.3 for LBB has the From the numerous studies conducted on LBB, we
following features. have learned many additional factors that were never
• The procedure uses a flaw tolerance-based initially envisioned. As with many things in life, we
analysis. have found out that the more we know, then we realize
• It is applied to analyzable sections of Class 1 how little we really knew in the beginning.
piping.
• It requires demonstration of adequate margin References
between a postulated leakage-size flaw under [1] Irwin, G. R., “Fracture of Pressure Vessels,”
normal operational loads, and a critical flaw size Materials for Missiles and Spacecraft, pp. 204-
under design-basis loads (load-controlled LBB). 229, McGraw-Hill, 1963.
• To demonstrate that the piping system is a [2] Kobayashi, A. S., Zii, M. and Hall, L. R.,
candidate for LBB approval it is necessary to show “Approximate Stress Intensity Factor for an
the following: Embedded Elliptical Crack Near Two Parallel
- There is no active degradation mechanisms Free Surfaces,” International Journal of Fracture
which would undermine LBB assumptions, i.e., Mechanics, Vol. 1, 1965, pp. 81-95.
long surface flaws could not occur, and [3] Folias, E. S., “The Stresses in a Cylindrical Shell
- There are no atypical loading conditions (e.g., Containing an Axial Crack,” ARL 64-174,
water hammer). Aerospace Research Laboratories, October 1964.
• With the above conditions satisfied, one can [4] Pellini, W. S., Evolution of Engineering
determine the smallest through-wall flaw (with Principles for Fracture-Safe Design of Steel
some margin), which can be detected by facility’s Structures, Naval Research Laboratory Report
leakage-detection system. NRL 6957, September 23, 1969.
• One needs to demonstrate by fracture analyses that [5] Milne, I., and others, “Assessment of the
the critical flaw (at design-basis loads) is larger Integrity of Structures Containing Defects,”
than leakage flaw by a specified margin. CEGB Report R/H/R6 - Revision 3, 1986.
[6] Kiefner, J. F., and others, “Failure Stress Levels
For the future LBB Regulatory Guide, a three-tiered of Flaws in Pressurized Cylinders,” Progress in
analysis concept is being developed. The following Flaw Growth and Fracture Toughness Testing,
levels are being established: ASTM STP 536, American Society for Testing
and Materials, pp. 461-481, 1973.
• Level 1 analysis - Simpler than current [7] ASME Boiler and Pressure Vessel Code, Section
methodology (with potentially larger margins) that XI, Rules for In-service Inspection of Nuclear
would accept piping systems which easily meet Power Plant Components, Article IWB-3600,
current requirements (i.e., most main coolant 1995 Edition, July 1995.
loops). [8] American National Standards Institute
(ANSI)/American Society of Mechanical
• Level 2 analysis - Slightly more complex than the Engineers (ASME) B31G “Manual for
current DSRP methodology, i.e., include effects Determining the Remaining Strength of Corroded
from pressure-induced bending, weld residual Pipelines,” 1984.
stresses, and dynamic strain aging[22]. [9] Kussmaul, K., and others, “Phänomenologische
Behälterberstversuche - Phase I,” BMFT Report
• Level 3 analysis - Significantly more complex than BMFT-TB-1500 279, by MPA-Stuttgart, July
current methodology, but allowing for nonlinear 1985.
seismic stress/fracture time-history analyses[20]. [10] Bryan, R. H., Bolt, S. E., Merkle, J. G., and
Whitman, “Quick-Look Report on Test of
Summary and Conclusions Intermediate Vessel V-8a – Tearing Behavior of
The concept of LBB has been around for over 40 Lower Upper,” ORNL/SST-4, Oakridge National
years, perhaps even longer. In the beginning, LBB Laboratory, August 1982.
analysis was limited to axial cracks in pipe and [11] Proceedings of IAEA Specialist meeting on
pressure vessels using linear elastic analysis Theoretical and Experimental Work on LMFBR
procedures. The methodology has now evolved to Steam Generator Integrity and Reliability with a
dynamic elastic plastic analysis with different stress Particular Reference to Leak Development and
sources occurring all at the same time. Probabilistic Detection, IAEA Document IWGFR/50,
approaches have also been established[23, 24] based on November 1983.
the many different deterministic models. [12] Moan, G. D., Coleman, C. E., Price, E. G.,
Rodgers, D. K., and S. Sagat, “Leak-Before-
These results and the many different industrial Break in the Pressure Tubes of CANDU
applications show that LBB is a very generic term that Reactors,” International Journal of Pressure
can be adjusted to fit the risk level of comfort for that Vessels and Piping, Vol. 43, 1990, pp. 1-21.
application/industry.

7
 [13] Pierce, W., “Effects of Surface and Through
Cracks on Failure of Pressurized Thin-Walled
Cylinders of 2014-T6 Aluminum,” NASA Report
TN D6099, November 1970.
[14] Eiber, R. J., Maxey, W. A., and Duffy A. R.,
“Fracture Investigation of Pipe for LNG
Service,” Battelle report to A.G.A. on Project
PR-3-42, December 1971.
[15] Rana, M. D., ‘Experimental Verification of
Fracture Toughness Requirements for Leak-
Before-Break Performance for 155-175 ksi
Strength Level Gas Cylinders,” Journal of
Pressure Vessel Technology, Volume 109,
November 1987, pp. 435-439.
[16] G. M. Wilkowski and R. J. Eiber, “Evaluation of
Tensile Failure of Girth Weld Repair Grooves in
Pipe Subjected to Offshore Laying Stresses,”
ASME Journal of Energy Technology, March,
1981, pp 48-57.
[17] Solicitation for Public Comment on Standard
Review Plan 3.6.3 “Leak-Before-Break
Evaluation Procedures,” Federal Register, Vol.
52, No. 167, pp 32633, Friday, August 28, 1987.
[18] G. M. Wilkowski, A. Zahoor, and M. F.
Kanninen, “A Plastic Fracture Mechanics
Prediction of Fracture Instability in a
Circumferentially Cracked Pipe in Bending - Part
II, Experimental Verification on a Type 304
Stainless Steel Pipe,” in Journal of Pressure
Vessel Technology, Vol. 103, No. 4, Nov. 1981,
pp 359-365.
[19] G. M. Wilkowski and G. Kramer, “An Energy
Balance Approach to Estimate the Initiation and
Arrest of Ductile Fracture Instability in
Circumferentially Cracked Pipe,” ASME Special
Technical Publication, Vol. 167, July 1989, pp
103-114.
[20] Olson, R., Wolterman, R., Scott, P.,
Krishnaswamy, P., and Wilkowski, G., “The
Next Generation Methodology for Cracked Pipe
System Subjected to Dynamic Loads,” ASME
PVP Vol. 275-1, June 1994, pp 159-172.
[21] Hopper, A. T., Wilkowski, G. M., Scott, P. M.,
Olson, R. O., Rudland, D., Kilinski, T., Mohan,
R., Ghadiali, N., and Paul, D., “The Second
International Piping Integrity Research Group
(IPIRG-2) Program - Final Report,” NUREG/CR-
6452, February 1997.
[22] Wilkowski, G. M., Olson, R. J., and Scott, P. M.,
“State-of-the-Art Report on Piping Fracture
Mechanics,” U.S. Nuclear Regulatory
Commission report NUREG/CR-6540, BMI-
2196, February 1998.
[23] Harris, D. O., Lim, E. Y., and Dedhia, D. D.,
“Probability of Pipe Fracture in the Primary
Coolant loop of a PWR Plant,” NUREG/CR-
2189, Vol. 5, August 1981.
[24] Rahman, S., and others, “Probabilistic Pipe
Fracture Evaluations for Leak-Rate-Detection
Applications,” NUREG/CR-6004, April 1995.

View publication stats