Professional Documents
Culture Documents
4, 2009
In this investigation, an attempt has been made to study the influence of welding consumables on the factors
that influence cold cracking of armour grade quenched and tempered (Q&T) steel welds. Flux cored arc welding
(FCAW) process were used making welds using austenitic stainless steel (ASS) and low hydrogen ferritic steel
(LHF) consumables. The diffusible hydrogen levels in the weld metal of the ASS and LHF consumables were
determined by mercury method. Residual stresses were evaluated using X-ray stress analyzer and implant
test was carried out to study the cold cracking of the welds. Results indicate that ASS welds offer a greater
resistance to cold cracking of armour grade Q&T steel welds.
KEY WORDS: Quenched and tempered steel; Flux cored arc welding process; Weld cold crack-
ing; Implant testing
Fig. 2 Schematic diagrams of implant test system: (a) dimensions in millimeters, (b) test block (base plate),
(c) implant specimen (dimension in mm)
[100/(M2 − M1 )] (1)
where DH is volume of diffusible hydrogen in
mL/100 g of deposited weld metal at NTP (0◦ C and
760 mmHg); Vg is volume of gas in burette in ml after
72 h; B is barometric pressure (mmHg); TR is room
temperature (30◦ C) when Vg is measured; H is head
of mercury in mm when Vg is measured; M1 is mass
of the sample in g before deposit of the weld metal;
and M2 is mass of the sample in g after removal from
hydrogen meter.
Five trials were carried out for each consumable
and the measured diffusible hydrogen values are pre- Fig. 3 Implant test results of stress vs time to failure
sented in Table 4. plot with the corresponding LCS values. Hori-
zontal arrows indicate the specimen did not fail
2.3 Implant testing
The implant test was conducted using an implant located in the HAZ. The sample was subjected to the
testing machine (ISHA 27105, India) as per the guide- desired stress under constant load within 5 min post-
lines given in literature [10, 11] and with modifica- welding. This arrangement enabled the coarse grained
tions in base plate dimensions as detailed in litera- HAZ of the specimen to experience the load. The
ture [12]. An implant test system (Fig. 2) was em- time required for the implant specimen to fail under
ployed to evaluate the susceptibility of the mater- each stress was noted and a plot of stress-time was
ial to hydrogen induced cracking. In this test sys- obtained. Three specimens were tested at each stress
tem, a helical threaded specimen (fabricated from the level and the average of time to failure was used for
base metal) was embedded in a single bead weld in plotting stress-time curve (Fig. 3). From this plot,
a 14 mm-thicked base plate. A single pass weld was lower critical stress (LCS) below which no failure oc-
deposited so that a portion of the notch section was curs, was obtained as presented in Table 5.
J. Mater. Sci. Technol., Vol.25 No.4, 2009 519
Table 6 Microhardness (Hv) variation across the weld (0.5 kg load)
Weld type Weld Region close to the Weld/HAZ HAZ region Base
region weld/HAZ interface in the interface boundary adjacent to metal
weld region side (fusion boundary) fusion boundary
Flux cored arc welded 245 320 405 425 455
Austenitic stainless steel weld (FA)
Flux croed arc welded low 294 380 415 430 455
hydrogen ferritic steel weld (FF)
2.4 Microstructure and hardness of metal deposited while the FF weld metal recorded
2.95 ml/100 g of metal deposited. The diffusible hy-
The implant specimen that did not fail after 72 h drogen level did not show larger variations in the
was subjected to microstructural examination to re- above weld metals. However, the FA weld showed
veal the presence of microcracks. Few implant speci- relatively lower level of diffusible hydrogen level than
mens were interrupted after 1000 min of loading and FF weld.
were subjected to microstructural examination to re-
veal the crack path. The microstructure analysis of 3.2 Lower critical stress
the weldments was carried out using an optical mi-
croscope (OM, ML7100, MEIJI, Japan). The spec- It has been reported that there were microcracks
imens were etched with 2% nital reagent to reveal in those specimens that did not fail under implant
the microstructure of the weld region of LHF weld, conditions even after 72 h[12] . In this study, the im-
base metal and HAZ regions. Aquaregia reagent was plant specimen that did not fail after 72 h was sub-
used to reveal the microstructure of the ASS weld jected to metallographic examination to reveal the
region. A Vickers0 s microhardness testing machine presence of microcracks. From Fig. 4, it is inferred
(HMV-T1, Shimadzu, Japan) was employed for mea- that no microcracks were found in the interface of all
suring the hardness in the weld metal region, fusion the welds. Hence, LCS was taken as the stress below
boundary, the region close to the fusion boundary on which no microcracks were present and also the high-
weld metal side and HAZ region. ASTM E 384-05a est stress at which the fracture did not occur (after
guidelines were followed for measuring the microhard- 72 h). From Table 5 and Fig. 3, it could be inferred
ness and the values are presented in Table 6. The frac- that the welds made by ASS consumable showed a
tured surface of the implant specimen was analyzed by higher LCS values than the LHF welds.
scanning electron microscopy ( SEM, 5610LV, JEOL,
Japan) at higher magnification. 3.3 Optical micrographs
Fig. 4 Optical micrographs of weld /HAZ interface of unfailed (after 72 h) implant specimen revealing non-
existence of microcracks. GPB: grain boundary phase; UTM: untempered martensite
tion are presented in Table 6. The hardness of the un- weld metal side (cracked region) are 320 and 380 Hv
wedded base metal is 455 Hv. The FA weld exhibits for FA and FF welds, respectively. The hardness of
a hardness of 245 Hv in the weld metal region, while the fusion boundary of FA weld is 405 Hv while the
FF weld recorded 294 Hv. Similarly, the hardness FF welds exhibit a hardness value of 415 Hv. The
in the region adjacent to the fusion boundary in the hardness in the HAZ region of the FA and FF welds
J. Mater. Sci. Technol., Vol.25 No.4, 2009 521
Fig. 6 Optical micrographs of weld/HAZ interface of interrupted (at 1000 min): (a) FA weld, (b) FF weld,
implant specimen revealing existence of cracks in the fusion boundary. GPB: grain boundary phase; TM:
untempered martensite
are found to be 425 and 430 Hv, respectively. Thus other hand, quasi-cleavage type of fracture is featured
the FA welds made using ASS consumables have a in the fractured surface of the FF implant specimen
lower hardness in the weld region, close to the fu- indicating brittle fracture (Fig. 7(b)). The fracture
sion boundary (where cracks are found) region, fusion surface analysis indicates that a higher energy frac-
boundary and the HAZ region than FF welds made ture occurred in FA welds under implant conditions
using LHF consumables. than FF welds.
The mechanical properties of the base metal and The transverse residual stresses measured from the
both weld metals are presented in Table 2. It is in- weld center are displayed in Fig. 8. From this figure,
ferred that the base metal has 1200 MPa of yield it is inferred that the transverse residual stresses of
strength and 1290 MPa of ultimate tensile strength. FA and FF welds are 310 and 360 MPa, respectively
The yield strength of the FA and FF weld metals is in the weld center. Similarly the transverse residual
565 and 680 MPa, respectively. The ultimate tensile stress at the fusion boundary for FA and FF welds are
strength of the FA and FF weld metals is 600 and 460 and 490 MPa, respectively. Moreover, FA welds
760 MPa, respectively. Thus, the FF weld exhibits imparted a lower transverse residual stress in HAZ
higher strength than FA welds. region. Thus a lower transverse residual stress is fea-
tured in FA welds than FF joints across the weld.
3.6 Fractured surface
4. Discussion
The fractured surface of the failed implant speci-
men is displayed in Fig. 7. An intergranular fracture is Hydrogen is carried to the atmosphere by shielding
featured in FA implant specimen (Fig. 7(a)). On the gas, or surface contamination in Q&T steels welds.
522 J. Mater. Sci. Technol., Vol.25 No.4, 2009
HAZ is raised to a sufficiently high temperature to in all cases and also the implant specimens failed in
produce a coarse grain region. This high tempera- the fusion boundary invariably in all cases.
ture region, because of its coarse grain, is not only The hardness and microstructure in fusion zone
more hardenable but also less ductile than the re- (i.e. grain boundary phase in ASS welds and region
gions further away from fusion boundary. It is the of untempered martensite in LHF welds) are the most
region, where the greatest risk of cracking exists. In influencing factors that contributes for greater LCS
case of quenched and tempered steels, the harder the values and higher resistance to cold cracking. The
microstructure, the greater the risk of cracking. Soft hardness in the fusion zone is inversely proportional
structure can tolerate more hydrogen than hard mi- to the critical stress (LCS)[7] . The lower the hard-
crostructure before cracking occurs[16] . ness in the fusion zone, the higher the LCS, and the
In case of similar welds, the microstructures of the greater the resistance to cold cracking. Thus, lower fu-
weld metals and HAZ will be unique and essentially sion zone hardness of the FA joints is one of the major
the zone adjacent to the fusion boundary is a charac- contributing factors for higher LCS values compared
teristic feature of the base metal and weld metal. But with that of FF joints. Thus, FA joint has a greater
this is not true in dissimilar welds. The fusion bound- LCS values (470 MPa) and it offers greater resistance
ary microstructure in dissimilar welds often possesses to cold cracking.
some unique features. Normal epitaxial nucleation
during solidification along the fusion boundary gives 4.4 Effect of HAZ microstructure
rise to grain boundaries that are continuous from the
base metal into weld metal across the fusion bound- Several factors affect the susceptibility of the ma-
ary. These boundaries are roughly perpendicular to terial to weld cold cracking such as strength, mi-
the fusion boundary and have been referred to as crostructure and alloy composition. It is difficult to
“Type I” boundaries. In dissimilar welds, where an separate the effects individually because these factors
austenitic weld metal and ferritic base metal exist, are interrelated[27] . The nature of the fusion zone and
the second type of boundary that runs roughly par- the HAZ has a significant effect on the HIC suscepti-
allel to the fusion boundary is often observed. This bility of armour grade Q&T steel. Microstructure (in
has been referred to as “Type II” boundary[28] . These the HAZ adjacent to the fusion boundary) is prob-
boundaries typically have no continuity across the fu- ably the most important variable in controlling the
sion boundary to grain boundaries in the base metal. susceptibility of weld cold cracking. The suscepti-
Several investigators have reported that hydrogen- bility of steels to weld cold cracking increases with
induced disbonding typically follows Type II grain strength of the weld and is also usually associated
boundaries[7, 28–30] . with hard microstructures. However, at the same
The fusion zone microstructure of the ASS welds hardness, different microstructures can have different
(FA) exhibits a soft grain boundary phase (white susceptibilities for weld cold cracking. For example,
phase) of the interface similar to that of the type II low carbon martensite/bainite structures have greater
boundaries as described above. The region of white susceptibility for cold cracking than fine grained aci-
phase and fusion boundary are located in close prox- cular ferrite at equivalent hardness levels. Further,
imity distance to each other. The micrographs of the it is generally accepted that un-tempered, twinned
interrupted implant specimen revealed that the crack or dislocated lath martensite are the most suscepti-
was developed in the grain boundary (white phase) ble microstructures, while the preferred substructure
phase region (Fig. 6(a)) in the FA welds. The forma- combines a well-tempered martensite or bainite with
tion of the white phase (rich in carbon and Cr) in the an ausworking process to produce a refined packet
FA welds is due to the diffusion of carbon from base size and a uniform dispersion of carbides[31] . It is ev-
metal region to weld metal region and migration of ident from the micrographs of HAZ region in close
Cr from weld metal region to base metal region and proximity (Fig. 4(c) and (d)) to the fusion bound-
it depends upon the weld thermal cycle employed for ary consists of invariably untempered martensite in
fabricating the joints[7] . On the other hand, the fusion all cases. However, due to the smaller difference in
zone microstructure of the FF has a hard untempered the heat input of the welding process, the hardness
martensite and no type II boundary exists adjacent to in this region shows minor variations. The hardness
the fusion boundary. The crack followed this region of in the HAZ region of the FA weld metal is relatively
untempered martensite (Fig. 6(b)) in FF welds. The lower than that in FF weld. In general, hard HAZ mi-
microhardness values (Table 6) reveal that FA weld crostructure is more susceptible to cold cracking than
has a softer (grain boundary phase) region compared soft (coarse) microstructures. FA weld shows rela-
to FF joints adjacent to fusion boundary, which are tively lower hardness in the HAZ region compared to
much harder than their respective weld center. other welds. Thus, lower HAZ hardness is also one of
The results have shown that hydrogen introduced the contributing factors for improved resistance of FA
during welding can lead to HIC in dissimilar joints welds against cold cracking.
(FA) and also similar joint (FF). Hydrogen in the
welding arc is detrimental in two ways: (1) it increases 4.5 Effect of residual stress
dilution by the carbon steel base metal, increasing
the amount of un-tempered martensite formed, and Owing to localized heating during the weld-
(2) it interacts with untempered martensite under ing process and subsequent rapid cooling, residual
stress to cause cracking[21,30] . The most susceptible stresses can arise in the weld and base metal. Such
microstructure (untempered martensite) forms in the stresses are usually of yield point magnitude. Resid-
vicinity of the fusion boundary. The cracks are found ual stresses attributed to welding pose significant
in the regions near to the fusion boundary invariably problems in the accurate fabrication of structures
J. Mater. Sci. Technol., Vol.25 No.4, 2009 525
Fig. 9 Macrostructure of the single ‘V’ butt joints revealing no evidence of delayed cracks due to HIC: (a) FA
joint, (b) FF joint
because those stresses heavily induce brittle fracturing maintaining the appropriate preheating and interpass
and degrade the buckling strength of welded struc- temperatures during welding of armour grade Q&T
tures. These residual stresses are developed in the steels are prerequisites that must be strictly followed
vicinity of a weld during arc welding. A weldment to achieve defect free welds[41] .
is locally heated by most welding processes. There-
fore, the temperature distribution in the weldment 5. Conclusions
is not uniform, and metallurgical changes take place
as welding progresses along the weld. Therefore, the (1) The flux cored arc welding consumables have
welding residual stress is sometimes called restraint significant effect on weld cold cracking of armour
stress[32,33] . These stresses can give rise to distor- grade Q&T steel welds.
tion and under certain circumstances even to pre- (2) The welds made using ASS consumables (FA)
mature failure. Recent investigations by Reddy and offered a greater resistance to weld cold cracking than
Mohandas[34,35] revealed that the magnitude of the FF welds made using LHF consumables.
residual stress was found to vary with weld thermal (3) The presence of widely spaced delta ferrite in
cycle. Their findings suggested that higher heat input a large plain austenitic matrix, lower diffusible hy-
welds were associated with lower residual stress. drogen level, lower weld metal strength, lower fu-
The residual stresses play an important role sion boundary and HAZ hardness, softer grain bound-
as far as the quality and reliability of a welded ary phase, and lower residual stress are the factors
construction[36,37] and is of greater significance in pro- that contribute for the greater resistance of FA welds
moting cold cracking in high strength steel welds. Lee against cold cracking compared with FF joints.
et al.[38,39] revealed that the occurrence of cold crack,
was caused by a complex interaction of the diffusible
hydrogen supply, susceptible microstructure, and ten- Acknowledgements
sile residual stress. The risk of cold cracking is greater The authors are thankful to Armament Research
for welds with higher magnitude of residual stresses. Board (ARMREB), New Delhi for funding this project
Thus, lower residual stresses are much beneficial to of- work (Project No MAA/03/41), M/s Combat Vehicle
fering greater resistance to weld cold cracking due to Research Development Establishment (CVRDE), Avadi,
hydrogen[40] . In this work, the FA welds have a lower Chennai for providing base material, Department of Man-
residual stress than FF welds (Fig. 8). Hence, lower ufacturing Engineering, Annamalai University for provid-
ing implant testing facility and M/s Defence Metallurgical
residual stress in FA welds is also one of the contribut-
Research Laboratory (DMRL), Hyderabad for providing
ing factors for greater resistance to cold cracking. the facility to carry out metallurgical characterization fa-
It is evident from the above discussion that the cility for this investigation.
LHF flux cored consumables can also be used for weld-
ing armour grade high Q&T steels as they have very REFERENCES
low level of diffusible hydrogen in their welds and ac-
ceptable LCS values under implant conditions. Single [1 ] N. Yurioka: Mater. Design, 1995, 6(4), 154.
‘V’ butt joints (14 mm in thickness) was fabricated [2 ] R.S. Parmar: Welding Processes and Technology, 3rd
using the above consumables and as per the welding edn, Khanna Publishers, New Delhi, 1999, 53-61.
conditions given in Table 3. The macrograph of the [3 ] Y. Matsusita, T. Maruyama and S. Nishiyama: Kobe
joints in Fig. 9 shows that there is no evidence of de- Res. Develop., 1995, 45, 21.
layed cracking (due to HIC). Thus, the LHF flux cored [4 ] F. Ade: Weld. J., 1991, 70, 53.
consumables can be accepted as an alternative to the [5 ] F.R Coe: Welding Steels without Hydrogen Cracking,
conventional ASS consumables. However, all the nec- 1st edn, Abington Publishing, Cambridge, 1999, 220.
essary precautions like proper storage of wires, and [6 ] G. Madhusudhan Reddy, T. Mohandas and G.R.N.
526 J. Mater. Sci. Technol., Vol.25 No.4, 2009
Tagore: J. Mater. Process. Technol., 1995, 49, 213. [22] J.D. Parker and G.C. Stratford: Sci. Technol. Weld.
[7 ] G. Madhusudhan Reddy, T. Mohandas and D.S. Join., 1999, 4, 29.
Sarma: Sci. Technol. Weld. Join., 2003, 8, 407. [23] T.A. Siewert: Weld. Des. Fabr., 1978, 1, 87.
[8 ] P. Deb, K.D. Challenger, and D.R. Clark: Mater. Sci. [24] R. Kacar: Mater. Design, 2004, 25, 1.
Eng., 1998, 77, 157. [25] V.S.R. Murti, Y.R. Sastry and G.M. Reddy: J. Mater.
[9 ] Standard Methods for Determination of Diffusible- Process. Technol., 1993, 37, 759.
Hydrogen Content of Martensitic, Bainatic and Fer- [26] R.J. Pargeter: TWI, UK Rep. No.46, 1992, 47.
ritic Steels Produced by Arc Welding, American Weld- [27] R.K. Dayal and N. Parvathavarthini: Sadhana, 2003,
ing Society Guidelines, Section 4.3, 1986, 1-30. 28, 431.
[10] Destructive Tests on Welds in Metallic Materials-Cold [28] K. Asami and T. Sakai: ISIJ Int., 1995, 21(6), 2.
Cracking Tests for Weldments, Arc Welding Processes, [29] F. Matsuda and H. Nakagawa: Trans. JWRI, 1984,
Part 3: Externally Loaded Tests, ISO 17642, 2005, 1- 13(1), 159.
24. [30] N. Bailey: Welding of Dissimilar Metals, 2nd ed, TWI,
[11] IIS/IIW-447-73: Weld. World, 1974, 1(2), 9. UK, 1996, 34.
[12] V. Ramasubbu, S.K. Albert, V. Guruprasad and A.K. [31] J.M.B. Losz and K.D. Challenger: Proc. of First
Bhaduri: Proc. National Welding Seminar, ed. A.K. United states-Japan Symp. on Advances in Welding
Bhaduri, Indian Institute of Welding, Chennai, India, Metallurgy, Yokohama, Japan, 1999, 207.
2006, 342. [32] S. Kou: Welding Metallurgy, 2nd edn, John Wiley and
[13] R.D. Stout: Weldability of Steels, 4th edn, Welding Sons, New York, 1987.
Research Council, New York, 1987, 204. [33] P.C. Brand, T.H. De Keijser and G. Den Ouden: Weld.
[14] B. Basu: Proc. National conf. on Welding- J., 1993, 72, 93s.
Productivity & Quality, ed. A.K. Shah, Naval Ma- [34] G.M. Reddy and T. Mohandas: Proc. Nat. Weld.
terials Laboratory, Ambernath, India, 2007, 73. Seminar, Indian Institute of Welding, 1993, 33.
[15] S. Lampman: Weld Integrity and Performance, 3rd [35] G.M. Reddy and T. Mohandas: J. Mater Sci. Lett.,
edn, ASM International, USA, 1997, 212. 1996, 15, 1633.
[16] H. Granjon: Fundamentals of Welding Metallurgy, [36] N. Yurioka and H. Suzuki: Int. Mater. Rew., 1990,
2nd edn, Jaico Publishing House, Bombay, India, 1994, 35(4), 217.
224. [37] P.H.M. Hart: Weld. J., 1986, 65(1), 14s.
[17] R.E. Dobly: Sci. Technol. Weld. Join., 2000, 5, 341. [38] H.W. Lee, S.W. Kang and D.S. Um: Weld. J., 1998,
[18] N. Bailey: Weldability of Ferritic Steels, 3rd edn, 77(12), 503s.
Ambington Publishing, Cambridge, 1994, 76. [39] H.W. Lee, S.W. Kang and J.U. Park: Weld. J., 2001,
[19] ASME Boiler Pressure Vessel: Code II C–SFA 5.29, 80(6), 137s.
2007, 641-686. [40] P.V. Ramana, G.M. Reddy and T. Mohandas: Non-
[20] L.E. Svensson: Control of Microstructures and Prop- Des. Testing Evaluation, 2007, 6(1), 33.
erties in Steel Arc Welds, 4th edn, CRC Press, 1994, [41] G. Magudeeswaran: Fatigue and Fracture Toughness
212. Behaviour of Armour Grade, Quenched and Tem-
[21] M.D. Rowe, T.W. Nelson and J.C. Lippod: Weld. J., pered, High Strength Steel Welds, Ph.D. Thesis, An-
1999, 78, 31s. namalai University, India, 2008.