Resistance spot welding of MS1200 martensitic advanced high strength steel: Microstructure-properties relationship

Journal of Manufacturing Processes 31 (2018) 867–874

Contents lists available at ScienceDirect

Journal of Manufacturing Processes journal homepage: www.elsevier.com/locate/manpro

Resistance spot welding of MS1200 martensitic advanced high strength steel: Microstructure-properties relationship M. Pouranvari ∗ , S. Sobhani, F. Goodarzi Department of Materials Science and Engineering, Sharif University of Technology, Tehran, Iran

a r t i c l e

i n f o

Article history: Received 5 December 2017 Received in revised form 5 January 2018 Accepted 11 January 2018 Keywords: Advanced high strength steel Martensitic steels Resistance spot welding Microstructure HAZ softening Failure mode

a b s t r a c t This paper addresses the microstructure and tensile-shear mechanical performance of MS1200 Gigagrade martensitic advanced high strength steel resistance spot welds. The key phase transformations in MS1200 welds were lath martensite formation in the fusion zone (FZ) and upper-critical heat affected zone (HAZ), new ferrite formation in the inter-critical HAZ and martensite tempering in the sub-critical HAZ. The MS1200 welds were featured by a near matching hardness in the fusion zone and undermatching hardness in the heat affected zone (HAZ) compared to the base metal. At certain process window a complete nugget pullout and separation was observed with high post-necking tearing energy. The interfacial to pullout failure mode transition was explained in the light of FZ hardness as well as the HAZ softening associated with martensite tempering in the sub-critical HAZ. The load bearing capacity of MS1200 welds failed at interfacial mode was strongly depends on the FZ size as well as the FZ hardness. However, the peak load of welds failed at pullout mode was a function of HAZ softening as well as the plastic constraint in the HAZ associated with the hard upper-critical HAZ/FZ and martensitic BM. © 2018 The Society of Manufacturing Engineers. Published by Elsevier Ltd. All rights reserved.

1. Introduction Advanced high strength steels are key materials for lightweight design strategies in automotive industry to achieve energy conservation, improvement of safety, and crashworthiness qualities [1,2]. The structural reinforcement components in vehicle’s safety cage composed of pillars, side sills, rockers, door reinforcement beams, roof rails and floor and roof cross members require Giga-grade steels with tensile strength more than 1 GPa. Therefore, extremely high strength steels, typically martensitic steels and press hardening grades, are one of the best promising candidates for the use in these components [3,4]. Resistance spot welding (RSW), which is the key joining process in automotive industry, plays critical role in vehicle manufacturing [5]. Body-in-white of the automotives contains several thousands spot welded joints depending on the size of the vehicle and the number of parts that need to be joined in combination and the joining strategy of the manufacturer [6]. Therefore, the mechanical performance of the RSWs plays key role in crashworthiness of the vehicle, the capability of a car structure to provide adequate protection to its passengers against injuries in the event of a crash [7]. Moreover, the failure of spot welds may affect the vehicle’s stiffness and NVH (Noise, Vibration and Harshness) performance on a

∗ Corresponding author. E-mail address: [email protected] (M. Pouranvari).

global level [8]. Therefore, the quality, performance and the failure characteristics of resistance spot welds (RSWs) are important for determination of durability and safety design of the vehicles. Hence, a fundamental knowledge of the failure process of the resistance spot welds is required to achieve sound, strong and reliable welds. Failure mode of resistance spot welds (RSWs) is a qualitative measure of mechanical properties [1–16]. The avoidance of interfacial failure (i.e. fracture through the fusion zone) is one of the key requirement for resistance spot welds in safety critical areas of the vehicle [5,6]. Generally, the pullout failure mode in which the failure occurs via withdrawal of the weld nugget from one sheet exhibits the most satisfactory mechanical properties. The pullout failure mode during quality control indeed indicates that the same weld would have been able to transmit a high level of force, thus cause severe plastic deformation in its adjacent components, and increased strain energy dissipation in crash conditions [17]. Therefore, it is needed to adjust welding parameters so that the pullout failure mode is guaranteed. Past researches have shown that the resistance spot welds made on AHSS steels display high susceptibility to fail in interfacial mode [5,10,13,17]. Complexity of failure process in AHSS spot welds is originated from complex weld phase transformation including maretnite formation in the fusion zone (FZ) and softening in the heat affected zone (HAZ) [1–20] and segregation phenomena [21,22]. It is shown that the IF to PF failure mode transition is largely depend on the complex interplay between weld geometry, fusion zone/HAZ/base metal properties, test geometry,

https://doi.org/10.1016/j.jmapro.2018.01.009 1526-6125/© 2018 The Society of Manufacturing Engineers. Published by Elsevier Ltd. All rights reserved.

868

M. Pouranvari et al. / Journal of Manufacturing Processes 31 (2018) 867–874

Table 1 Chemical composition and mechanical properties of investigated MS1200 steel. Chemical composition (%wt.) C 0.09

Mn 1.5

Si 0.18

Cr 0.04

Mechanical property Ni 0.04

YS (MPa) 1240

UTS (MPa) 1330

El (%) 5

YS: Yield Strength, UTS: Tensile Strength, El: Elongation.

shear test are illustrated in Fig. 1b.1 The tensile-shear tests were performed at a cross head of 10 mm/min. Failure modes were determined by observing the weld fracture surfaces. Mechanical performance of the welds is described in terms of peak load (Pmax ) and failure energy (Wmax ). The Wmax was digitally calculated by measuring the area under the load-displacement curve up to the peak point. The data points for peak load and failure energy are average of two specimens. Optical microscopy and scanning electron microscopy (SEM) was used to examine the structure of the spot welds at both macro and micro scales. Samples for metallographic examination were prepared using standard metallography procedure and were etched using Nital solution. Vickers micro-hardness test was performed using an indenter load of 100 g for a period of 20 s to obtain hardness values in different zones of the weldment. To increase the accuracy in reading of the indentations sizes, they were measured using image analyzer software (ImageJ) under optical microscopy. 3. Results and discussion 3.1. Tensile-shear properties

Fig. 1. (a) Schedule of resistance spot welding used in this study for joining MS1200 martensitic steel sheets, (b) schematic of tensile-shear samples.

and the stress state in each weld [1–25]. Therefore, a fundamental knowledge of microstructural evolutions during welding of AHSS automotive steels is vital to produce sound, strong and reliable joints and to predict the failure behavior of welded joint. Compared to other AHSS types, the welding behavior of martensitic grades are less investigated [26–29] and therefore further clarifications on structure-properties relationships in martensitic AHSS steel resistance spot welding are required. This paper aims at understanding the microstructure and mechanical properties of the MS1200 martensitic advanced high strength steel during resistance spot welding.

2. Experimental methods This study concerned the joining of uncoated cold rolled MS1200 martensitic AHSS (Docol 1200 M) sheet using resistance spot welding. Table 1 shows the chemical composition and the tensile properties of the steel. The sheet thickness was 1.5 mm. Resistance spot welding was performed using a 120 kVA AC pedestal type resistance spot welding machine operating at 50 Hz controlled by a programmable logic controller (PLC). Welding was conducted using a 45-deg truncated cone RWMA (resistance welding manufacturing alliance) class 2 electrode with 8-mm face diameter. Fig. 1a shows the welding schedule. In this work, the effect of welding current, as the main variable of RSW, was studied on the microstructure/properties of the welds. Therefore, welding current was incrementally increased from 7 to 14 kA with a step size of 1 kA. Three samples Mechanical performance of the welds was evaluated using quasi-static tensile-shear testing. The dimensions for the tensile-

Results showed that the welding current has a profound effect on the load-displacement characteristics of MS1200 resistance spot welds (Fig. 2a). Three distinct types of load-displacement curves were observed as a function of welding current. To describe the mechanical performance of the welds two parameters were extracted from the load-displacement curves: peak load (i.e. the peak point at load-displacement curve) and failure energy (i.e. the area under load-displacement curve up to the peak point). The peak load (Pmax ) represents the maximum force sustainable by the spot welds before failure initiation and the failure energy (Wmax ) represents the energy absorption capability of the welds before failure initiation. Fig. 2b shows the effect of welding current on the Pmax and Wmax indicating that increasing welding current up to 11 kA enhances both load bearing capacity and energy absorption capability. However, increasing welding current beyond 11 kA does not improve the mechanical properties of the welds. According to AWS D8.7:2005 [30], the minimum acceptable peak load for 1.5 mm MS1200 spot welds is 24.9 kN. Therefore, according to Fig. 2b, the minimum acceptable welding current is 10 kA. Four failure modes were observed during the tensile-shear loading of the MS1200 weld, as illustrated in Fig. 3. The failure mode was a function of welding current. Effect of welding current on failure mode is indicated in Fig. 2b. Welds made using current lower than 10 kA were failed in full interfacial failure (IF) mode. Joints made using welding current of 10 kA was failed at partial interfacial accompanied with partial thickness-partial pullout (PT-PP). Based on the first failure path, this failure mode was categorized an interfacial fracture. Increasing welding current to 11 kA changes the failure mode from IF to complete double side nugget pullout and separation. However, upon increasing welding current to 13 kA and more, where welding is accompanied with severe expulsion, the failure mode was changed to PT-PP. Each failure mode exhibits a characteristics load-displacement curve (Fig. 2a). The failure mode affects the shape of tail in load-displacement curves. In IF mode, the load suddenly dropped to zero due to the rapid progression of the failure process. In complete nugget pullout and separation mode, the peak point corresponds to the necking of the nugget

1 It should be noted that the width of test sample in this study (i.e. 30 mm) is less than that of recommended by AWS D8.7 standard (i.e. 60 mm) [31]. The narrow specimens used in this study provide lower restraint to the weld, and therefore, a weld failing in pull-out mode in such specimens may fail in interfacial mode if the specimens are wider [8].

M. Pouranvari et al. / Journal of Manufacturing Processes 31 (2018) 867–874

869

Fig. 2. (a) Load-displacement characteristics of MS1200 resistance spot welds made on various welding currents (b) Effect of welding current on Pmax and Wmax and failure mode. IF: interfacial failure, CPS: complete pullout and separation, PP-PT: partial pullout-partial thickness. Expulsion experienced welds are marked.

Fig. 3. Typical failure mode observed in this study: (a) interfacial failure mode observed when welding current is less than 10 kA, (b) partial interfacial failure accompanied with pullout-partial thickness mode observed when is welding current is 10 kA, (c) complete bottom separation mode observed when welding current is 11 and 12 kA, (d) partial pullout-partial thickness mode re-observed when welding is accompanied with expulsion at welding current higher than 12 kA.

circumference. In this failure mode, the load-displacement curve has a long t¨ ail¨corresponding to the post necking elongation due to the continued necking of the nugget circumference and complete tearing and shearing of the base metals. In PP-PT mode, similar to pullout mode, the peak point corresponds to the necking of the nugget circumference. However, as the failure progresses, the crack redirected through the FZ as slant crack resulted in some part of mating sheet thickness is removed upon final separation, as indicated by the change in the slope of load-displacement curve in the tail section. The tendency to fail in PP-PT in expulsion experienced welds has been also reported in the other works [31,32]. This can be related to the increased indentation and its associated stress concentration in the through thickness direction which can promotes the crack redirection toward FZ in thickness direction. Among these three failure types, the failure under complete nugget pullout and separation exhibited the best mechanical properties. The process of nugget pullout and separation failure is accompanied with higher plastic deformation than traditional pullout failure mode. This fact contributes to the total energy absorption of the spot welds (i. e. the area under the load-displacement curve up to the final fracture

point). This can increase the strain energy dissipation in crash conditions. This type of failure mode has been only reported during impact testing of a high strength steel spot weld [33]. Generally, the mechanical properties of the spot welds are controlled mainly by two distinct factors, geometrical factors (i.e. FZ size and electrode indentation) and metallurgical characteristics of the weld. How the geometrical/metallurgical attributes of the weld affect the mechanical properties of the RSWs is a function of failure mode. Here, for simplicity, only two distinct failure modes was considered: (i) The peak load of spot welds in IF mode can be estimated using following formula [5]: PIF = fIF D2 FZ

(1)

Where, D is the FZ, ␶FZ is the shear strength of the FZ and fIF is a constant. (ii) The peak load of spot welds in PF mode can be estimated using following formula [5]: PPF = fPF D t PFL

(2)

870

M. Pouranvari et al. / Journal of Manufacturing Processes 31 (2018) 867–874

Fig. 4. Metallurgical changes induced by resistance spot welding in MS1200 martensitic advanced high strength steels (a) a typical macrostructure (b) base metal (BM) showing very fine martensite packets (c) fusion zone (FZ) showing large martensite packets, (d) microstructure of high temperature upper-critical heat affected zone, (e) microstructure of low temperature upper-critical heat affected zone, (f) microstructure of inter-critical heat affected zone showing ferrite-martensite dual phase microstructure, (g) microstructure sub-critical heat affected zone showing partial martensite decomposition due to tempering.

Where, ␴PFL is the ultimate tensile strength of the PF location and fPF is a constant. According to Eq. (1), the PIF depends on FZ size and shear strength of the FZ. The latter is a function of fusion zone hardness. According to Eq. (2), the PPF depends on FZ size, sheet thickness and tensile strength of the pullout failure location (␴PFL ). The pullout failure location depends on the hardness characteristics of the weldment. Therefore, to analyze the mechanical behavior of spot

welds both geometrical and metallurgical factors should be considered. 3.2. Macro/micro scale metallurgical characteristics of the MS1200 RSW Fig. 4a shows a typical macrostructure of MS1200 resistance spot welds. As can be seen three distinct microstructural zones

M. Pouranvari et al. / Journal of Manufacturing Processes 31 (2018) 867–874

871

Fig. 5. FZ size and critical FZ size to avoid interfacial failure mode versus welding current. 4t0.5 recommendation is also superimposed to the plot indicating that sizing based on this recommendation is not sufficient for obtaining pullout failure mode during tensile-shear loading of MS100 resistance spot welds.

were created in the weld and its surrounding consisting of (i) fusion zone (FZ) or weld nugget which is melted and re-solidified during welding process showing a cast structure, (ii) Heat affected zone (HAZ) which is not melted but undergoes microstructural changes during welding and (iii) base metal (BM) which does not show any metallurgical modifications. From geometrical view of point, the FZ size is the most important parameters controlling the mechanical performance of the RSW, as it determines the overall bonding area of the joint. Weld nugget or fusion zone size (D) which is defined as the width of the weld nugget at the sheet/sheet interface in the longitudinal direction, was measured for each welding condition. Fig. 5 shows the effect of welding current on the FZ size. The trend is predictable based on Joule heating effect. Fig. 4b shows the microstructure of the MS1200 base metal indicating fine martensitic structure. Fig. 4c shows the SEM micrograph showing FZ microstructure of MS1200 indicating an almost martensitic structure. According to Fe-Fe3 C phase diagram (not shown here), the solidification/transformation path during RSW of the FZ can be summarized as follows: L→F →A→M

(3)

The formation of martensite depends on steel hardenability and the cooling rate. The formation of martensite in the FZ of MS1200 can be rationalized based on the following. The critical cooling rate to achieve martensite in the steel microstructure can be estimated using the following equation [34]: Log CCR = 7.42 − 3.13 C − 0.71 Mn − 0.37 Ni − 0.34 Cr − 0.45 Mo (4) where, CCR is the critical cooling rate in Kh−1 and the element symbols refer to their concentrations in weight percent. Using Eq. (4), the CCR for MS1200 is calculated as 247 ◦ Cs−1 . According to the simple analytical model of Gould et al. [35], the cooling rate of 1.5 mm thick steel RSW is about 4000 ◦ Cs−1 . The cooling rate experienced by FZ is well above the CCR of MS1200 steel. Therefore, martensite formation in FZ is not surprising. Compared to the FZ, there is a large microstructural gradient in the HAZ. The microstructural evolution in the HAZ depends on how the microstructure of the base metal interacts to the welds thermal cycle. Therefore, the experienced peak temperature plays critical role in determining microstructure of the HAZ. The critical temperature of A1 and A3 for MS1200 are calculated using well-

Fig. 6. Hardness characteristics of MS1200 resistance spot welds: (a) hardness profile showing two soft zones in the HAZ, yellow highlighted zone is ICHAZ and the orange highlighted zone is SCHAZ, (b) effect of welding current on the hardness of the FZ and minimum hardness of the SCHAZ.

known equations2 ([36]) as 712 and 845 ◦ C, respectively. Therefore, the HAZ can be divided into three distinct regions, based on the peak temperature: • Upper-critical HAZ (UCHAZ): The peak temperatures in this region are above Ac3 and lower than the melting point of the steel. Therefore, the BM undergoes full austenitization. The subsequent transformation during cooling depends on the cooling rate and the steel hardenability. The steel hardenability is a function of steel chemistry and primary austenite grain size. The latter is a function of the peak temperature in the upper-critical region. Therefore, the UCHAZ can be divided to two distinct zones: coarse-grained HAZ (CGHAZ) and fine-grained HAZ. Fig. 4 (d–e) shows the microstructure of UCHAZ at two different location, CGHAZ and FGHAZ, respectively. As can be seen, despite their different primary austenite grain size, both zone exhibited full martensitic microstructure. This is due to combined effect of high cooling rate during RSW and high composition-related hardenability of the MS1200 steel. • Intercritical HAZ (ICHAZ): The peak temperatures in this region are ranging between Ac1 and Ac3 . Therefore, the BM undergoes partial austenitization during heating (i.e. the BM microstructure transforms into ferrite plus austenite). Due to high hardenability of intercritical austenite and the fast RSW cooling rates, austenite is transformed into martensite, as it is evidenced by observation a ferrite-martensite two phase regions in the ICHAZ, Fig. 4f. • Sub-critical HAZ (SCHAZ): In the sub-critical HAZ, the peak temperature is below the Ac1 . The most prominent feature in this region is tempering of the initial martensite. Fig. 4g shows the microstructure of SCHAZ. The martensite decomposition followed by broken lathy appearance and sub-micron particles carbide precipitation evident in this region indicating tempering of martensite [19,37]. It should be noted that accurate characterization of the sub-micron particles requires detailed transmission electron microscopy (TEM) study.

2

Ac1 (◦ C)= 751-16.3C-275Mn-5.5Cu-5.9Ni+349Si+12.7Cr+3.4Mo

872

M. Pouranvari et al. / Journal of Manufacturing Processes 31 (2018) 867–874

Fig. 7. Peak load and failure energy vs. FZ size: welds failed in the complete pullout and separation and partial-thickness/partial-pullout (PP-PT) are highlighted.

Fig. 9. Peak load versus FZ size of welds failed at pullout/partial-pullout mode. The predictions based on the Eq. (2) (without consideration of HAZ softening) and Eq. (10) (with consideration of HAZ softening) are also superimposed. A range of fPF (2.5–3.4) was considered to predict peak load.

SCHAZ is significantly higher than the full tempered martensite. As suggested by Biro [38], the hardness of the fully tempered martensite (HTM ) can be calculated using the volume weighted average hardness of the ferrite (H␣ ) and cementite (HFe3C ), as follows: HTM = (1 −

Fig. 8. Correlation between load bearing capacity of MS1200 spot welds failed in 2 D. interfacial mode andHFZ

3.3. Hardness characteristics Fig. 6a shows a typical Vickers microhardness profile across the weldment of the MS1200 RSW. The average microhardness of the BM was 405 HV which is consistent with martensitic microstructure of the base metal. As can be seen, the microhardness profile of the welds is featured by formation of soft zones, where the average micro-hardness value is significantly lower than the BM. Examination of location of indentation revealed that the soft zones are associated with ICHAZ and SCHAZ. The local martensite tempering in the HAZ in the SCHAZ and formation of ferrite in the ICHAZ are responsible for HAZ softening in MS1200 RSW. It is of note that the SCHAZ is wider than the ICHAZ. The relative size of the HAZ subzones depends on the temperature range corresponding to each zone as well as heat transfer phenomena during welding. For example, the size of ICHAZ is significantly lower than the SCHAZ. The temperature range in ICHAZ is from 712 ◦ C (i.e. Ac1 ) to 845 ◦ C (i.e. Ac3 ). However, the temperature range in SCHAZ is from the 400 ◦ C (considered as the effective tempering temperature) to 712 ◦ C (i.e. Ac1 ). According to metallographic cross-section of welds (for example see Fig. 4a), the ICHAZ size ranges from about 100 to 140 ␮m and SCHAZ size ranges from 520 to 750 ␮m. The extent of HAZ softening is a function of the experienced peak temperature during welding. The minimum hardness value in the SCHAZ hardness corresponded to where experienced peak temperature of AC1 during welding, where more pronounced martensite tempering is expected. It should be noted that the minimum SCHAZ hardness was a function of welding current. According to Fig. 6b, the HAZ softening became more pronounced as the welding current increased. It is of note that the minimum hardness of the

C C )H˛ + ( )HFe3 C 6.67 6.67

(5)

Where, the average hardness values for ferrite and cemenite are 80 HV and 1270 HV [38], respectively. Eq. (5) account for the role of C, as no contribution to hardness has been reported for solution strengthening of Cr, Mo and less than 2% Mn. According to Eq. (5) and the carbon content of the base metal (i.e. 0.098 wt.%), the hardness of the full tempered martensite can be calculated as 97 HV. However, the minimum hardness in the SCHAZ for MS1200 welds is about 205 HV which is well above the full tempered martensite. This confirms that tempering of the initial martensite in the base metal was occurred partially due to rapid heating and cooling cycle of the RSW process. In the inter-critical HAZ, there is a sharp increase of the hardness, which is up to the hardness of UCHAZ (see Fig. 5a). The hardness of the ICHAZ depends on the volume fraction of martensite which in turn in controlled by the experienced peak temperature during welding. As can be seen, the hardness of ICHAZ is lower than the BM due to formation of new ferrite. Increasing the temperature in the inter-critical regions increases the volume faction of austenite which subsequently will be transformed to martensite upon rapid cooling. The hardness values in the UCHAZ and FZ are relatively uniform and equivalent to the hardness in the BM. This is due to the fact that both UCHAZ and FZ experienced full austenitization during heating and transformed to martensite during cooling. The hardness of the martensite (HM ) can be estimated using the following Equation [39]: HM = 127 + 949C + 27Si + 11Mn + 8Ni + 16Cr + 21logV Kh−1

(6)

and the element symbols Where V is the cooling rate in refer to their concentrations in weight percent. The 95% confidence limits are ± 26 HV. Using Eq. (6), the expected hardness value for the full martensitic MS1200 steel is found to be 395 ± 26 HV which is consistent with hardness of FZ and UCHAZ. It is of note that the hardness of the FZ is slightly higher than that of the BM. This can be related to the light-tempering of the base metal during its manufacturing process. Moreover, as can be seen in Fig. 5a, despite their same microstructure, the hardness of high-temperature UCHAZI (averaged at 414 HV) is lower than low-temperature UCHAZII (aver-

M. Pouranvari et al. / Journal of Manufacturing Processes 31 (2018) 867–874

aged at 442 HV). This can be explained in terms of martensite packet size. As can be seen in Fig. 4(d and e), the martensite packet size in UCHAZII is significantly finer than that of in UCHAZI which can explain the slight higher hardness values in the UCHAZII . According to Fig. 6b, the hardness of the FZ is independent of the welding heat input which confirms that the variation of heat input and the associated cooling rate does not influence the phase transformation in the FZ. 3.4. Microstructure-properties relationship in MS1200 resistance spot welds 3.4.1. Failure mode transition Due to its significant impact on the joint reliability, the failure mode is an important issue for automotive crashworthiness. The transition from IF mode to PF mode is generally related to the increase in the size of FZ above a minimum value. According to Fig. 5, sizing based on 4t0.5 rule does not guarantee pullout failure mode. Therefore, it seems that the metallurgical factors should be considered when analyzing the failure mode of spot welds. Failure mode of spot welds can be considered as a competition between ductile shear failure of the FZ and necking of the nugget circumference. The failure occurs by a mechanism which needs less force to happen. Therefore, pullout failure will be preferred mode if the following conditions meet [5,10]: (PPF )Eq.2 ≺ (PIF )Eq.1 ⇔ Pullout Failure

(7)

Therefore, D≥t

fPF PFL fIF FZ

(8)

According to theoretical calculations, fIF and fPF are ␲/4 and ␲ [5,10]. Assuming linear relationships between tensile strength and shear strength [40] and between tensile strength of the materials and their Vickers hardness [41], the condition to avoid interfacial mode can be written as follows [5,10]: D≥

4t HPFL ( ) m HFZ

(9)

where, t is the sheet thickness (mm), m is the ratio of shear strength to tensile strength of the FZ (it has been shown experimentally that f can be approximated by Von Miss Criterion [42]), HFZ and HPFL are hardness values (HV) of the fusion zone and pullout failure location, respectively. The pullout failure location during the tensile-shear loading is where the lowest hardness has [5,43]. The minimum hardness for MS1200 occurs at SCHAZ. According to this model, IF to PF transition can be explained in terms of the FZ size and the changes in FZ and SCHAZ hardness with welding current. According to Fig. 6b, the hardness ratio of FZ to SCHAZ increases as the welding current increases due to formation a softer sub-critical HAZ. The critical FZ size (DC ) corresponding to each welding current is also calculated according to Eq. (9). The calculated DC is superimposed to Fig. 5, where the variation of FZ size with welding current is also shown. As can be seen in Fig. 5, in welding current lower than 11 kA, the FZ size is lower than the DC , consequently the welds were failed in IF mode. On the other hand, welds made at welding current equal or higher than 10 kA exhibit higher FZ size that the DC and were failed in PF mode (or partial pullout). Indeed, the transition in failure mode from IF to PF can be well correlated to the (i) increasing FZ size caused by increasing welding current and (ii) increasing the hardness ratio of FZ to sub-critical HAZ via encouraging martensite tempering and therefore reduction of HAZ hardness. Heat affected zone softening reduced the strength of the HAZ and resulted in strain localization and hence encouraged the PF mode.

873

3.4.2. Factors affecting the mechanical response Fig. 7 shows the effect of FZ size on Pmax and Wmax of the spot welds. The following points can be drawn from this figure: (i) When the FZ size is smaller than the DC (where the welds failed in IF mode), there is strong correlation between FZ size and weld mechanical properties. This can be explained by the fact that load bearing area during interfacial failure mode is proportional to square of FZ size (Eq. (1). Fig. 8 shows that there is direct relationship between Pmax versus D2 HFZ for welds failed at interfacial mode confirming that load bearing capacity of welds during interfacial failure is dictated by square of FZ size as well as FZ hardness. (ii) When the FZ size is larger than DC (where the welds failed in PF mode), the Pmax does not changed significantly with the FZ size. According to Eq. (2), the peak load in PF mode is linearly proportional to FZ size. Therefore, it is expected that increasing FZ size from 6.8 to 8.3 mm should accompany with 22% increase in peak load. However, as can be seen from Fig. 7, the peak load is nearly same for all welds made using welding current ranging from 11 to 15 kA. This can be related to the effect of HAZ softening on the joint strength. The effect of HAZ softening on the joint strength can be estimated using Eq. (2). When spot welds exhibit no softening in the HAZ and the pullout failure occurs at the BM, the peak load in the PF mode can be estimated using the well known empirical relation in the form ofFPF = f t DBM , where ␴BM is the tensile strength of the BM and f is material-dependent coefficient which has been reported to be in the range of 2.5–3.4 [5]. When spot welds exhibit softening in the HAZ, the pullout failure occurs at the SCHAZ, therefore, the peak load in the PF mode can be estimated usingFPF = f t DSCHAZ , where ␴SCHAZ is the tensile strength of the SCHAZ. The ␴SCHAZ was calculated using softening ratio (i.e. ratio of minimum hardness in the SCHAZ to the BM hardness). Therefore, the peak load of spot welds accounting for HAZ softening can be calculated using following formula: SCHAZ = fPF DtBM ×

min HSCHAZ

HBM

(10)

min are base metal hardness and the minimum Where, HBM and HSCHAZ hardness in the SCHAZ, respectively. Fig. 9 shows the predicted Pmax for MS1200 with FZ size of 6.8–8.3 mm without consideration of HAZ softening (Eq. (2) and with consideration of HAZ softening (Eq. (10). Also, the experimental data are superimposed to this plot. As can be seen, Eq. (2) could not accurately predict the load bearing capacity of the welds. This confirms that the peak load of MS1200 welds failed in PF mode is not governed by the BM strength and is significantly affected by the HAZ softening. However, as can be seen in Fig. 9, the measured peak loads of the spot welds is somewhat higher than those predicted using Eq. (10). This can be related to the constraint effect of the hard BM and the hard UCHAZ and FZ on the plastic deformation of the soft HAZ. It is reported that this constraint effect can lower the detrimental effect of the HAZ softening on the weld load bearing capacity [44,45]. Therefore, it can be concluded that the mechanical properties of the MS1200 welds failed at the PF mode is governed by both softening phenomena and the plastic constraint in the HAZ.

4. Conclusions Martensitic steel (MS) are designed to consist of hard martensite, which offers a high tensile-strength with unique performance in anti-intrusion structural reinforcement components in vehicle’s safety. However, the resistance spot welding, as the critical sheet metal joining in body-in-white manufacturing, produces significant microstructural changes in a narrow heat-affected zone (HAZ) and fusion zone (FZ), which can result in dramatic changes in mechanical properties. The most prominent metallurgical fea-

874

M. Pouranvari et al. / Journal of Manufacturing Processes 31 (2018) 867–874

ture of MS1200 martensitic steel spot welds was the HAZ softening which led to producing a local under-match across the weld zone. Both ferrite formation in inter-critical HAZ and martensite tempering in the sub-critical HAZ are responsible for HAZ softening in MS1200 martensitic steel. However, considering the larger area of SCHAZ compared to the ICHAZ, the martensite tempering plays more important role in mechanical properties of the welds. Despite the fact the load bearing capacity of spot welds failed in interfacial mode was not affected by the HAZ softening phenomena, it significantly affect the peak load of welds failed in pullout mode. The loss of joint strength due to HAZ softening was a function of minimum hardness of the SCHAZ as well as the plastic constraints created by the presence of two hard zones (UCHAZ/FZ and BM) surrounding the soft HAZ. The latter reduces the effect of HAZ softening on the joint strength. Acknowledgements The authors would like to acknowledge the SSAB Company for providing the raw material, Docol M1200 steel, for this research. References [1] Tasan CC, Dieh M, Yan D, Bechtold M, Roters F, Schemmann L, Zheng C, Peranio N, Ponge D, Koyama M, Tsuzaki K, Raabe D. An overview of dual-phase steels: advances in microstructure-oriented processing and micromechanically guided design. Annu Rev Mater Res 2015;45:391–431. [2] Pouranvari M. Critical assessment: dissimilar resistance spot welding of aluminium/steel: challenges and opportunities. Mater Sci Technol 2017;33:1705–12. [3] Zuidema BK. Bridging the design–manufacturing–materials data Gap: material properties for optimum design and manufacturing performance in light vehicle steel-intensive body structures. JOM 2012;64:1039–47. [4] Tamarelli CM. The evolving use of advanced high-strength steels for automotive applications. International Iron and Steel Institute; 2011. [5] Pouranvari M, Marashi SPH. Critical review of automotive steels spot welding: process, structure and properties. Sci Technol Weld Join 2013;18:361–403. [6] Uijl, Resistance Spot Welding of Advanced High Strength Steels, Phd Thesis. [7] Yang YP, Gould J, Peterson W, Orth F, Zelenak P, Al-Fakir W. Development of spot weld failure parameters for full vehicle crash modelling. Sci Technol Weld Joining 2013;18:222–31. [8] Zhang H, Senkara J. Resistance welding: fundamentals and applications. Taylor & Francis CRC Press; 2005. [9] Ma C, Chen DL, Bhole SD, Boudreau G, Lee A, Biro E. Microstructure and fracture characteristics of spot-welded DP600 steel. Mater Sci Eng A 2008;485:334–46. [10] Pouranvari M, Marashi SPH. Failure mode transition in AHSS resistance spot welds. part I. controlling factors. Mater Sci Eng A 2011;528:8337–43. [11] Pouranvari M, Marashi SPH, Safanama DS. Failure mode transition in AHSS resistance spot welds. part II: experimental investigation and model validation. Mater Sci Eng A 2011;528:8344–52. [12] Pouranvari M. Fracture toughness of martensitic stainless steel resistance spot welds. Mater Sci Eng A 2017;680:97–107. [13] Sun X, Stephens EV, Khaleel MA. Effects of fusion zone size and failure mode on peak load and energy absorption of advanced high-strength steel spot welds. Weld J 2007;86:18s–25s. [14] Pouranvari M. Susceptibility to interfacial failure mode in similar and dissimilar resistance spot welds of DP600 dual phase steel and low carbon steel during cross-tension and tensile-shear loading conditions. Mater Sci Eng A 2012;546:129–38. [15] Moharrami R, Hemmati B. Numerical stress analysis in resistance spot-welded nugget due to post-weld shear loading. J Manuf Processes 2017;27:284–90. [16] Chen J, Yuan X, Hu Z, Li T, Li C. Improvement of resistance-spot-welded joints for DP 600 steel and A5052 aluminum alloy with Zn slice interlayer. J Manuf Processes 2017;30:396–405. [17] Marya M, Wang K, Hector LG, Gayden X. Tensile-shear forces and fracture modes in single and multiple weld specimens in dual-phase steels. J Manuf Sci Eng 2006;128:287–98. [18] Dancette S, Massardier-Jourdan V, Fabregue D, Merlin J, Dupuy T, Bouzekri M. HAZ microstructures and local mechanical properties of high strength steels resistance spot welds. ISIJ Int 2011;51:99–107.

[19] Baltazar Hernandez VH, Nayak SS, Zhou Y. Tempering of martensite in dualphase steels and its effects on softening behavior. Metall Mater Trans A 2011;42:3115–29. [20] Pouranvari M, Marashi SPH. Key factors influencing mechanical performance of dual phase steel resistance spot welds. Sci Technol Weld Join 2012;15:149–55. [21] Krajcarz F, Gourgues-Lorenzon A-F, Lucas E. Influence of carbon content on the primary solidification mode of high strength steels in resistance spot welding conditions. Scr Mater 2016;120:98–102. [22] Amirthalingam M, van der Aa EM, Kwakernaak C, Hermans MJM, Richardson IM. Elemental segregation during resistance spot welding of boron containing advanced high strength steels. Weld World 2015;59:43–755. [23] Krajcarz F, Gourgues-Lorenzon A, Lucas E, Pineau A. Fracture toughness of the molten zone of resistance spot weld. Int J Fract 2013;181:209–26. [24] Radakovic D, Tumuluru M. Predicting resistance spot weld failure modes in shear tension tests of advanced high-strength automotive steels. Weld J 2008;87:96. [25] Sawanishi C, Ogura T, Taniguchi K, Ikeda R, Oi K, Yasuda K, et al. Mechanical properties and microstructures of resistance spot welded DP980 steel joints using pulsed current pattern. Sci Technol Weld Joining 2014;19:52–9. [26] Jong Y, Lee Y, Kim D, Kang M, Hwang I, Lee W. Microstructural evolution and mechanical properties of resistance spot welded ultra high strength steel containing boron. Mater Trans 2011;52:1330–3. [27] Safanama DS, Marashi SPH, Pouranvari M. Similar and dissimilar resistance spot welding of martensitic advanced high strength steel and low carbon steel: metallurgical characteristics and failure mode transition. Sci Technol Weld Join 2012;17:288–94. [28] Lopez-Cortez VH, Reyes-Valdes FA. Understanding resistance spot welding of advanced high-strength steels. Weld J 2008;87(12):36–40. [29] Tamizi M, Pouranvari M, Movahedi M. Welding metallurgy of martensitic advanced high strength steels during resistance spot welding. Sci Technol Weld Join 2017;22:327–35. [30] AWS D8.7M. Recommended practices for automotive weld Quality— resistance spot welding. American Welding Society (AWS); 2005. [31] Pouranvari M, Marashi SPH, Mousavizadeh SM. Failure mode transition and mechanical properties of similar and dissimilar resistance spot welds of DP600 and low carbon steels. Sci Technol Weld Join 2010;15:625–31. [32] Pouranvari M, Marashi SPH, Mousavizadeh SM. Dissimilar resistance spot welding of DP600 dual phase and AISI 1008 low carbon steels: correlation between weld microstructure and mechanical properties. Ironmaking Steelmaking 2013;38(6):471–80. [33] Karve GS, Zhang H. Dependence of impact performance on process parameters and weld attributes for spot-welded advanced high strength steels Anaheim, California USA. Proceedings of IMECE04 2004 ASME International Mechanical Engineering Congress and Exposition November 2004:13–20. [34] Easterling KE. Modelling the weld thermal cycle and transformation behavior in the heat-affected zone. In: Cerjak H, Easterling KE, editors. Mathematical modelling of weld phenomena. The Institute of Materials; 1993. [35] Gould JE, Khurana SP, Li T. Predictions of microstructures when welding automotive advanced high-strength steels. Weld J 2006;85:111s–6s. [36] Andrews KW. Empirical formulae for the calculation of some transformation temperatures. JISI 1965;203:721–7. [37] Bhadeshia HKDH, Honeycombe R. Steels: microstructure and properties. 3rd edn Oxford: Butterworth-Heinemann; 2006. [38] Vignier S, Biro E, Hervé M. Predicting the hardness profile across resistance spot welds in martensitic steels. Weld World 2014;58:297–305. [39] Blondeau R, Maynier P, Dollet J, Vieillard-Baron B. Mathematical model for the calculation of mechanical properties of low alloy steels metallurgical products: a few examples of its applications. In: Proc Int Conf Heat Treatments’76. London: Metals Society; 1976. [40] Hertzberg RW. Deformation and fracture mechanics of engineering materials. 2nd ed John Wiley & Sons; 1996. [41] Zhang P, Li SX, Zhang ZF. General relationship between strength and hardness. Mater Sci Eng A 2011;529:62–73. [42] Dancette S. Comportement mécanique des soudures par points: mécanismes et stratégies de prédiction dans le cas des tôles en acier pour automobile. Ph.D. thesis. Lyon, France: Université de Lyon; 2009. May. [43] Pouranvari M, Marashi SPH. On the failure4 of low carbon steel resistance spot welds in quasi-static tensile–shear loading. Mater Des 2010;31(8):3647–52. [44] Hochhauser F, Ernst W, Rauch R, Vallant R, Enzinger N. Influence of the soft zone on the strength of welded modern HSLA steel. Weld World 2012;56:77–85. [45] Rodrigues DM, Menezes LF, Loureiro A, Fernandes JV. Numerical study of the plastic behaviour in tension of welds in high strength steels. Int J Plast 2004;20:1–18.