Available online at www.sciencedirect.

com

ScienceDirect
Procedia Materials Science 9 (2015) 419 – 427

International Congress of Science and Technology of Metallurgy and Materials, SAM –

CONAMET 2014

Fatigue life of GMAW and PAW welding joints of boron

microalloyed steels
Hernán G. Svobodaa,b,*, Horacio C. Nadalea
a
GTSyCM3, INTECIN, Faculty of Engineering, University of Buenos Aires, Las Heras 2214, Buenos Aires (1127), Argentina
b
CONICET, Rivadavia 1917, Buenos Aires, Argentina

Abstract

Recently boron microalloyed steels have become an alternative in weight saving in structural elements in automotive industry,
especially for parts like pillars and columns and reinforcements. These materials are used as quenched and tempered, reaching
strength values between 1000 and 1500 MPa. Welding is a relevant issue as a fabrication process in automotive industry, in
particular for high strength steels. Furthermore, fatigue life of these welded joints plays a very important role for these
applications. Gas metal arc welding (GMAW) is one of the more widely used welding processes in automotive industry. Indeed,
different aspects of plasma arc welding (PAW) make it an interesting option for these applications. The objective of this work
was to evaluate the fatigue life of GMAW and PAW welded joints of boron microalloyed steels, in a quenched and tempered
condition. Samples of 100×100 mm were butt welded by GMAW and PAW and microstructural and mechanically characterized.
Fatigue curves were obtained for both processes in a four point bending configuration. It was observed that PAW welded samples
had a higher fatigue life than those welded by GMAW. Weld bead toes of GMAW joints acted as geometric stress concentrators,
being the main responsible for fatigue resistance reduction.
© 2015
© 2015TheTheAuthors.
Authors.Published
Published by by Elsevier
Elsevier Ltd.Ltd.
This is an open access article under the CC BY-NC-ND license
Peer-review under responsibility of the Scientific Committee of SAM– CONAMET 2014.
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the Scientific Committee of SAM–CONAMET 2014
Keywords: Boron microalloyed steels; GMAW; PAW; microstrucutre; fatigue

* Corresponding author. Tel.: +054-11-4514-3009; fax: +054-11-4514-3013

E-mail address: [email protected]

2211-8128 © 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the Scientific Committee of SAM–CONAMET 2014
doi:10.1016/j.mspro.2015.05.012
420 Hernán G. Svoboda and Horacio C. Nadale / Procedia Materials Science 9 (2015) 419 – 427

1. Introduction

Boron microalloyed steels (BMAS) are part of advanced high strength steels and in the last years have become in
an alternative for weight reduction of structural parts in automotive industry. Particularly with the development of
hot stamping process these steels have reach a special impulse for its use in parts with complex geometries like
columns, reinforcements, pillars, etc. These materials are hot forged and then quenched and tempered, achieving
tensile strengths between 1000-1500 MPa and reducing significantly the springback [World Steel Association
(2014), Dilthey and Ste (2005)].

Nomenclature

AHSS Advance High Strength Steels

b Basquin exponent
BM Base Metal
BMAS Boron Microalloyed Steels
FI Fatigue Index
GMAW Gas Metal Arc Welding
HAZ Heat Affected Zone
CG-HAZ Coarse-grained HAZ
FG-HAZ Fine-grained HAZ
HI Heat input
HV Vickers microhardness
HVmin Minimum Vickers microhardness
IC Inter-critical HAZ
If Fatigue Index
IH Hardness Index
,V Strength Index
Kt Stress concentration factor of weld toe
LM Light Microscopy
MHZ Minimum hardness zone
Nf Cycles to Failure
PAW Plasma Arc Welding
Q+T Quenching and Tempering
T Weld toe angle
R Fatigue Stress Ratio
U Weld toe radius
SEM Scanning Electron Microscopy
SC Sub-critical HAZ
Va Stress amplitude
Vf Fatigue limit
V’f Basquin coefficient
Vuts Ultimate Tensiles Stress
WM Weld Metal

Welding of these materials is a relevant aspect for automotive industry. Among welding processes employed in
this industry, GMAW is one of the more used, as well as RSW, LW and PAW [World Steel Association (2014)].
For AHSS, the joint efficiency decrease with the increase in the material strength, due to the formation of softened
zones in the HAZ [World Steel Association (2014); Farabi et al. (2010)]. The fatigue life of these welded joints is
also a very important issue [World Steel Association (2014)]. Geometric aspects as the stress concentration
produced by the weld toe or other welding defects, residual stresses and microstructural features will define the
 Hernán G. Svoboda and Horacio C. Nadale / Procedia Materials Science 9 (2015) 419 – 427 421

fatigue life of the welded joints [Radaj (1996)]. In this context the information available about the fatigue life of
welded joints involving BMAS is scarce, as well as the effect of the welding process and its parameters.
The objective of this work was to study the fatigue life of welded joints of BMAS in Q+T condition, produced by
GMAW and PAW.

2. Experimental procedure

To reach the proposed objectives a sheet of BMAS with a thickness of 1 mm was used. Chemical composition
was determined and heat treatment of quenching and tempering was done on samples of 50×100 mm. It consisted in
austenization to 875 ºC during 10 min followed by quenching in oil, then tempering at 475 ºC during 30 min. The
obtained material in Q+T condition was characterized microstructural and mechanically by LM, SEM, Vickers
microhardness and tensile test.
Q+T samples were butt welded by GMAW and PAW in mechanized form. Fig. 1 shows the experimental set up
for both welding processes. In Table 1, the welding parameters used for both welding processes are presented.

Table 1. Welding parameters.

Proc. I (A) V (V) V (mm/seg) Gas Q (L/min) HI (J/mm)
GMAW 85 19 15.6 Ar-20CO2 20 0.11
PAW 50 13.5 7.7 Ar/Ar-5H2 0.8/20 0.09

For GMAW process the filler metal was a solid wire ER70S6 of 0.8 mm in diameter. The torch angle was 15º
pushing, the contact tip-work piece distance was 13 mm and it was used a backing plate of 5 mm thick of aluminum.
For PAW process the torch angle was 10º pushing, the retreating of the electrode was 1 mm, and the tip-work piece
distance was 2 mm.
It can be seen that the electric parameters are slightly lower in PAW than in GMAW, but it was compensated
with welding velocity, looking for a similar heat input (HI) in both processes.
The welded samples obtained by both processes were microstructural and mechanically characterized by means
of LM, SEM, HV and transversal tensile test.
Indeed, for both processes fatigue tests in four point bending, with the face of the samples in tension were done,
using a stress ratio R=0.1 and a test frequency of 20 Hz, to obtain the fatigue curve Va-N. The span used was 80 mm
and 40 mm for the upper actuator. Fig. 2(a) shows a scheme for test probes extraction for fatigue and tensile tests.
On every GMAW fatigue test sample, previously to the fatigue testing, the geometric aspects related to the weld
toe, according to Fig. 2(b). From the radius and angle the Kt factor was calculated using the methodology proposed
by Lagoda (2008). On each sample the four toes were analyzed, obtaining finally an average value for Kt.
After fatigue testing, some of the testing samples were characterized using LM and SEM.

a b
Fig. 1. Experimental set-up: (a) GMAW; (b) PAW.
422 Hernán G. Svoboda and Horacio C. Nadale / Procedia Materials Science 9 (2015) 419 – 427

a) b)

Fig. 2. (a) Samples extraction for tensile and fatigue testing of welded joints; (b) geometrical aspects measurements in GMAW weld toe.

3. Results and discussions

In Table 2 the chemical composition of the analyzed material is shown.

Table 2. Chemical composition (wt%).

C Mn Si P S B Fe
0.28 1.17 0.16 0.016 0.012 0.0031 Bal.

It can be seen that the material corresponds to medium carbon steel, with Mn and 30 ppm of B. In Fig. 3, the
microstructure obtained after the Q+T heat treatment is shown.
The obtained microstructure consisted in a uniform tempered martensite, with a Vickers microhardness of 300
+/-8 HV. The 0.2% proof stress was 877 MPa and the ultimate tensile strength was 940 MPa, with a specific
elongation to fracture of 5%, obtained in longitudinal samples respect to the rolling direction.
In Fig. 4, images of the welded joints obtained for both processes are shown.

Fig. 3. Microstructure of base material in Q+T condition.

a b

Fig. 4. Surface aspect of welded samples: (a) GMAW; (b) PAW.

 Hernán G. Svoboda and Horacio C. Nadale / Procedia Materials Science 9 (2015) 419 – 427 423

Fig. 5. Macrographs and Vickers microhardness profiles: (a) GMAW; (b) PAW.

GMAW presents the typical reinforcement of welding process with filler material while PAW does not have it,
due to its autogenous condition. Indeed a good surface aspect and the absence of defects can be seen.
Macrographs and microhardness profiles of both welded joints are shown in Fig. 5.
In Fig. 5(a), in the macrograph of GMAW joint, the different zones produced by the welding thermal cycle are
observed: the weld metal (WM), the heat affected zone (HAZ) and the base metal (BM). Also, it can be seen the
absence of internal defects like porosity or cracking, showing a very good root conformation with complete
penetration. Indeed the wetting of WM on BM in the weld toes can be observed, which represent a stress raisers that
will affect the fatigue life of the welded joint [Radaj (1996)].
Vickers microhardness profiles show a slightly decrease from the BM hardness (300 HV) to the weld centerline,
reaching a minimum at the subcritical/intercritical zone of 266 HV. This could be associated to the tempering of the
martensite. Then the microhardness values increase up to reach 500 HV near the fusion line, into the HAZ. This fact
is related with the partial or total austenization reaching that will form fresh martensite during cooling. Finally in the
WM the microhardness decrease up to 380 HV, due to the dilution of the BM with the filler metal.
For PAW welded joint [Fig. 5(b)] there are also no internal defects, a good root conformation and full
penetration, without undercut or reinforcement. In this sense, this process does not present stress raisers related with
geometric aspects.
The evolution of microhardness along the welded joint, shows a profile very similar to that obtained for GMAW
welded joint, with HVmin=239 HV and peak values and WM values very similar. Indeed, HAZ and WM sizes are
also very similar to the GMAW ones. This could be related with the similar HI applied in both cases.
Figs. 6(a) and (b) show the microstructural evolution for both GMAW and PAW welded joints, respectively.
In both cases a very similar microstructural evolution is observed. From BM up to WM it passes thru different
zones which constitute the HAZ. Adjacent to the BM is the subcrititcal zone (SC) where the martensite is submitted
to a high temperature tempering. Then it is the intercritical zone (IC) where the partial austenization allows the
formation of a fraction of fresh martensite. Next to IC zone there is the fine grain HAZ (FG-HAZ) and the coarse
grained HAZ (CG-HAZ) where complete austenization takes place and grain size increase with the proximity to the
fusion line. This last zone corresponds to the maximum hardness value. In the WM due to the dilution of the BM
with the filler metal the hardness is lower and no martensitic structures (bainite) could be form. These observations
are consistent with the microhardness profiles of Fig. 5.
424 Hernán G. Svoboda and Horacio C. Nadale / Procedia Materials Science 9 (2015) 419 – 427

a)

b)

Fig. 6. Microstructural evolution in the several zones of the welds: (a) GMAW; (b) PAW.

In Table 3, the mechanical properties of both welded joints are presented, as well as joint efficiency and hardness
index IH = HBM/HVmin

Table 3. Mechanical properties of welded joints.

Proc. HVmin Vuts (MPa) IH IV
GMAW 266 912 0.89 0.97
PAW 239 915 0.80 0.97

The joint efficiency for both processes was high, reaching 97%. The IH was something lower, especially for
PAW. The fracture was located at MHZ, for both cases. In Table 4, the geometric aspects results of the GMAW
fatigue samples are presented, as well as the average calculated value of Kt and Ktmax.
 Hernán G. Svoboda and Horacio C. Nadale / Procedia Materials Science 9 (2015) 419 – 427 425

Table 4. Kt values measured on GMAW fatigue test samples.

Proc. U (mm) T º  Ktprom .tmax

GMAW 0.61 40 1.17 1.21

The values of UandT are the average of four measures on ten samples. Ktmax was determined as the average of
the Ktmax of each sample. It can be seen that the local stress is 20% higher for samples welded with GMAW, due to
the effect of the stress raiser of the weld toe.
In Fig. 7, the fatigue curves obtained for both welding processes are shown.
PAW welded joints presents a higher fatigue strength than GMAW ones with more difference in high cycle
regime. This fact could be explained by the stress concentration produced in the weld toe of the GMAW joints,
which magnify locally the stress value, requiring a lower number of cycles to produce the fatigue failure than PAW
ones. This effect is amplified at lower stress level [Xu et al. (2013)].
Basquin constants were obtained from the correlation with the experimental data. In Table 5, the fatigue limit,
Basquin constants values and the fatigue index (If=Vf/Vuts) obtained for both GMAW and PAW are presented. The
fatigue limit was defined as the fatigue strength for 107 cycles.

Table 5. Fatigue limits and Basquin constants for both welding processes.
Proc. Vf (MPa) Vuts (MPa) If V`f (MPa) b

GMAW 150 912 0.16 624 -0.13

PAW 250 915 0.27 1345 -0.05

Fatigue index for GMAW welded joints was lower than for PAW ones, which reaches 0.27. This value is higher
than those obtained previously for welded joints of AHSS [Xu et al. (2013), Bonello and Svoboda (2013)]. The
Basquin constants are values of technological interest to estimate fatigue life of welded structures with these
materials and processes.
For GMAW fatigue tested samples in all cases the cracks nucleated at the weld toe, as was expected [Figs. 8(a)-
(c)].

Fig. 7. Fatigue curves for GMAW and PAW welded joints.

426 Hernán G. Svoboda and Horacio C. Nadale / Procedia Materials Science 9 (2015) 419 – 427

a) b) c)

Fig. 8. Fracture location for GMAW fatigue tested samples.

a b c

Fig. 9. Fracture location for PAW fatigue testing samples: (a) MS; (b) MB; (c) ZRG.

In Fig. 8(b), it can be observed that the fissure nucleates at the weld toe and propagates in a direction orthogonal
to the tensile stresses. In Fig. 8(c), a detail of the nucleation site can be observed, locating where the fusion line
reach the free surface, and then the propagation proceeds thru CG-HAZ and FG-HAZ at the bottom.
For PAW welded joints, the crack nucleates in different zones, depending on the stress level. For high stress level
the fracture was located into the WM in the central shrinkage or in the fusion line [Fig. 9(a)]. For medium stresses it
was at MHZ or in BM [Fig. 9(b)], and for lower stress levels the nucleation site was positioned in the CG-HAZ [Fig.
9(c)]. These observations has been previously reported for steels of lower strength [Xu et al. (2013)] and they can be
associated to the absence of severe geometric stress raisers, which allows that metallurgical aspects became relevant.
In Fig. 10, fracture surface SEM images of a GMAW sample are shown.
In Fig. 10(a) the initiation zone is shown whereas in Fig. 10(b) striations of propagating zone can be seen.
Secondary cracks are also observed. These aspects are consistent with the previously observed for this type or
welded joints [Xu et al. (2013)]. There were no significant differences in the fracture surfaces for different welding
processes.

a b

Fig. 10. Fracture surfaces for fatigue tested samples: (a) initiation zone, (b) propagation zone.
 Hernán G. Svoboda and Horacio C. Nadale / Procedia Materials Science 9 (2015) 419 – 427 427

4. Conclusion

In the present work butt welded joints of BMAS of 1 mm thick in Q+T condition, obtained by GMAW and PAW
processes were microstructural and mechanically analyzed. Tensile and fatigue behavior were evaluated.
The procedures for both welding processes were successfully applied producing defects free welded joints.
Both obtained joints presented inside the HAZ a softened zone (240-270 HV) with respect to the base material
(300 HV) associated to the tempering of the martensite, as well as a hardened zone (500HV) near the fusion zone,
associated to the formation of fresh martensite. In the WM the microhardness was something lower reaching 360-
400 HV, due to dilution of the base metal with the filler material. In tensile test, both processes achieve high joint
efficiencies of 97%, being the location of the fracture in the MHZ in the HAZ.
Fatigue life of PAW welded joints was significantly higher than the GMAW ones, with a fatigue limit 78%
higher. This could be associated to the local stress magnification at the toe of GMAW beads. In this process the
cracks nucleated at the mentioned weld toe. For PAW fatigue samples the fracture was located at WM for high
stresses, at BM or MHZ for medium stresses and at HAZ-CG for lower applied stresses.

Acknowledgements

Authors are grateful to INTI-Mecánica for chemical analysis and SEM, Ternium Siderar for material provision
and to Ing. Horacio De Rosa. This Project is partially supported by University of Buenos Aires and ANPCYT.

References

Bonello, T., Svoboda, H., 2013, Efecto del modo de transferencia en el proceso GMAW sobre la vida a la fatiga de uniones soldadas de aceros
DP750, XI Congreso Regional de Ensayos No Destructivos y Estructurales, Mar del Plata, Argentina.
Dilthey, U., Ste, L., 2005, Multimaterial car body design: challenge for welding and joining., Science and Technology of Welding and Joining
11(2), 135-142.
Farabi, N., Chen, D., Shou, Y., 2010, Fatigue properties of laser welded dual-phase steel joints, Procedia Engineering 2(1), 835-843.
Lagoda, T., 2008, Lifetime estimation of welded joints, Springer, Berlin.
Radaj, D., 1996, Review of fatigue strength assessment of nonwelded and welded structures based on local parameters, International Journal of
Fatigue 18, 153-170.
World Steel Association, Committee of Automotive Applications, Advanced High Strenght Steels (AHSS): Application Guidelines (5.0 Version).
Brussels: World Steel Association, 2014.
Xu, W., Westerbaan, D., Nayak, S., Chen, D., Goodwin, F., Zhou, Y., 2013, Tensile and fatigue properties of fiber laser welded high strength low
alloy and DP980 dual-phase steel joints, Materials and Design 43, 373-383.