64 COLD CRACKING IN WELDMENTS OF STEEL S 690 QT
COLD CRACKING IN WELDMENTS
OF STEEL S 690 QT
B. Alexandrov1, K. Theis2, M. Streitenberger3, H. Herold3, I. Martinek3
1Technical University of Sofia (Bulgaria), currently visiting The Ohio State University (USA)
2 Techniques for Testing and Measurement, Irxleben (Germany)
3 Otto-von-Guericke-University, Magdeburg (Germany)
E-mail: 1alexandrov.1@osu.edu
2pmt-ok@t-online.de
3ifst@uni-magdeburg.de
ABSTRACT
The cold cracking process in shielded metal arc and gas metal arc welding of steel S 690 QT is investigated by Tekken
test. The conditions for cold cracking are varied by changing the specimens’ thickness, heat input, initial weld metal
hydrogen concentration and preheating temperature. The kinetics of initiation and propagation of cracking is moni-
tored by acoustic emission equipment. It is found out that the crack’s propagation path shifts from weld metal to fusion
line and HAZ with decreasing initial weld metal hydrogen concentration and increasing hardness of HAZ. In the
investigated range of cold cracking conditions the magnitude of cracking is mainly controlled by the preheating tem-
perature and heat input. The acoustic emission signal provides valuable information about the factors controlling the
kinetics of crack initiation and propagation and the intensity of cold cracking. The initial hydrogen concentration in
weld metal controls the incubation period’s duration and affects the cracking intensity at higher heat inputs. The pre-
heating temperature does not influence the incubation period, but significantly affects the kinetics of crack propa-
gation. The heat input has a complex influence on the cracking kinetics and intensity. This is related to the contra-
dictive effects of heat input on the initial hydrogen concentration per unit length of weld metal and on the behaviour
of hydrogen during cooling. The obtained results provide a basis for combined experimental – modelling investiga-
tions on the cold cracking phenomenon in weldments of higher strength structural steels, aiming at quantitative eval-
uation of the influence of the main controlling factors.
IIW-Thesaurus keywords: GMA welding; Arc welding; Gas shielded arc welding; MMA welding; Arc welding; Cold
cracking; Cracking; Defects; Microstructure; Heat affected zone; Weld zone; Weld metal; Hardness tests; Mechanical
tests; Acoustic emission; Practical investigations.
1 INTRODUCTION joints of a higher strength structural steel at the lower
yield strength level, where weld metal cracking was
Despite the significant advance in the development of expected to occur. For that purpose quenched and tem-
high strength structural steels with improved weldability pered steel of grade S 690 QT was selected together
and welding consumables with lower hydrogen potential, with matching filler materials for shielded metal arc and
cold cracking is still a major problem in the welding pro- gas metal arc welding. The experiments were performed
duction. For many years cold cracking has been consid- by Tekken test in order to clarify the conditions for alter-
ered a typical heat-affected zone (HAZ) defect in the joints native HAZ/weld metal cracking. Acoustic emission
of conventional structural steels. The implementation of equipment was implemented to register the acoustic sig-
the new higher strength thermo-mechanically treated and nals, generated by the cracking events. This technique
quenched and tempered steels with reduced carbon provides unique opportunities for monitoring the kinet-
equivalent has shifted cold cracking to the weld metal, ics and for evaluating the intensity of the cracking
whose matching strength is achieved by higher level of process [5, 6, 7].
alloying [1, 2]. The yield strength ranges above 680 MPa
and 800 MPa are reported as typical for weld metal hydro-
gen assisted cracking in multipass welds [3, 4]. 2 EXPERIMENTAL PROCEDURES
The aim of the present research was to investigate the
conditions for hydrogen assisted cold cracking in welded 2.1 Investigated materials
The chemical composition and mechanical properties of
Doc. IIW-1673-04 (ex-doc. IX-2115-04) recommended the investigated steel S 690 QT plates, covered elec-
for publication by Commission IX “Behaviour of metals trode for SMAW (EN 757: E69 4 Mn2NiCrMo B T 42 H5,
subjected to welding”. AWS A 5.5: E11018-M) and wire for GMAW (EN 12534:
Welding in the World, Vol. 49, n° 5/6, 2005
COLD CRACKING IN WELDMENTS OF STEEL S 690 QT 65
Table 1 – Chemical composition of investigated materials, %
Material C Si Mn P S Al Cu Ni Cr Mo V Nb
S 690 QT 10 mm 0.220 0.298 1.14 0.0180 0.0547 0.0592 0.0145 0.0284 0.0387 0.126 0.0486 < 0.005
S 690 QT 20 mm 0.200 0.302 1.13 0.0157 0.0059 0.0631 0.0101 0.0557 0.0862 0.128 0.0486 < 0.005
E69 4 Mn2NiCrMo
B T 42 H5 0.05 0.30 1.70 2.00 0.40 0.40
G3CrNi1Mo 0.09 0.50 1.62 1.4 0.30 0.27 0.1
Table 2 – Mechanical properties of investigated materials
Material ReH Rm A5 KV
MPa MPa % J
S 690 QT 10 mm > 840 873 > 16,7 > 148 (– 40 °C)
S 690 QT 20 mm > 802 868 > 16,8 > 86 (– 40 °C)
E69 4 Mn2NiCrMo B T 42 H5 > 730 830 > 18 > 47 (– 40 °C)
G3CrNi1Mo > 690 790 > 16 > 47 (– 40 °C)
G3CrNi1Mo, AWS A5.28: ER100-SG) are given in false acoustic signals during the slag self-detachment.
Table 1 and Table 2, correspondingly. The acoustic emission examinations started within 5 min-
utes after welding and continued until the completion of
cracking. The acoustic signals generated by the cold
2.2 Testing procedure cracking process were recorded by two channel
analysing system for acoustic emissions of the type
The Tekken Test experiments were performed accord-
LOCAN 320 (Physical Acoustic corporation, USA).
ing to EN ISO 17642-2:2003. Plates with thickness of
10 and 20 mm were welded by both the SMAW and The aim of the measurement was to ascertain the suit-
GMAW processes, using the respective filler materials. ability of the acoustic emission signal as a proof of the
The welding parameters, the drying conditions for the cold cracking events and to determine optimal testing
covered electrodes and the preheat temperatures are parameters, device adjustments and criteria for evalua-
given in Table 3. The SMAW process was performed tion and coordination. The records were performed at the
by mechanized electrode feeding and travel, and the following test conditions: frequency 150 kHz; pre-ampli-
GMAW process by mechanized travel, thus providing fying 40 dB; main amplifying 52 dB; peak detection time
for constant travel velocity and heat input along the 200 s; hit definition time 400 s; hit lock-out time
welds. Carbon dioxide was used as a shielding gas in 2 000 s.
the GMAW process.
The first uncomplicated evaluation of measurement
results was done by conventional representation of hits
2.3 Acoustic emission examinations versus time. In this, all acoustic emission signals regis-
tered during the time of measurement are included in the
Immediately after welding the specimens were carefully measurement results, i.e. also all kinds of noise signals
cleaned from the slag in order to avoid generation of (process noise, electrical noise etc.) and signals from dif-
Table 3 – Experimental parameters and conditions
Thickness, Heat input*, Current, Voltage, Velocity, Drying of
Test No mm kJ/mm A V mm/s electrode Preheat
10/1.0/N/RT 10 1.03 140 23 2.5 Non dried RT
10/1.0/D/RT 10 1.03 140 23 2.5 2 h/350 °C RT
10/1.0/N/RT 10 1.03 140 23 2.5 Non dried RT
10/1.0/D/RT 10 1.03 140 23 2.5 1 h/350 °C RT
10/1.7/N/RT 10 1.7 180 25 2.12 Non dried RT
10/1.7/D/RT 10 1.7 180 25 2.12 1 h/350 °C RT
20/1.7/D/RT 20 1.7 180 25 2.12 1 h/350 °C RT
20/2.0/D/RT 20 2.0 190 26 1.97 1 h/350 °C RT
20/2.0/D/70 20 2.0 190 26 1.97 1 h/350 °C 70 °C
20/2.0/D/50 20 2.0 190 26 1.97 1 h/350 °C 50 °C
20/1.0/MAG**/RT 20 1.053 280 29 6.17 – RT
20/1.3/MAG**/RT 20 1.299 280 29 5 – RT
* Calculated with efficiency factor 0.8.
** GMAW, shielding gas carbon dioxide.
66 COLD CRACKING IN WELDMENTS OF STEEL S 690 QT
ferent sources of acoustic emission (micro- and macro- LTotal (mm) is the total length of all cracks in the respec-
cracks, crack periphery friction, and other). tive test cross-section.
By aimed selection of equipment parameters (frequency, The cracks lengths were measured by the image sys-
boosting, time threshold values for signal registration) as tem Quantimed Q570 using the linear vertical length for
well as selection and evaluation of different acoustic H and HC as well as the overall length of each crack for
emission measurement values (amplitude, energy, time measuring LHAZ, LFL and LTotal on all the five cross sec-
of signal increase, signal time and other), measurement tions. Whereas Cs characterizes the proportion of the
results can be co-ordinated to different sources of cracked weld cross section, the HAZ and fusion-line
acoustic emission as crack signals, crack periphery fric- cracking ratio quantifies the crack path location.
tion, noise, and other.
2.5 Hardness measurements
2.4 Cracking estimations and microstructure evaluation
The estimations on the magnitude of cold cracking were The hardness in HAZ and weld metal was measured by
performed in accordance with EN ISO 17642-2:2003. the Vickers method with load of 1 kg (HV1). The mea-
The cold cracks identification and measurements were surements were performed over the middle (3rd) cross
done by light microscope at magnification x50 or higher sections of each tested specimen along a line starting
in five cross-sections for each tested specimen. The from the base metal at X-side of the groove (BMX), pass-
section crack ratio (CS, %) was calculated using the ing the weld metal (WM) and reaching the base metal
measuring procedure presented in Figure 1 and formula at Y-side of the groove (BMY).
(1). The HAZ cracking ratio (CHAZ , %) and the fusion The HAZ microstructure was classified by microstruc-
line cracking ratio (CFL, %) were calculated by formulae tural constituents as ferrite, bainite and martensite. The
(2) and (3), correspondingly: weld metal microstructure was classified according to
HC the IIW procedure for light microscope examination of
CS = 100, % (1)
H ferritic steel weld metal [8, 9].
LHAZ
CHAZ = 100, % (2)
LTotal
LFL 3 RESULTS AND DISCUSSION
CFL = 100, % (3)
LTotal
where 3.1 Tekken test
H is the minimum thickness (mm) of the test bead, The Tekken Test results (% CS, % CHAZ, % CFL and LTotal)
HC is the height (mm) of the root crack shown in are summarized on Table 4, along with the cumulative
Figure 1, energy of acoustic signal, deposited metal cross sec-
tion area, crack initiation sites, hardness and microcon-
LHAZ (mm) is the total length of HAZ cracks,
stituents’ content in weld metal. Typical examples of
LFL (mm) is the total length of fusion line cracks, crack initiation and propagation are shown in Figure 2
to Figure 5.
All the specimens with thickness of 10 mm cracked com-
pletely throughout the weldments’ thickness and length
(tests No. 1-6, Table 4). In the cases of higher hydrogen
concentration (the specimens welded with non-dried
electrodes) cracking occurred predominantly into the
weld metal (see Figure 2). Welding with dried electrodes
resulted in fusion line cracking of about 19 to 39 % of
the total crack length, followed by weld metal cracking
(see Figure 3). The proportion of fusion line cracking
was the highest in the case of shorter drying time and
lower heat input (test No. 4), where the later was asso-
ciated with higher hardness in HAZ than in weld metal.
The two 20 mm non-preheated shielded metal arc
welded specimens (tests No. 7 and 8) cracked com-
pletely through the weld metal with some fusion line
cracking in the case of the lower heat input (see
Figure 4). Preheating of 50 oC resulted in 81 % section
crack ratio with 7 % fusion line cracking (test No. 10).
The preheating temperature of 70 oC, combined with
heat input of 2 kJ/mm and dried electrodes, turned out
to be the threshold for avoiding cold cracking in the
Figure 1 – Crack measuring procedure 20 mm Tekken test specimens (test No. 9).
COLD CRACKING IN WELDMENTS OF STEEL S 690 QT 67
Table 4 – Tests results
Test No. CS % CFL/CHAZ LTotal. Cumulative A* Crack Hardness HV1* Microconstituents
% mm acoustic mm2 initiation* in WM*, %
energy BMX HAZX WM HAZY BMY AF FS PF
(1) 10/1.0/N/RT 118 0 28.94 – 28.6 WMY** 277 294 322 293 253 19 64 17
(2) 10/1.0/D/RT 123 19/0 23.18 – 27 FLX 279 301 328 297 274 14 69 17
(3) 10/1.0/N/RT 107 0 31.94 17.23 106 32.8 WMY** 272 286 316 279 268 11 72 17
(4) 10/1.0/D/RT 107 39/0 33.43 17.96 106 33.1 FLX** 279 340 319 304 282 17 73 10
(5) 10/1.7/N/RT 111 3/0 43.15 113.48 106 42.2 WMY 234 248 283 252 268 37 36 27
(6) 10/1.7/D/RT 102 25/0 34.97 68.65 10 6 47.5 FLX 272 277 296 267 273 46 32 22
(7) 20/1.7/D/RT 105 2/0 33.93 37.78 106 41.3 WMY 288 353 348 312 279 7 82 11
(8) 20/2.0/D/RT 103 0 37.44 85.15 10 6 45.3 WMY 279 313 322 330 274 15 70 15
(9) 20/2.0/D/70 0 – 0 0.24 106 41.9 – 285 296 303 275 284 39 32 29
(10) 20/2.0/D/50 81 7/0 16.11 37.85 106 41.8 FLX** 269 294 320 273 288 26 51 23
(11) 20/1.0/MAG/RT 90 9/50 49.12 1.59 106 45.8 FLX 282 421 294 398 282 48 30 22
(12) 20/1.3/MAG/RT 98 0/57 47.48 6.61 10 6 58.2 FLX 283 372 263 315 283 58 21 21
* Deposited metal cross-section area, crack initiation site, hardness and microconstituents are determined at the middle (3rd) cross sections of
the tested specimens.
** Crack initiation site varies along the five cross sections of the specimen.
Figure 4 – Typical WMY crack initiation
Figure 2 – Typical WMY crack initiation and WM propagation in the cases of welding
and WM propagation in the cases of welding with dried electrodes, no preheat and higher
with non-dried electrodes, no preheat and lower restraint (Tests No. 7, 8)
restraint (Tests No. 1, 3, 5)
Figure 3 – Typical FLX crack initiation
and FL/WM propagation in the cases of welding Figure 5 – Typical FLX crack initiation
with dried electrodes, no preheat and lower and FL/WM propagation in the cases of GMAW
restraint (Tests No. 2, 4, 6) (Tests No. 11, 12)
The GMAW process was applied at two heat inputs with- Typical examples of crack initiation and propagation are
out preheating, resulting in cold cracking about CS of 90 shown in Figures 2, 3, 4 and 5. In the cases of pre-
to 98 % and HAZ crack proportion of nearly 60 % (see dominant weld metal cracking (tests No. 1, 3, 5, 7, 8) the
Figure 5, tests No. 11 and 12). The later is related to preferred crack initiation site is the weld metal at Y-side
the very high HAZ hardness (370-420 HV1). of the root, which is the side with higher stress con-
68 COLD CRACKING IN WELDMENTS OF STEEL S 690 QT
centration. For the lower restraint level (10 mm thick Depending on the welding conditions, the weld metal
specimens) weld metal cracking is typical for higher consists of varying amounts of PF, AF and FS (see
hydrogen concentrations (non dried electrodes) and Figures 6, 7 and 10). The HAZ in the SMA welded spec-
higher hardness of weld metal compared to the HAZ. For imens is predominantly bainitic (see Figures 8 and 9),
the higher restraint level (20 mm thick specimens) pre- while in the case of GMAW it contains significant
dominant weld metal cracking occurs in non preheated amounts of martensite. It can be seen from Table 4 that
specimens where weld metal hardness is almost equal the microstructure of weld metal is controlled by the
to or higher than HAZ hardness. plate thickness, heat input and preheating. Reducing
the plate thickness and increasing the heat input and
In the cases of considerable fusion line and HAZ crack-
preheating results in lower content of FS and respec-
ing the crack initiation normally occurs at the fusion line
tively higher amounts of AF and PF in weld metal. This
at X-side of the root. Due to the specific groove shape,
relation is clearly reflected by the weld metal hardness.
HAZ in that area is subjected to local overheating
accompanied by significant grain growth and hardness
The alternating crack initiation between the weld metal
increase (see Table 4). Crack initiation in this highly sus-
and HAZ in the 10 mm specimens is obviously controlled
ceptible to cracking microstructure requires less hydro-
by both the hydrogen content and microstructure. The
gen concentration than in the cases of dried electrodes
weld crack propagation at the low hydrogen levels can
for the 10 mm thickness (test No. 2, 4, 6) and of GMAW
be attributed to the higher weld metal hardness com-
process (test No 11 and 12).
pared to HAZ. Although the higher HAZ hardness in the
The microstructures of HAZ and weld metal for some non-preheated 20 mm specimens (tests No. 7 and 8),
typical combinations of thickness, heat input and preheat the cracking initiated in weld metal. Preheating resulted
temperature are shown in Figure 6 trough Figure 11. in more significant hardness reduction in HAZ (tests
Figure 6 – Weld metal of test 5: 10 mm specimen Figure 8 – HAZ of test 5: 10 mm specimen non-
non-preheated, 1.7 kJ/mm SMAW, AF 37 %, preheated, 1.7 kJ/mm SMAW, bainite
FS 36 %, PF 27 %
Figure 7 – Weld metal of test 8: 20 mm specimen Figure 9 – HAZ of test 8: 20 mm specimen
non-preheated, 2.0 kJ/mm SMAW, AF 15 %, non-preheated, 2.0 kJ/mm SMAW, bainite
FS 70 %, PF 15 % and martensite
COLD CRACKING IN WELDMENTS OF STEEL S 690 QT 69
Figure 10 – Weld metal of test 11: 20 mm Figure 11 – HAZ of test 11: 20 mm specimen
specimen non-preheated, 1.0 kJ/mm GMAW, non-preheated, 1.0 kJ/mm GMAW, bainite
AF 48 %, FS 30 %, PF 22 % and martensite
No. 9 and 10) and, respectively, in a crack-free speci- cumulative energy of the acoustic signal is directly
men and crack initiation at the fusion line. Like in the related to the cracking intensity. It was evaluated both
10 mm specimens, the crack initiation in the 20 mm by the conventional parameter energy versus time and
specimens is also controlled by the hydrogen – the rather crack-classifying parameters as hits versus
microstructure relation. time and amplitude versus time with respect to the fre-
Obviously the weld metal is more susceptible to crack quency distribution of hits versus amplitude and ampli-
initiation at higher hydrogen concentrations and higher tude versus energy.
levels of restraint and stress concentration (WMY initia- The conventional parameter energy versus time was
tion). At lower hydrogen levels HAZ becomes more sus- considered as most suitable, among the obtained
ceptible than weld metal even being softer and sub- acoustic emission parameters, for representing the inten-
jected to lower stress concentration (FLX initiation). sity of cracking activities. The crack area was not mea-
The low heat input in the case of GMAW resulted in sured during the present investigations, but it could be
very high HAZ hardness and considerable fusion line seen from Table 4 and Figure 16 that the total energy
cracking. of the acoustic signal is in reasonable correlation with
the total crack length of the shielded arc welded non-pre-
heated specimens (tests No. 3 to 8).
3.2 Acoustic emission results
The change of cumulative energy in time was used to
The results of acoustic emission investigations are sum- characterise the kinetics of crack nucleation and prop-
marised in Table 4 and in Figures 12 to Figure 15. The agation. The initial concentration of hydrogen in weld
Figure 12 – Influence of hydrogen concentration and heat input on intensity of cracking at lower level
of restraint in SMAW specimens
70 COLD CRACKING IN WELDMENTS OF STEEL S 690 QT
Figure 13 – Influence of heat input and preheating on intensity of cracking at higher level of restraint
in SMAW specimens
Figure 14 – Influence of level of restraint on intensity of cracking in SMAW specimens
Figure 15 – Influence of heat input on intensity of cracking in GMAW specimens
COLD CRACKING IN WELDMENTS OF STEEL S 690 QT 71
Figure 16 – Correlation of the cumulative energy of the acoustic signal to the total crack length
in the SMA – welded non-preheated specimens
metal has significant influence on crack nucleation and microstructure and hydrogen effusion from the weld on
propagation. The lower hydrogen concentration prolongs the one hand and the higher initial hydrogen concen-
the incubation periods and delays the stage of stable tration on the other hand. The balance of these two
crack growth, considerably reducing the cracking inten- effects determines the final weld hydrogen concentration
sity at higher heat input (see Figure 12). at room temperature and thus controls the intensity of
cracking.
By controlling both the hydrogen concentration in the
temperature range of cracking and the microstructure The plate thickness also has significant influence on the
of HAZ and weld metal (see Table 4), the applied low kinetics of cracking (see Figure 14). At equal other con-
temperature preheating mainly influences the kinetics ditions the harder microstructure and the higher levels
and intensity of fracture during stable growth stage (see of restraint and stress concentration (WMY initiation) in
Figure 13 and Table 4). Preheating of 50 oC slows down the thicker specimen (test No. 7) caused considerably
the crack propagation and reduces the cumulative faster crack nucleation and propagation than in the thin-
acoustic energy more than twice compared to the non- ner one (test No. 6).
preheated specimen. The critical preheating tempera- More detailed information on the cracking process in
ture for avoiding cold cracking at the given combination these two specimens is presented in Figure 17 and
of restraint, initial weld metal hydrogen concentration Figure 18. The change of the acoustic signal’s parame-
and chemical composition of the weld and base metal ters (amplitude, energy and number of hits) in time
is 70 oC. The recorded low level acoustic activity in this clearly shows that main cracking activity in the thicker
specimen is most probably due to non proper cleaning specimen is completed in less than 1 000 seconds, while
of slag in weld root, stress relaxation and deformation in the thinner one it spreads over 6 000 seconds with
during cooling the specimen to room temperature. some interruption of about 1 000 s in the middle. The
comparison of the distribution of hits by amplitude in the
Increasing the heat input had an expected effect of pro-
two tests shows that the cracking process in test No. 7
longing the incubation period, which was related to the
generated less hits with equal amplitude than in test
reduced crack susceptibility of weld metal and HAZ
No. 6. The distribution of amplitude by energy in test
microstructure and to the increased time for attaining
No. 7 is more scattered. The faster cracking, the lower
critical hydrogen concentration in the area of cracking.
energy acoustic activity and the more scattered acoustic
On the other hand the increase of heat input resulted in
parameters in the thicker specimen are indicators of a
an increase in the cumulative acoustic energy of two to
brittle fracture process.
six times (Table 4, Figures 12, 13 and 15). This unex-
pected effect might be related to the contradictive influ-
ence of heat input on the hydrogen behaviour. The
higher heat input results in bigger cross-section of the 4 CONCLUSIONS
deposited metal (Table 4) and correspondingly in higher
initial hydrogen content per unit weld length. At the same The performed experimental investigations on weld-
time it facilitates the effusion of hydrogen from weld ments of steel S 690 QT by Tekken test have clearly
metal to the ambient atmosphere and through the HAZ. shown that within the covered range of welding condi-
The total effect of heat input is obviously a question of tions cold cracking may occur in both the weld metal
balance between the effects of slower cooling rate on the and the HAZ. Due to the specific combination of micro-
72 COLD CRACKING IN WELDMENTS OF STEEL S 690 QT
Figure 17 – Acoustic emission activity of test No. 6 (10 mm, non-preheated, 1.7 kJ/mm, dried electrode)
Figure 18 – Acoustic emission activity of test No. 7 (20 mm, non-preheated, 1.7 kJ/mm, dried electrode)
constituents in weld microstructure, the weld metal crack metal hydrogen and stress concentration. The fusion
initiation requires a higher level of initial weld hydrogen line crack initiation is followed by weld metal propaga-
concentration and a higher stress concentration at the tion in the case of SMAW and by HAZ and weld metal
cracking initiation site, typically occurring at the Y-side propagation in the case of GMAW.
of the Tekken test groove. The fusion line crack initia- The initial hydrogen concentration in weld metal con-
tion is typical for the X-side of the groove where coarse- trols the incubation period’s duration and affects the
grained susceptible to cold cracking microstructure is cracking intensity at higher heat inputs. The preheating
developed in HAZ, requiring lower levels of initial weld temperature does not influence the incubation period,
COLD CRACKING IN WELDMENTS OF STEEL S 690 QT 73
but significantly affects the kinetics and intensity of crack- Service (DAAD) in the framework of the “Research Stays
ing. The heat input has a complex influence over the and Study Visits for University Academics and Scientists”
cold cracking process. Increasing the heat input simul- Programme.
taneously results in longer incubation periods and in a
higher cracking intensity. The first effect is related to the
well known effects of heat input on the microstructure, REFERENCES
on the time for reaching critical hydrogen concentration
and on the value of that concentration. We consider that [1] Widgery D.J.: High strength weld metals – Routes for
the unexpected second effect could be attributed to the development, IIW Doc. II-1459-02.
influence of heat input on the initial hydrogen concen-
[2] Hart P.H.M.: Resistance to hydrogen cracking in steel
tration per unit length of weld metal.
weld metals, Welding Journal, January 1986, pp. 14-22.
The acoustic emission provides valuable information [3] Nevasmaa P.: Hydrogen induced cracking in high-
about the cold cracking process. It has been shown that strength multipass weld metal - predicting the cracking risk
the energy of the acoustic signal is directly related to the and necessary precautions for safe welding, IIW Doc. IX-
kinetics of crack initiation and propagation and to the 2066-03.
intensity of cracking. The acoustic signal parameters as [4] Åström H. et al.: Consumables for welding high strength
hits, amplitude, and energy can be potentially used for steels, Development and metallurgical problem areas, IIW
more precise investigation of the characteristics and Doc. II-1460-02.
stages of the cracking process.
[5] Savage W.F. et al.: Hydrogen induced cracking during
The obtained results provide a basis for combined exper- implant testing of alloy steel, Welding Journal, 1976, vol. 55,
imental – modelling investigations on the cold cracking No 12, pp. 400-407.
phenomenon in weldments of higher strength structural [6] Vasudevan R. et al.: Hydrogen – Assisted Cracking in
steels, aiming at quantitatively evaluating the influence HSLA Pipe Line Steels, Welding Journal, 1981, vol. 60,
of the main controlling factors. No 9, pp. 155-168.
[7] Kikuta Y., et al.: Acoustic emission detection of crack
propagation modes in hydrogen delayed cracking, IIW Doc.
ACKNOWLEDGMENTS IX-837-73.
[8] Guide to the light microscope examination of ferritic
The presented research has been conducted at the steel weld metal, IIW Doc. IX-1533-88.
Institute of Joining and Beam Technology, Otto von [9] Abson D.J.: Quantitative assessment of low alloy steel
Guericke University, Magdeburg, Germany, and has weld metals using the light microscope – Results of an
been sponsored by the German Academic Exchange international point counting exercise, IIW Doc. IX-1687-92.
