3.1. Visual Inspection and Magnetic Particle Testing
Visual inspection tests were performed on the joined samples of RSW to investigate whether there were any surface defects. Possible surface deformations, cracks, etc. around the welding area were examined, and the welding core diameter, height, electrode penetration depth, and spot diameter were measured in mm using calipers and micrometers. The images of the measurements are presented in Fig. 5, and the data in tabular form are given in Table 4. Figure 5 represents a-normal and b-RPH processes.
Table 4
Dimensions of the welding nugget and spot.
| Nugget diameter (mm) | Nugget height (mm) | Electrode immersion depth (mm) | Spot diameter (mm) |
Normal | 5,04 | 1,18 | 0,52 | 6,13 |
RPH | 5,32 | 1,27 | 0,58 | 6,18 |
There are different standards for determining the minimum size of a weld in the automotive sector [30, 31]. In order for the welded joint to be able to withstand the desired load, it is required that the size of the weld nugget be as large as possible. When examining Table 4, it can be seen that while the electrode penetration depth is 0.52 mm, the nugget diameter is 5.04 mm, and similarly, when the electrode penetration depth is 0.58 mm, the nugget diameter is 5.32 mm. It can be observed that as the electrode penetration depth increases, the weld nugget diameters also increase.
When the welded joints were visually inspected, it was determined that there were no defects such as cracks, superficial deformations, or interface protrusions in the welding zone. In this context, it can be inferred that the parameter selections were appropriate in the DNA method that was performed. Of course, the fact that no welding errors were encountered is due to the adaptation of the welding machine to today's technology, allowing for software applications to minimize errors that could occur due to human factors.
After visual inspection, magnetic particle (MP) testing was performed to detect discontinuities that cannot be detected by the naked eye on the surface or near-surface internal structures of the test pieces joined by RSW. MP testing was performed on both sides of the STRENX and DP welded connections, and the obtained images are given in Fig. 6. Figures <link rid="fig6">6</link>-a and 6-b represent the normal welded connections, while Figs. 6c and 6d represent the RPH-treated ones.
Upon examining Fig. 6, it was determined that liquid metal brittleness, which can occur due to the formation of intermetallics between the zinc coating and the steel, did not occur due to the residual austenite. As a result, it is believed that the absence of the austenite phase in the joined steels or the inability of residual austenite, which may exist in very small proportions, to diffuse with the melted liquid zinc, and the suHAZility of the selected parameters for the chemical and mechanical properties of the steels and welding processes support a flawless weld.
Başer [32], joined galvanized beynitic ferrite-supported TRIP (TBF) steel using direct current medium frequency technology (MFDC) with the RSW method and applied magnetic particle (MP) testing to detect liquid metal embrittlement in the joints. As a result of the magnetic particle testing, it was reported that no liquid metal embrittlement was detected in the weld core and HAZ.
3.2. Microstructure Investigations
In all welding methods, microstructure studies performed in the welding zone are crucial in order to understand the changes that occur in the microstructure or the joining effects after the welding process. Therefore, macro and microstructure studies were performed on the welded specimens to determine the effects of the RPH process on the welding core and especially on the surrounding HAZ. Macro-microstructure images for normal welded connections are shown in Fig. 7, while welded connections obtained using the RPH process are shown in Fig. 8.
In the microstructure analyses of the welded samples, which were obtained using normal welding and RPH, in the cross-sections of the samples, as shown in Fig. 7, 1 represents the weld metal, 2 and 3 represent the transition zone between the weld core and the HAZ, and finally, 4 and 5 represent the transition zone from the HAZ to the base material. When the macro and microstructures of the welded samples subjected to RPH, as shown in Fig. 8, were examined, it was observed that similar transition zones existed as in Fig. 7. In addition, as shown in Fig. 8 (visual 3), the weld core-HAZ was divided into 4 different regions. On Fig. 8, a represents fine-grained HAZ, b represents coarse-grained HAZ, c represents the fusion boundary zone, and d represents the fusion zone.
When examining the microstructure images of the welded joint with RPH applied in Fig. 8, which belongs to the transition area from HAZ to the base metal, relatively lighter-colored perlitic-ferritic structures are observed in these areas, while the formation of tempered martensitic structures around this region is noteworthy. Nikosohbat et al. [33], welded DP980 steel using the RSW method and examined the macro and microstructures of the welded specimens by taking sections. As a result of the examinations, they stated that the HAZ was composed of ferrite and martensite phases, there was softening in the HAZ as it went from the base metal to the welding zone, and this situation was tempered with the martensite phase due to the heat effect.
Although the chemical compositions of STRENX 700 CR and DP 800 high-strength steels are different, the cross-section images show similarities due to their similar resistances. It has been determined that the effect of STRENX 700 CR material on HAZ and the welding core is greater than that of DP800.
When the microstructure images of the welds produced in Fig. 8 are examined, it is thought that the coarse and fine-grained structures observed in HAZ were transformed into austenite phase due to the peak temperature reached during the DNA process being above the Ac3 temperature, and then coarse martensitic phases were formed due to cooling. It is also believed that this contributed to the formation of coarse grains in the welded joint due to the delayed cooling of the joint caused by the RPH process. The region where the highest temperature is reached in coarse-grained HAZ is above the Ac3 value and is the area where there is enough time for the growth of austenite grains [34, 35].
3.3. Hardness Test
Hardness measurements were taken along a line on the test pieces joined by the DNA method to investigate the effect of hardness changes in the welding area on mechanical properties. In this context, measurements were made from DP 800 and STRENX 700 CR materials and welding core regions to determine the hardness profiles of welded joints, and graphs were created. Graphs showing the changes in hardness strengths of RPH and normally joined DNA welds are shown in Figs. 9 and 10, respectively, to examine the changes in hardness strengths of welded joints. These graphs, with and without RPH, were obtained at 120 µm intervals along the cross-sectional directions of the welded joints and from an average of 145 hardness measurement points.
The two hardness graphs given below are very similar to each other regardless of the values. When the graphs are evaluated in general, it is seen that the highest hardness is in the normal welded joints of DP800 HAZ, followed by the welding metal. The lowest hardness values are found in the welded joints subjected to the RPH process, and it is determined that the lowest hardness is observed in STRENX 700 CR HAZ, and then hardness values are formed in the welding metal. Sanchez et al. [36], stated in their study that the hardness values of high-strength steels after DNA are caused by changes in ferrite and martensite phases in the structure, especially in HAZ. They also stated that the decrease in hardness in HAZ occurred as a result of the tempering of the martensite phase.
When Figs. 9 and 10 were examined, it was found that the highest hardness value in the weld metal in normally welded joints was 489 HV, while the highest hardness value in the weld metal after RPH application was found to be 465 HV. This result showed that RPH caused approximately a 5% reduction in hardness in the weld metal. Thus, it was seen that the reduction in hardness values in this region with the RPH process allowed for a more homogeneous structure by bringing it closer to the hardness of the base metal.
It was observed that the lowest hardness of RSW-joined normal sources in STRENX 700 CR HAZ was 291 HV, while the highest hardness was 487 HV. Welded joints made with RPH showed that the lowest hardness in STRENX 700 CR HAZ was 287 HV, while the highest hardness was 454 HV, as seen from the hardness measurements taken and the graphs created. According to these results, there was almost no change in hardness values measured in the martensitic softening zone of HAZ in STRENX 700 CR (Figs. 9 and 10), while there was a 33 HV decrease in intercritical HAZ. As a result, a hardness decrease of approximately 6.8% occurred in STRENX 700 CR HAZ due to RPH prior to welding. Thus, it can be seen that the RPH process is effective in MFDC technology and reduces the upper critical hardness of HAZ, enabling more ductile ruptures to occur. At the same time, the changes in hardness values in test samples subjected to both normal and RPH treatments in critical sub-HAZ were negligible, and it is thought that the RPH treatment causes the forces applied to welded joints to be distributed over a wider area with the expansion of HAZ. It is believed that this prevents the formation of a sharper zone for the starting point of deformation when a mechanical force is applied to RSW joint, and thus makes welded joints more resistant under mechanical stress.
When the DP 800 HAZ was examined through the graphs, it was seen that the martensite softening observed in the STRENX 700 CR HAZ did not occur here. The highest hardness values observed in the DP 800 HAZ were 493 HV in normal welded joints and 458 HV in welded joints with RPH. The measurement results showed that a 7.8% decrease in hardness occurred in the DP HAZ with the application of the RPH process.
When the lowest hardness values observed in the welding metal were compared with each other through the graphs (Figs. 9 and 10), a 35 HV decrease in hardness values was observed. Similarly, in the measurements, it was observed that the difference between the highest hardness values observed in the welding metal was 24 HV. As a result, a decrease of approximately 8% in the lowest hardness and 5% in the highest hardness values of the welding metal was achieved as a result of the heat treatment applied in the joints.
Chabok vd. [37], reported in their study that intercritical HAZ was present in the Ac1 and Ac3 range, which they considered as peak temperatures, and that the volume fraction of martensite in the weld zone was higher than in the base metal. In a similar study, Jonardhan et al. [38], stated that hardness values increased as they moved from the base metal towards the fusion zone, and that this was due to the presence of martensite and different zones formed in HAZ. Khan et al.[39] also found that hardness increased from the base metal towards the fusion zone, and attributed this to the chemical composition of the steels and the rapid cooling process.
3.4. Tensile-Shear Test
Three welded samples were made for each parameter by joining the welded samples, and the graphs of all samples were created separately for each. As an example, the pulling-shearing graph obtained from the samples processed with PWHT is given in Fig. 11-a, and the pulling-shearing graph created to better understand the effects of welding parameters used during welding is given in Fig. 11-b.
It can be seen that the pulling-shearing measurements obtained from all welded connections with both normal and PWHT treatments gave similar results in both force (kN) and extension (mm). When Fig. 11-a is examined, the highest pulling-shearing strength in the welded joints performed with PWHT was measured as 15.58 kN. It is thought that the increase in strength was due to the expansion of HAZ as a result of the slow cooling of the welded connections after the PWHT process, and as a result of the wider area of reaction of the applied force against the pulling-shearing strength. As a result, it was understood that the RPH process contributed significantly to the pulling-shearing strength. Wang et al. [40], stated in their study that the pre-heating process increased the pulling-shearing strength, while Manladan et al. [41] stated that the pre-heating process helped to clean the oxide layer and improved the contact resistance and welding quality. Lia et al. [42] reported that pre-heating in spot welding promoted the formation of a fusion region and also increased the welding strength.
When Fig. 11-b is examined, it is clearly seen that the application of RPH to the joined welded specimens led to an increase in their strengths compared to the welded specimens without RPH. Looking at the tensile-shear strengths in the graph, it can be observed that the strength increase in the RPH-applied welded specimens (15.45 kN) compared to the non-RPH applied joined specimen (14.51 kN) was approximately 6.7%.
When evaluated as a whole, the tensile-shear results showed parallelism with hardness studies and it was seen that the changes in hardness strengths had an effect on the tensile-shear strength capacities [43]. At the same time, in order to determine the tensile-shear strengths, it is necessary to have a good knowledge of the morphology of the welded connections. High-strength dual-phase steels generally experience tensile-shear damage not from the base metal but from the HAZ, unlike other steels (low-carbon steels, high-strength low-alloy steels, etc.), and it is believed that this situation occurs due to HAZ softening, which is different from other steels [44].
In conclusion, when looking at the tensile-shear strength results, it is observed that RPH, by reducing the hardness in the upper critical HAZ and allowing it to expand, causes the mechanically applied tensile-shear force to spread over a wider area, thereby increasing the tensile-shear strength, in other words, the load-carrying capacity, with the application of RPH compared to normal welded connections.
3.4.1. Tensile-Shear Failure Modes
The rupture images, or in other words, the failure modes obtained after the tensile-shear tests on welded connections are shown in Fig. 12. In Fig. 12, a and b represent the RPH process, while c represents the normal welded connections. As seen in Fig. 12, the post-rupture failure mode is buttoning, and the ruptures occur between the HAZ and the main material.
When the morphological structure of the damage modes was examined, it was concluded that the martensite phase, especially those present within the structure, was significant in thermal cycling, and that the softening of the tempered martensite that occured below the critical HAZ in microhardness measurements played a role in the buttoning-type ruptures. Additionally, the weakest area, the HAZ, was where the buttoning-type damage modes seen in Figs. 12-a, b, and c occured, and it was understood that the broader HAZ resulting from the RPH process not only led to a more homogeneous structure due to the reduction in upper critical HAZ hardness between Ac3 and Ac1 seen in hardness test results but also enabled the load stresses applied during the tensile-shear strength testing to be borne in a broader area. Hernandez [45] and Shojaee [46] reported that softening zones occured due to martensite tempering in the non-critical heat-affected zone (HAZ) close to the base metal and that the presence of these softening zones could affect the type of ruptures that occur during tensile-shear strength testing.
SEM images of the rupture damages obtained from tensile-shear tests conducted on STRENX 700 CR and DP 800 steels after different process treatments are given in Figs. 13 and 14, respectively. Figure 13 includes SEM images of the rupture damages of normal specimens, while Fig. 14 shows the SEM images of the rupture damages of specimens treated with RPH.
Upon examination of all the SEM images provided in Figs. 13 and 14, it was seen that the ductile-brittle and brittle rupture modes occured due to the formation of shear rupture. When the structure morphology was examined, it was evaluated that the results obtained were one of the consequences of martensitic softening, which could vary depending on the martensite volume fraction, tempering degree, and severity in critical sub-HAZ in dual-phase steels. The region where martensitic softening occured was surrounded by the fusion boundary and the martensite phase that occured due to rapid cooling, which led to shear rupture in the form of failure after the tensile-shear strength test.
The SEM images of resistance spot welded joints were further examined, and it was observed that cracks started from the lower critical HAZ of STRENX 700 CR, continued to the upper critical HAZ where high microhardness values were obtained, and then led to rupture by surrounding the fusion zone. It was also considered that the presence of heterogeneous structures in HAZ triggered internal stresses and led to crack formation. Similar studies support this finding [47].
3.5. Cross-Tension Test
Tensile-shear testing is one of the commonly used methods for determining the mechanical behavior of resistance spot welded joints [48, 49]. Therefore, tensile-shear tests were conducted on all normal and PEO-treated welded joints, and the results were used to create a force (kN)- displacement (mm) graph (Fig. 15). For the tensile-shear testing, three welded joints were made for each parameter, and the normal graph was created as an example, which is presented in Fig. 15-a.
When the averages of the cross-tensile strengths were examined, the lowest strength value was measured as 4.73 kN in the normal samples, while the highest strength value was obtained as 4.97 kN in the samples where the PWHT process was applied. When a general comparison was made between the welding operations, it was determined that an increase of approximately 5.3% was achieved in the cross-tensile strengths of the samples where the PWHT process was applied. A graph, composed of the triple averages of the welded samples (a total of 6) in both different parameters, is given in Fig. 15-b to be able to see the cross-tensile results together and for easy understanding. When the results are examined, it is clearly seen that the PWHT process contributes significantly to the increase in cross-tensile strength.
The graph obtained in this way was actually an expected outcome. It was believed that the increase in cross-tensile strength was due to the changes that occured as a result of the effect of martensitic softening that occured after welding, critical sub-HAZ formation, upper critical HAZ, and also due to the lower heat input to the welding zone with the use of medium-frequency direct current technology. At the same time, the high cross-tensile strength values of the welded connections where the PWHT process was applied, as seen in Fig. 15-b, were attributed to the fact that the force applied to the test samples was distributed over a wider area due to the expansion of the critical sub-HAZ zone (soft zone) observed in HAZ with the PWHT process. As a result, it was predicted that the tears spread over a wider area compared to normal welded connections, and thus, the cross-tensile load-bearing capacity increased. In addition, it was found that the cross-tensile strength values obtained from the welded samples were approximately three times lower than the tensile-shear strength values. It is known that the way the force was applied during testing was the reason for this situation in the joints made using the RSW method. Hernandez [50], joined DP steels using the RSW method and stated that the cross-tensile values were lower than the tensile-shear test values. Chao [51] applied tensile-shear and cross-tensile tests to welded connections by joining high-strength steels with the RSW method in his study. According to the results obtained, he stated that the cross-tensile strength was lower than the tensile-shear strength.
3.5.1. Cross-Tension Failure Modes
In the automotive sector, the maximum load carrying capacity during collisions is determined using the cross-tension test, and the reliability of vehicles produced in this way is determined, and this method is frequently used. Therefore, the rupture surfaces of DNA samples obtained after the cross-tension test were examined macroscopically in detail, and images taken from the rupture regions of the samples after the test are given in Fig. 16, where a represents the RPH and b, c represent normal operations.
Cross-tension tests were applied to all welded samples combined in two different parameters, and the rupture images of all cross-tension test samples (normal) made in a series are given in Fig. 16-a for an example. The rupture images obtained from the samples subjected to cross-tension tests under normal conditions are shown in Fig. 16-b, and the rupture images obtained from the samples subjected to RPH are shown in Fig. 16-c. Rupture damage modes are crucial in determining the toughness and load-carrying capacities of DNA connections. Looking at Fig. 16, it can be seen that all damage modes occured in the form of buttoning.
The buttoning damage mode seen as a result of the cross-tension test indicates that it occurred around the weld or HAZ. It is believed that the formation of buttoning damage mode begins with cracking around the weld core due to the applied force, and it progresses as the force continues to load. Damage modes that occur in this way indicate that the welded joint will show the necessary strength when the structure is forced to separate under the static force application.
The ruptures that occured as a result of the cross-tension test are seen to occur in a brittle and semi-brittle manner on the STRENX 700 CR side. When the structure morphology was examined, it was thought that factors such as martensite softening and upper critical HAZ triggered these types of ruptures. In this context, it was confirmed that the cross-tension results obtained were parallel to the tensile-shear results, and therefore, the optimum welding parameters determined by the tensile-shear test were also confirmed by the cross-tension test results. Tamizi et al. [52], stated that during the formation of cross-tension damage modes, the crack starts at the upper critical HAZ and then continues towards the lower critical HAZ.
3.6. Fatigue Test Results
The effects of welding parameters on the fatigue life of DP and STRENX steels joined by RSW were investigated by subjecting them to fatigue tests. Fatigue tests were conducted at different load levels (0.2, 0.3, 0.5, 0.75 kN) and a constant frequency of 5 Hz to predict possible discontinuities (such as crack formation) that may occur as a result of continuous cyclic loading under future usage conditions. Force-life curves were created to understand the fatigue behavior of the materials using the data obtained from the fatigue tests.
The result graph produced based on the numerical data obtained from the fatigue tests of welded connections (normal and RPH) is shown in Fig. 17. When the force-cycle relationship of the RPH process was examined, it was determined that there were significant increases in fatigue cycle numbers due to load reductions.
When the welded joints were examined among themselves, it was observed that the average cycle numbers of the normal processed samples under 0.75 kN load were 4,850, while this rate increased by 109% to 10,179 cycle numbers in the fatigue strength of the welded joints with HSLA steel and DP steel joined by the RPH process. Similarly, while the average fatigue cycle numbers of normal processed joints under 0.2 kN load were 327,116, this rate increased by 189% to 947,632 cycle numbers in the welded joints joined by the RPH process.
As a result, it was determined that there was a significant increase in the number of cycles due to the decrease in the applied force during the fatigue strength testing. The RPH process before welding was clearly effective in the formation of increased cycle numbers. Considering all these results, it is thought that the high fatigue strengths of welded connections with the RPH treatment compared to normal welded connections, which was previously observed during hardness measurements and explained in detail as the widening of the soft zone in HAZ, can be attributed to this soft zone. Ordonez et al. [53], joined DP980 dual-phase steel using RSW and applied fatigue testing to the resulting welded joints. They stated that a soft zone was formed due to the tempering of martensite as it could not reach the over-tempering temperature within the structure after welding, and that the soft zone and the decrease in hardness values due to softening could improve fatigue damage that could occur on HAZ. Banarjee et al. [54], examined the effects of weld size and geometry on the fatigue performance of DP590 steel welded joints using RSW. According to the results obtained from the fatigue tests, it was observed that the applied force and weld size had an effect on the fatigue strength, and insufficient fatigue strengths were obtained in cases of small weld sizes and high stress loads.
3.6.1. Examination of Ruptured Surfaces after Fatigue Testing
The examination of rupture surfaces obtained after the fatigue strength tests revealed that ruptures occurred at the weld core and HAZ interface in all welded joints. It is known that in RSW studies in the automotive industry, fatigue strengths are important along with material performance, microstructure, and geometric properties [54, 55]. To analyze the detected ruptures after fatigue, one sample from each series was examined. It can be said that the crack initiation occurred around the weld core during fatigue testing, and due to the increase in dislocation density with the heat effect, the crack continued from the HAZ and then progressed along the material. Xu et al. [56], investigated the fatigue performance of dual-phase steels using RSW and found that the fatigue crack initially started around the core.
The rupture patterns observed in welded joints after fatigue testing appeared to be very similar and consistent. It is believed that all ruptures occured due to the softening of the weld nugget and the HAZ interface that occured after welding, followed by crack propagation. The softening caused by thermal transformations in the STRENX 700 CR region could be considered as a significant factor in crack initiation and rupture formation, as well as HAZ expansion and martensite softening.
Figures 18 and 19 show SEM images of ruptured surfaces of samples subjected to normal and RPH treatment, respectively.
Upon examination of the SEM images (Figs. 18–19), it is believed that cracks formed during the fatigue test started from the tempered zone, i.e., STRENX 700 CR HAZ, and then continued along the cross-section, reaching the coarse-grained HAZ and surface coating, causing brittle ruptures. It is also highly likely that there are areas of ductile-brittle rupture formation. When a literature search was conducted, it was suggested that the decrease in fatigue strength after the RSW process could be due to the notch effect that occurred in the RSW and the microstructure that was created in the base metal, HAZ, and welding metal [55, 57, 58]. Therefore, due to the heterogeneous microstructure present in the structure, the rupture images of the areas taken for SEM analysis may vary.
Holovenko et al. [59], investigated the fatigue life of welded samples in their study and found that welds made by the RSW method showed a noticeable decrease in fatigue life compared to laser welding. The researchers explained that the cause of this decrease was the notch effect caused by the welding geometry, the different mechanical properties exhibited by the microstructures, and the presence of HAZ.
According to the SEM images, fatigue cracks are thought to occur as a result of the breaking of interatomic bonds in certain planes, and these crack progressions are called cleavage mechanisms. These progressions can be seen within grains or at grain boundaries. It is known that the yield strength of the ferrite phase in the structure is lower than that of the martensite phase. Therefore, while the martensite phase remains in the elastic state, the ferrite phase is subjected to plastic deformation. With this effect, the stress in the ferrite increases, and thus, regional deformations can be observed. As a result, ductile or cleavage-type brittle ruptures can be seen depending on different morphologies. In Fig. 18, the width of one measured crack was 380 µm, while the length of another crack was measured as 4,435 µm. In this context, it can be seen from SEM images that cleavage cracks occur in related fatigue ruptures, and it is also possible to observe cracks occurring at grain boundaries, as seen in Fig. 19.