investigation on fatigue fracture behaviors of spot welded q&p980 steel

$: Investigation on fatigue fracture behaviors of spot welded Q&P980 steel$
International Journal of Fatigue xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

International Journal of Fatigue

journal homepage: www.elsevier .com/locate / i j fa t igue

Investigation on fatigue fracture behaviors of spot welded Q&P980 steel

http://dx.doi.org/10.1016/j.ijfatigue.2014.03.0040142-1123/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +86 24 23971043.E-mail address: [email protected] (Z.F. Zhang).

Please cite this article in press as: Wang B et al. Investigation on fatigue fracture behaviors of spot welded Q&P980 steel. Int J Fatigue (2014),dx.doi.org/10.1016/j.ijfatigue.2014.03.004

B. Wang a,b, Q.Q. Duan a, G. Yao a, J.C. Pang a, X.W. Li b,c, L. Wang d, Z.F. Zhang a,⇑a Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, 72 Wenhua Road, Shenyang 110016, Chinab Key Laboratory for Anisotropy and Texture of Materials, Ministry of Education, Northeastern University, Shenyang 110819, Chinac Institute of Materials Physics and Chemistry, College of Sciences, Northeastern University, Shenyang 110819, Chinad Baoshan Iron & Steel Co., Ltd., Shanghai 201900, China

a r t i c l e i n f o

Article history:Received 17 November 2013Received in revised form 24 February 2014Accepted 6 March 2014Available online xxxx

Keywords:Resistance spot weldQ&P980 steelFailure modeFatigue properties

a b s t r a c t

Microstructural characterization, micro-hardness tests, tensile and fatigue tests of spot welded Q&P980steel were performed using tensile-shear and cross-tension specimens. The hardness values of nuggetand base material were measured to be 497 HV and 334 HV, respectively. It is found that the fatiguecracks in heat affected zone (HAZ) initiate at the interface between two sheets. The fatigue failure modesconsist of the fracture along the circumference or along the direction of width for tensile-shear specimensand pullout or fracture along the direction of width for cross-tension specimens. It is also found that thefatigue properties of spot welded Q&P980 and DP780 specimens are approximately the same in the caseof tensile-shear and cross-tension specimens.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

In order to satisfy the weight reducing, energy saving and secu-rity principles of new generation cars, the advanced high-strengthsteels (AHSS) have been widely used in automobile industry. Sheetsteels for automotive application have been defined as the ‘‘FirstGeneration’’ of AHSS and the ‘‘Second Generation’’ of AHSS. Thedual phase (DP) steels, transformation induced plasticity (TRIP)steels, complex phase (CP) steels and martensitic (MART) steelsare considered as the first generation AHSS [1]. Second generationAHSS includes twinning induced plasticity (TWIP) steels, Al-addedlightweight steels with induced plasticity (L-IP�), and shear bandstrengthened steels (SIP steels) [1]. Recently, there have beenincreased interests in the development of the ‘‘Third Generation’’of AHSS, namely, steels with strength-ductility combination signif-icantly better than that of the first generation AHSS but signifi-cantly cheaper than the second generation AHSS [1]. In 2003,Speer et al. [2] defined a new heat treatment process that can makesteel microstructures which have residual austenite and controlledamounts of martensite, and the process was referred to as‘‘quenching and partitioning (Q&P)’’. In this process, austenitewas quenched below martensite start temperature (Ms) to containmartensite–austenite, and then heat treated at partitioningtemperature which can either be equal to or greater than quenchtemperature. At partitioning temperature, carbon diffused from

martensite to residual austenite to enhance the austenite stability[1,3]. Currently, the Q&P steel becomes the third generationadvanced high-strength steel [1]. More recently, Paravicini Baglianiet al. [4] compared the properties of a low alloy medium carbonsteel (0.28C wt%) obtained after Q&P and quenching and tempering(Q&T) treatments, and it was found that the strength–toughnesscombination of specimens treated by the Q&P process was betterthan that of the Q&T process; Sun and Yu [5] also found that themechanical properties of low-carbon steels (0.2C wt%) treated bythe Q&P process exhibited better combination of strength and duc-tility than that of the Q&T process. Besides, Cerny et al. [6] reportedthat the fatigue endurance of the Q&P treated material was excel-lent, followed by classically heat treated and as received conditionsin sequence; and Wang et al. [7] found that Q&P exhibited muchhigher formability than other same grade high-strength steels.Therefore, the Q&P steel has potential application in automobileindustry due to its good combination of high strength and goodductility, as well as the excellent fatigue performance and form-ability. Q&P steels are well suited for cross members, longitudinalbeams, B-pillar reinforcements, sills and bumper reinforcements[7].

On the other hand, the materials used in auto manufacturingcan be joined by a variety of methods, such as resistance spotwelding, fiber laser welding and weld bonding [8]. But the resis-tance spot welding remains the primary method in automobilemanufacturing. For example, a typical vehicle could contain morethan 3000 spot welds [9]. Wang et al. [7] reported that when resis-tance spot welded, Q&P980CR steels required less current and

http://

http://dx.doi.org/10.1016/j.ijfatigue.2014.03.004

mailto:[email protected]


http://www.sciencedirect.com/science/journal/01421123

http://www.elsevier.com/locate/ijfatigue



2 B. Wang et al. / International Journal of Fatigue xxx (2014) xxx–xxx

higher electrode force than conventional steels, and the 1.6 mmQ&P980CR showed good spot weld strength performance. In gen-eral, the spot weld undergoes fatigue damage during the serviceperiod. Thus, it is of significant importance to investigate the fati-gue performance of spot welds for achieving a safe and reliabledesign. There are many reports about the fatigue behavior of steelsafter resistance spot welding. For example, Rathbun et al. [10]found that the fatigue performance of spot welded high-strengthsteels was mainly determined by the geometric factor rather thanthe material and microstructure. As well as, Vural and Akkus [11]found that the fatigue life increased as the nugget diameterincreased; and Shariati and Maghrebi [12] found the longer fatiguelife in thicker tensile-shear specimens. Xu et al. [13] concluded thatfor thin spot welded sheet its microstructure played an importantrole in the fatigue strength, when sheet thickness increased thefatigue strengths of spot welds of different materials were almostthe same. The fatigue strengths of spot welded DP600 GI(1.53 mm), TRIP600 (1.64 mm) and HSLA340YGI (1.78 mm) speci-mens were approximately the same in the case of tensile sheartests and coach peel type specimens for both low- and high-cyclefatigue lives [9]. In addition, Nakayama et al. [14] confirmed thatthe fatigue limit of spot welded 270 MPa-grade steel and590 MPa-grade steel was determined by the fatigue limit of heataffected zone (HAZ) and the residual stress. However, there wereno reports about the fatigue behaviors of spot welded Q&P steel.

Based on the investigations above, the purpose of the currentstudy is to further investigate the fatigue behavior of Q&P980 steelafter spot welding for both tensile-shear and cross-tension speci-mens. Firstly, the microstructure change, micro-hardness distribu-tion and tensile properties of spot welded steel were examined.Then, the fatigue failure modes were studied and the crack initia-tion and propagation mechanisms were discussed. Finally, the fati-gue properties of spot welded Q&P980 steel were compared withthose of spot welded DP780 steel.

2. Experimental material and procedure

The material used in this study was Q&P980 steel, the chemicalcomposition and mechanical properties of which are shown inTables 1 and 2, respectively. The dimensions of fatigue specimensare shown in Fig. 1, and the sheet thickness was 1.6 mm. The ten-sile-shear specimens had length of 125 mm and width of 40 mm.The cross-tension specimens had length of 150 mm and width of50 mm, respectively.

The metallographic samples were cut along the center of thespot welding steel in the direction of the width of specimens, pol-ished, and etched by 4% Nital (4 mL HNO3 and 96 mL alcohol). Thesize of nugget and spot weld were measured by Keyence VHX-1000E Digital Microscope. The nugget diameter and spot welddiameter of fatigue specimens were equal to 6.62 ± 0.04 mm and7.93 ± 0.04 mm, respectively; as shown in Fig. 2. The microstruc-ture was observed by LEO SUPER35 scanning electron microscope(SEM) and confirmed by the D/max 2400 diffraction instrument.Micro-hardness was carried out at intervals of 0.2 mm using LECOAMH-43 Micro-Hardness Tester at 200 g load for holding 13 s, andthe sample and paths are shown in Fig. 4.

The tensile tests of spot weld were carried out according to GB2651-89, and the tensile experiments of spot weld were carried outon INSTRON 5982 (Fig. 3(a)) testing machine with a strain rate of

Table 1Chemical composition of Q&P980 steel (wt%).

C Si Mn P S Al

0.23 1.55 1.92 0.010 0.002 0.040

Please cite this article in press as: Wang B et al. Investigation on fatigue fracdx.doi.org/10.1016/j.ijfatigue.2014.03.004

1 mm/min. The fatigue tests of spot weld were performed accord-ing to GB/T 15111-94, and three specimens were required for eachload level. The fatigue experiments of tensile-shear and cross-ten-sion specimens were performed on INSTRON 8801 (Fig. 3(b)) andINSTRON 8871 (Fig. 3(c)) testing machine, respectively. The speci-mens were under sinusoidal load with a load-ratio R = 0.1, in thefrequency range from 10 to 50 Hz. In order to prevent a bendingmoment applied at the spot weld, shims with the same thicknessas the sheet were glued at both ends of the tensile-shear speci-mens. As the specimens ruptured or the length of cracks was equalto the diameter of spot weld, the failure was considered to takeplace, and specimens that survived 107 cycles were called as run-ning out. As the three specimens performing under the same loadlevel all survived 107 cycles, the corresponding fatigue load wasconsidered as the conditional fatigue limit load (107 cycles).Finally, LEO SUPER35 (SEM) was used to identify the fatigue crackinitiation site, crack propagation route and to observe the charac-teristics of fracture surface.

3. Results and discussion

3.1. Micro-hardness distribution and microstructure change

The micro-hardness distribution is shown in Fig. 4. It could beseen that the hardness of nugget and base material (BM) were497 HV and 334 HV, respectively; and the hardness of nuggetwas nearly 1.5 times higher than the base material. The SEM andX-ray diffraction (XRD) examination indicated that the microstruc-ture of nugget was predominantly martensite (Figs. 5(a) and 6(a)),and the microstructures of BM were ferrite, martensite and resid-ual austenite (Figs. 5(d) and 6(b)). The microstructure in the nuggetdepended on heat input and cooling rate during resistance spotwelding progress. It was reported that the cooling rates rangedfrom over 105 �C/s for sheet thickness less than 0.5 mm, to nearly2000 �C/s for 2 mm thick sheet [15]. Thus, the cooling curve missedto touch the nose of the continuous cooling transformation (CCT)curves, resulting in only martensitic transformation [16]. Also Xuet al. [13] and Ma et al. [17] reported that the nugget was full ofmartensite. There was a hardness peak between the nugget andheat affected zone (HAZ) (see Fig. 4). The reason for the occurrenceof hardness peak could be that the thermal history of this regionresults in smaller grain size compared with that of nugget [13].Besides, Ma et al. [17] found that there was a transition zonebetween nugget and HAZ, and that the microstructure comprisingof martensite in the transition region was finer than that of thenugget; however, such a transition zone was not found in thisstudy. The microstructure of HAZ was mainly composed of mar-tensite and carbide (Figs. 5(b) and 6(b)), but the martensite inthe HAZ was smaller and its weight fraction (�84.3%) was lowerthan that in the nugget. In the HAZ, the hardness value varies,i.e., the region near the nugget had a higher hardness value thanthe region far away from nugget, similar results were also foundin fiber laser welded DP980 [16] and DP600 [18]. In addition, therewas a region with a lower hardness value (�300 HV) even than theBM (334 HV) between the base material and HAZ, which was calledthe soft zone as indicated in Fig. 4. The similar soft zone was alsofound in spot welded DP600 [19] and M190 [20] steels. Fig. 5(c)showed that the soft zone of Q&P980 steel after spot welding con-tained ferrite and tempered martensite. The softening degreedepends on material chemistry and welding heat input [21].

3.2. Tensile properties of spot weld

The ultimate tensile loads of spot welded Q&P980 tensile-shearand cross-tension specimens were 23.7 ± 0.6 kN and 10.8 ± 0.5 kN,

ture behaviors of spot welded Q&P980 steel. Int J Fatigue (2014), http://



Table 2Mechanical properties of Q&P980, DP980 [16] and DP780.

Material r0.2/MPa rb/MPa Uniform elongation (%) Elongation to fracture (%)

Q&P980 776 ± 14 1018 ± 0.2 19.4 ± 1.5 29.9 ± 2.1DP980 720 1095 14.2DP780 500 ± 7.1 857 ± 3.5 18.4 ± 1.3 28.7 ± 1.8

Fig. 1. The dimensions of fatigue specimens (mm). (a) Tensile-shear specimen; and (b) cross-tension specimen.

Fig. 2. The illustration of spot weld diameter and nugget diameter.

B. Wang et al. / International Journal of Fatigue xxx (2014) xxx–xxx 3

respectively. The fractured tensile specimens are shown in Fig. 7,and it could be seen that the failure modes of tensile-shear andcross-tension tensile tests were interfacial failure (Fig. 7(a)) andpullout failure (Fig. 7(b)), respectively. Fig. 8 shows the load–dis-placement curves of the tensile tests. Apparently, for the tensile-shear tensile test, the load–displacement curve exhibited a linearrelationship; and the load dropped quickly to zero immediatelyafter failure. In contrast, for the cross-tension tensile test, theload–displacement curve had a nonlinear region before reachingthe maximum load. The high displacement value of cross-tensiontensile test was due to the plastic deformation of base material;and the shape of the ‘‘tail’’ of the curve was consequent upon a par-tial nugget pullout and subsequent tearing of base material [22], asshown in Fig. 7(b).

In addition, the pullout failure of tensile-shear specimens andinterfacial failure of cross-tension specimens were reported insome existing works by others [22–25]. For tensile-shear tensiletest, there was a competition between the pullout and the interfa-cial failure modes [25]. Radakovic and Tumuluru [25] noted thatthe interfacial failure mainly initiated at the notch that was createdat the sheet interface and the pullout failures nucleated outside theweld nugget at the base material. For tensile-shear tensile tests,the minimum nugget size (Dc) to ensure pullout failure modewas proportional to the ratio of the hardness of the base materialto the hardness of nugget [24], as expressed by [23]:


Dc ¼ 8tðHÞFL

ðHÞNð1Þ

where t is the thickness of sheet, (H)FL is the hardness of failurelocation, and (H)N is the hardness of nugget.

In this study, the thickness of the sheet was 1.6 mm, and thebase material can be considered as the failure location. The valuesof hardness of base material and nugget were 334 HV and 497 HV,respectively. Taking them into Eq. (1), it was obtained that the crit-ical nugget size (Dc) was equal to 8.60 mm. However, the nuggetsize in this study was 6.62 ± 0.04 mm, which meant that the nug-get size in this study was smaller than the minimum nugget size(Dc), ensuring pullout failure mode, so the failure mode of ten-sile-shear tensile test was interfacial failure. For cross-tensiontensile test, the critical nugget size was lower than that of ten-sile-shear test [24], and Chao [26] derived an equation for criticalweld nugget size (Dc) in the cross-tension tensile test as below:

Dc ¼ 0:86sf

Kc

� �2=3

t4=3 ð2Þ

where sf is the shear strength of spot weld, Kc is the fracture tough-ness of spot weld, and t is the thickness of sheet. However, the frac-ture toughness of spot weld was hard to measure, and the criticalnugget size of cross-tension test could not be calculated in thisstudy.

In the tensile-shear tensile test, the driving force for the interfa-cial failure mode was the shear stress at the interface; and the ten-sile stresses around the nugget cause the plastic deformation in thesheet thickness direction was the major reason for pullout failure[23]. The elongated (shear-type) dimples in Fig. 9(a) indicated thatthe fracture mechanism was ductile and the failure occurred undershear stress; which was similar with the results by Pouranvariet al. [23]. While in the cross-tension tensile test, the notch atthe interface experienced open loading, which was the drivingforce for interfacial failure; and the driving force for pullout failurewas the shear stress at the nugget circumference [24]. Fig. 9(b)




Fig. 3. The equipments of tensile and fatigue testing. (a) INSTRON 5982; (b) INSTRON 8801; and (c) INSTRON 8871.

Fig. 4. Micro-hardness distribution of Q&P980 steel spot weld.

Fig. 5. Microstructure of Q&P980 steel spot weld. (a) Nu



shows the fractograph of specimen under cross-tension tensiletest, and the fracture surface was mainly featured by the formationof elongated (shear-type) dimples, indicating that the fracturemechanism was ductile and the failure occurred under shearstress. Chao [22] also found that the fracture surface of a pulloutfailure sample was full of elongated or ‘‘fish scale’’ dimples. Sodespite the fact that the global loading mode to the cross-tensionspecimen was tensile, the driving force for pullout failure modewas shear stress.

3.3. Fatigue strength and failure mode

The F–N curves of Q&P980 steel spot weld are shown in Fig. 10.Clearly, the fatigue life increased as the applied load decreased. Forthe same fatigue life, the fatigue strength (load) of tensile-shearspecimens was higher than that of cross-tension specimens. Theconditional fatigue limit loads (107 cycles) of tensile-shear and

gget; (b) HAZ; (c) soft zone; and (d) base material.




Fig. 6. XRD patterns of spot welded Q&P980. (a) Nugget and HAZ; and (b)BM. (a forferrite or martensite; c for austenite; e for carbide).

Fig. 7. The tensile failure mode of spot welded Q&P980 steel. (a) Tensile-shearspecimen; and (b) cross-tension specimen.

Fig. 8. Load–displacement curves of tensile-shear and cross-tension specimens.


cross-tension specimens were 1300 N and 280 N, respectively. Inaddition, the F–N curves of tensile-shear and cross-tension


specimens exhibited a linear relationship (Fig. 10), and the F–Ncurves of tensile-shear and cross-tension specimens could bedrawn as Eqs. (3) and (4), respectively.

log Fmax ¼ 5:08� 0:28 log Nf ð3Þ

log Fmax ¼ 4:28� 0:25 log Nf ð4Þ

the coefficients of determination for Eqs. (3) and (4) were 0.998 and0.990, respectively.

It was apparent that the spot weld of tensile-shear and cross-tension specimens were undergoing shear stress (s�1) (Fig. 11(a)and (c)) and bending stress (rb) (Fig. 11(b) and (d)). The shearstress (s�1) could be expressed as follows:

s�1 ¼F1

pr2 ð5Þ

where F1 is the applied load of tensile-shear specimens, and r is thespot weld radius.

The bending moment (M) of cross-tension spot weld could beexpressed as below:

M ¼Z l=2

0pxdx ¼

Z l=2

0

F2

lxdx ¼ F2l

8ð6Þ

where p is the applied stress, l is the length of the sheet, and F2 isthe applied load of cross-tension specimens.

The bending stress (rb) could be drawn as follows:

rb ¼My

I¼ F2l=8

pr4=4� r ¼ F2l=2pr3 ð7Þ

where I is the moment of inertia, y is the distance from the center toone side of the spot weld, and r is the spot weld radius.

It was well known that the stress for the crack opening canmake the crack to propagate. In this study, the stress for the crackopening of tensile-shear and cross-tension spot weld was definedas r1 and r2, as shown in Fig. 11(c) and (d). Then, r1 and r2 couldbe expressed as follows:

r1 ¼ s�1 � sina ¼ F1

pr2 � sin a ð8Þ

r2 ¼ rb � cos a ¼ F2l2pr3 � cos a ð9Þ

where a is the kinked angle as shown in Fig. 8.In addition, as Paris equation [27] shown:

dadN¼ cðYDr

ffiffiffiapÞn ð10Þ




Fig. 9. SEM fractograph of spot welded Q&P980 steel under tension. (a) Tensile-shear test; and (b) cross-tension test.

Fig. 10. F–N curves of Q&P980 steel spot weld. (a) Tensile-shear specimens; and (b)cross-tension specimens. The insert images show fatigue failure modes of Q&P980spot weld.

Fig. 11. Stress analysis in the spot weld. (a) Schematic illustration of tensile-shearspecimen; (b) schematic illustration of cross-tension specimen; (c) the normalstress distribution versus angle a of tensile-shear specimen; and (d) the normalstress distribution versus angle a of cross-tension specimen.


Nf ¼2

ðn� 2ÞcðYDrÞn1

aðn�2Þ=20

� 1

aðn�2Þ=2c

" #ð11Þ

where c and n are the material constants. For the same material andsheet thickness, c, n, Y, a0 and ac could be approximately consideredas the same. So for the same fatigue life of tensile-shear and cross-tension specimens, the Dr was similar. It could be considered thatfor the same fatigue life of tensile-shear and cross-tension speci-mens, the crack opening stress r1 equaled to r2, so

tan a ¼ F2l2pr3 =

F1

pr2 ð12Þ

where l was 150 mm, and r was 3.97 mm. according to the dataindicated in Fig. 10, for the same fatigue life (Nf = 107), the fatigue


strengths (loads) of tensile-shear and cross-tension specimenswere1300 N (F1) and 280 N (F2), respectively. The kinked angle awas thus calculated to be 76�. The experiment result showed thatthe kinked angle was 75� (Fig. 12). The calculated result was in goodagreement with experiment result; and it could explain why thefatigue strength (load) of tensile-shear specimens was higher thanthat of cross-tension specimens.

The spot welds of different materials have different fatigue fail-ure modes. Rathbun et al. [10] reported three types of failuremodes for DQSK tensile-shear specimens, namely, a plug failurewith a surrounding ring of plastically deformed material, plug fail-ure without the ring of deformed material, and fatigue crack prop-agation around the circumference and then into the base material.Ma et al. [17] noted the interfacial fracture, fracture along the cir-cumference, and fracture along the straight line normal to theloading direction for DP600 tensile-shear specimens. The interfa-cial fracture, and plug and hole type failure were reported inM190 and DP780 tensile-shear specimens [20,28]. In this study,the tensile-shear specimens showed fracture failure along the cir-cumference or along the direction of width (Fig. 10(a)) after fatiguetests. However, plug and hole type failure modes were not found inthis study due to the high hardness of nugget. When the tensile-shear specimens were deformed under high load (6000–8000 N),it showed fracture along the circumference (Fig. 10(a)). When theapplied load was less than 6000 N, the fracture was found to occuralong the direction of width (Fig. 10(a)).

Failure modes of DQSK [10] cross-tension specimens such as thecomplete plug failure, weld separation and rotation of the button,




Fig. 12. Photograph of the cross-section of a fatigued Q&P980 spot weld.


button rotation and crack propagation into base material wereinvestigated previously. Hilditch et al. [29] noted a mixture ofinterfacial fracture and weld pullout, weld separation resultingfrom the fatigue crack penetration through the thickness of thesheet, and the weld rotating out of plane, followed by crackpropagation into the base sheet for TRIP590 cross-tension speci-mens. There were two failure types for the present Q&P980cross-tension specimens. As the applied load was higher than360 N, the cross-tension specimens showed weld pullout or plugfailure (Fig. 10(b)), while the applied load was less than 360 N,specimens ruptured along the direction of width (Fig. 10(b)).

3.4. Fatigue crack initiation, propagation and fractography

In the past decade, Satoh et al. [30] used a three-dimensionalelastic-plastic stress analysis and found that the greatest strainoccurred at a site slightly away from the HAZ, in the base material;and Lanciotti and Polese [31] found that fatigue crack nucleated atthe site of 0.93 mm or 0.48 mm away from the nugget in spot weldof stainless steel AISI 301. More recently, Hassanifard et al. [32]reported that the circumference of the nugget had the maximum

Fig. 13. SEM micrograph of fatigue fracture surface of Q&P980 spot weld. (a) Crack initiafracture area.


stress concentration factor and the fatigue crack initiation wouldoccur at such sites. In the present study, for tensile-shear andcross-tension fatigue tests, the fatigue cracks mainly initiated nearthe notch tip in HAZ and at the interface between two sheets, asshown in Fig. 12. Fig. 13(a) showed the SEM micrograph of crackinitiation area, consisting of radial stripe, and the stripe initiatedfrom a ‘‘ring’’ of material existing on the periphery of the spot weldat the interface of two sheets. The ‘‘ring’’ resulted from meltedmaterial which was forced out from the weld during welding pro-cess [10]. The micrograph of the ‘‘ring’’ is shown in Fig. 13(b), itwas found that the surface of the ‘‘ring’’ had been oxidized, andsome globular scales formed due to exposure to air [10], mean-while, the pre-cracks as a result of metallic evaporation and highcooling rate were found.

After initiation, the fatigue cracks propagated through thethickness. Fig. 12 showed that the fatigue cracks penetrated thethickness of sheet, and the cracks located in the HAZ. It was knownthat the fatigue crack propagation path was determined by thestress distribution around the welding region [10]. Xu et al. [13]considered the fatigue crack as a kinked crack with respect to ori-ginal notch tip and found that the kinked angle was close to 80�.Hassanifard et al. [32] also calculated the stress intensity factorand J-integral by using finite element method, and found that thekinked angle was nearly 75� with respect to the longitudinal direc-tion. In this study, the kinked angle was found to be nearly 75� aswell (Fig. 12). It was reported that the fatigue crack propagationlife for tensile-shear specimen contained the number of cyclesfor crack to penetrate the thickness and propagate through thewidth of specimen [32,33]. In this study, the fatigue cracks of ten-sile-shear specimens first propagated through the thickness, thethrough-thickness cracks could be considered as the secondaryfatigue crack source, which grew into the base material along thedirection of width up to the final fracture. As for cross-tensionspecimens, the fatigue cracks penetrated the thickness until frac-ture as the applied load was higher than 360 N, while the fatiguecracks first penetrated the thickness, and then grew into the basematerial along the direction of width up to fracture as the loadwas less than 360 N, as the case of tensile-shear specimens. Thefractograph of crack propagation area was shown in Fig. 13(c),

tion area; (b) high magnification of ‘‘Ring’’; (c) crack propagation area; and (d) final





and that transgranular fracture, featuring cleavage facets and sec-ondary cracks were observed, but no typical fatigue striationformed. The final fracture area consisted primarily of equiaxialdimples, as shown in Fig. 13(d).

3.5. Fatigue properties comparison

It was reported that the Q&P980 steel had higher yielding ratioin comparison with DP590 and DP980 steels, and that the elonga-tion was equivalent with DP590 steel, while the strength wassimilar to DP980 steel [34]. Table 2 shows the mechanical proper-ties of Q&P980, DP980 [16] and DP780. It could be seen that thestrengths of Q&P980 and DP980 were approximately the same,but the elongation to fracture of Q&P980 was nearly two timeshigher than that of DP980. Meanwhile, the elongation of Q&P980was approximately equivalent with DP780 steel, and the strengthof Q&P980 was nearly 1.4 times higher than that of DP780. TheQ&P980 steel has better combination of high strength and ductilitythan that of DP980 and DP780 steel. Whether the fatigue proper-ties of Q&P980 spot weld is better than those of DP780 steel spotweld?

The fatigue data of spot welded Q&P980 were compared withthat of spot welded DP780, and the fatigue experimental proce-dures of spot welded DP780 were the same as those of spot weldedQ&P980. The sheet thickness and nugget diameter of DP780 spotweld were 1.60 mm and 6.78 ± 0.07 mm, respectively; and the geo-metric factor of Q&P980 and DP780 spot weld was approximatelythe same. Fig. 14 shows the fatigue properties comparison betweenspot welded Q&P980 and DP780 in the case of tensile-shear andcross-tension specimens. It was observed that at low cycles, there

Fig. 14. Fatigue property comparisons between spot welded Q&P980 and DP780.(a) Tensile-shear specimens; and (b) cross-tension specimens.


was a little gap between the F–N curves of spot welded Q&P980and DP780 specimens, and the spot welded DP780 specimenshad a little higher fatigue strength (load) than that of spot weldedQ&P980 specimens for both tensile-shear and cross-tension speci-mens. As the fatigue load decreasing, the F–N curves of spotwelded Q&P980 and DP780 specimens tended to overlap. For thecase of high cycles, the fatigue strengths (loads) of spot weldedQ&P980 and DP780 specimens were the same. Thus, the fatigueproperties of spot welded Q&P980 and DP780 specimens wereapproximately the same in the case of tensile-shear and cross-ten-sion specimens for both high and low load. Though, Q&P980 steelhad better combination of high strength and ductility than that ofDP780 steel, it did not mean that the spot welded Q&P980 speci-mens had better fatigue properties than those of spot weldedDP780 specimens. These results agreed with the earlier reports[9,10,13] that the base material did not affect the fatigue perfor-mance of spot weld.

4. Conclusions

In this study, we have investigated the fatigue behaviors andfracture modes of Q&P980 spot weld specimens, and comparedthe fatigue properties of spot welded Q&P980 with those of spotwelded DP780. Based on the analysis and discussion above, the fol-lowing conclusions can be drawn:

(1) The hardness values of nugget and base material were equalto 497 HV and 334 HV. The hardness values of the nuggetand HAZ were higher than the base material due to the for-mation of martensite.

(2) The ultimate tensile forces of tensile-shear and cross-tensionspecimens were 23.7 ± 0.6 kN and 10.8 ± 0.5 kN, respec-tively. The failure modes of tensile tests for tensile-shearand cross-tension specimen were interfacial failure and pull-out failure, respectively.

(3) The conditional fatigue limit loads (107 cycles) of tensile-shear and cross-tension specimens were 1300 N and 280 N,respectively. For tensile-shear and cross-tension specimens,the fatigue cracks initiated near the notch tip in HAZ and atthe interface between two sheets; and the kinked angle offatigue cracks was nearly 75�.

(4) The fatigue cracks of tensile-shear specimens first propa-gated through the thickness, and then grew into base mate-rial along the direction of width up to the final fracture. Asfor cross-tension specimens, the fatigue cracks penetratedthe thickness until fracture as the applied load was higherthan 360 N; while the load was less than 360 N, the propaga-tion of fatigue cracks was similar to that of tensile-shearspecimens.

(5) For the similar geometric factor of spot welded Q&P980 andDP780, the fatigue properties of spot welded Q&P980 andDP780 specimens were approximately the same in the caseof tensile-shear and cross-tension specimens for both highand low fatigue load; and it indicated that the fatigue prop-erty of spot weld was independent of base material at leastfor the steels studied in this investigation.

Acknowledgements

Special thanks go to the Baoshan Iron & Steel Co., Ltd. for pro-viding all the spot weld specimens by authors, and go to Dr. L.L.Li and Dr. R.T. Qu for their assistance in SEM observations. Mean-while, the authors are thankful to Prof. X. P. Song for the assistancein XRD analysis.





References

[1] Matlock DK, Speer JG. Third generation of AHSS: microstructure designconcepts. In: Haldar A, Suwas S, Bhattacharjee D, editors. Microstructure andtexture in steels and other materials. Godalming: Springer-Verlag London Ltd.;2009. p. 185–205.

[2] Speer J, Matlock DK, Decooman BC, Schroth JG. Carbon partitioning intoaustenite after martensite transformation. Acta Mater 2003;51:2611–22.

[3] Nayak SS, Anumolu R, Misra RDK, Kim KH, Lee DL. Microstructure-hardnessrelationship in quenched and partitioned medium-carbon and high-carbonsteels containing silicon. Mater Sci Eng A 2008;498:442–56.

[4] Paravicini Bagliani E, Santofimia MJ, Zhao L, Sietsma J, Anelli E. Microstructure,tensile and toughness properties after quenching and partitioning treatmentsof a medium-carbon steel. Mater Sci Eng A 2013;559:486–95.

[5] Sun J, Yu H. Microstructure development and mechanical properties ofquenching and partitioning (Q&P) steel and an incorporation of hot-dippinggalvanization during Q&P process. Mater Sci Eng A 2013;586:100–7.

[6] Cerny I, Mikulová D, Sís J, Mašek B, Jirková H, Malina J. Fatigue properties of alow alloy 42SiCr steel heat treated by quenching and partitioning process.Procedia Eng 2011;10:3310–5.

[7] Wang L, Zhong Y, Feng WJ, Jin XY, Speer JG. Industrial application of Q&P sheetsteels. In: Proceedings, Intl. Symp. on new developments in advanced high-strength sheet steels. AIST; 2013. p. 141–51.

[8] Weber G, Thommes H, Gaul H, Hahn O, Rethmeier M. Mechanical properties ofweldbonded joints of advanced high strength steels. J Adhe Sci Technol2011;25:2369–89.

[9] Long X, Khanna SK. Fatigue properties and failure characterization of spotwelded high strength steel sheet. Int J Fatigue 2007;29:879–86.

[10] Rathbun RW, Matlock DK, Speer JG. Fatigue behavior of spot welded high-strength sheet steels. Weld J 2003;82:207–18.

[11] Vural M, Akkus A. On the resistance spot weldability of galvanized interstitialfree steel sheets with austenitic stainless steel sheets. J Mater Process Technol2004;153–154:1–6.

[12] Shariati M, Maghrebi MJ. Experimental study of crack growth behavior andfatigue life of spot weld tensile-shear specimens. J Appl Sci 2009;9:438–48.

[13] Xu J, Zhang YS, Lai XM, Chen GL. Experimental investigation of fatigueperformance of spot welded dual phase sheet steels. Sci Technol Weld Joining2008;13:726–31.

[14] Nakayama E, Fukumoto M, Miyahara M, Okamura K, Fujimoto H, Fukui K, et al.Evaluation of local fatigue strength in spot weld by small specimen. FatigueFract Eng Mater Struct 2010;33:267–75.

[15] Gould JE, Khurana SP, Li T. Predictions of microstructures when weldingautomotive advanced high-strength steels. Weld J 2006;85:111–6.

[16] Xu W, Westerbaan D, Nayak SS, Chen DL, Goodwin F, Zhou Y. Tensile andfatigue properties of fiber laser welded high strength low alloy and DP980dual-phase steel joints. Mater Des 2013;43:373–83.

[17] Ma C, Chen DL, Bhole SD, Boudreau G, Lee A, Biro E. Microstructure andfracture characteristics of spot-welded DP600 steel. Mater Sci Eng A2008;485:334–46.


[18] Farabi N, Chen DL, Li J, Zhou Y, Dong SJ. Microstructure and mechanicalproperties of laser welded DP600 steel joints. Mater Sci Eng A2010;527:1215–22.

[19] Hernandez VHB, Kuntz ML, Khan MI, Zhou Y. Influence of microstructure andweld size on the mechanical behaviour of dissimilar AHSS resistance spotwelds. Sci Technol Weld Joining 2008;13:769–76.

[20] Pal TK, Chattopadhyay K. Resistance spot weldability and high cycle fatiguebehavior of martensitic (M190) steel sheet. Fatigue Fract Eng Mater Struct2010;34:46–52.

[21] Biro E, Lee A. HAZ properties of various DP600 chemistries. Can Welding AssocJ 2005:17–21. Spring.

[22] Chao YJ. Ultimate strength and failure mechanism of resistance spot weldsubjected to tensile, shear, or combined tensile/shear loads. J Eng MaterTechnol 2003;125:125–32.

[23] Pouranvari M, Asgari HR, Mosavizadch SM, Marashi PH, Goodarzi M. Effect ofweld nugget size on overload failure mode of resistance spot welds. SciTechnol Weld Joining 2007;12:217–25.

[24] Pouranvari M. Susceptibility to interfacial failure mode in similar anddissimilar resistance spot welds of DP600 dual phase steel and low carbonsteel during cross-tension and tensile-shear loading conditions. Mater Sci EngA 2012;546:129–38.

[25] Radakovic DJ, Tumuluru M. Predicting resistance spot weld failure modes inshear tension tests of advanced high-strength automotive steels. Weld J2008;87:96–105.

[26] Chao YJ. Failure mode of spot welds: interfacial versus pullout. Sci TechnolWeld Joining 2003;8:133–7.

[27] Paris PC, Gomez MP, Anderson WP. A rational analytic theory of fatigue. TrendEng 1961;13:9–14.

[28] Pal TK, Bhowmick K. Resistance spot welding characteristics and high cyclefatigue behavior of DP 780 steel sheet. J Mater Eng Perform 2012;21:280–5.

[29] Hilditch TB, Speer JG, Matlock DK. Effect of susceptibility to interfacial fractureon fatigue properties of spot-welded high strength sheet steel. Mater Des2007;28:2566–76.

[30] Satoh T, Abe H, Nishikawa K, Morita M. On three-dimensional elastic-plasticstress analysis of spot-welded joint under tensile shear load. Trans Jpn WeldSoc 1991;22:46–51.

[31] Lanciotti A, Polese C. Fatigue crack propagation in tensile shear stainless steelspot welded specimens. Fatigue Fract Eng Mater Struct 2008;31:76–84.

[32] Hassanifard S, Bonab MAM, Jabbari Gh. Investigation of fatigue crackpropagation in spot-welded joints based on fracture mechanics approach. JMater Eng Perform 2013;22:245–50.

[33] Zhang Y, Taylor D. Sheet thickness effect of spot welds based on crackpropagation. Eng Fract Mech 2000;67:55–63.

[34] Hu X, Liu YF, Zhu L. Study and numerical analysis on formability of quenchingand partitioning steel sheets of auto-body. AIP Conf Proc 2013;1532:547–54.


http://refhub.elsevier.com/S0142-1123(14)00086-3/h0005






























































































investigation on fatigue fracture behaviors of spot welded q&p980 steel

Documents