Comparion of RSW technologies on DP steels with modified instrumented Charpy impact test

In recent decades the toughness of welded structures becomes more and more important. This trend is particularly valid for the advanced high-strength steels (AHSS), which have reduced toughness comparing to the low-strength structural steels. Dynamic characteristics of the welded joints of the newly developed high-strength steel sheets required by the automotive industry have been neglected, with welding procedures being optimized on the static joint properties, mainly on the tensile-shear force. Up to the present time, testing of dynamic properties of the spot-welded joints has been performed by increased testing speed during tensile testing. The authors have developed a new dynamic testing method and designed new testing equipment for impact bending, which can give a numeric result to characterize the resistance of spot-welded joints against dynamic load. In this paper, this method will be used to evaluate resistance spot-welded joints made on DP600 and DP800 steels with three different technological parameters, comparing long-time welding, short-time welding, and two-pulse welding. The different parameters cause different weld nugget sizes, failure modes, impact energies, and force–time diagrams. All test results show that bigger weld nugget diameters cause better impact energies and impact forces, but the differences are not perfectly correlated.


Introduction
Over the years several impact testing methods have been developed for spot-welded joints, for example, the dropweight system, pendulum impactor, and high-speed servohydraulic equipment [1][2][3][4][5][6][7][8][9]. There are two basic types of dynamic tests suitable for testing of welded joints: the drop weight and the pendulum impactor version. Drop-weight equipment is suitable for the testing of medium and large section sizes due to its lower kinetic energy (100-1000 J), while hammer-type impact testers (like Charpy) can be used for smaller cross-sections, such as resistance spot-welded (RSW) joints. The Charpy impact test was invented in 1900 by Georges Augustin Albert Charpy , and it is regarded as one of the most frequently used tests to evaluate the relative toughness of a material in a fast and economic way [10]. The Charpy impact test is a standard test to measure the impact energy (also referred to as notch toughness) absorbed by a material during fracture. The notch provides a point of stress concentration within the specimen and improves the reproducibility of the results [11]. The absorbed energy is computed by working out the potential energy lost by a pendulum through breaking a specimen [11]. In the early 1970s, French researchers developed test methods for the dynamic testing of thin plates [12]. Grumbach and Sanz at the IRSID Institute in Paris (Materials Research Center and Development) developed a new technique for studying the dynamic characteristics of 5-mm-thick plates and ⌀5 mm cylindrical specimens by redesigning the previously used traditional Charpy seat and hammer (see Fig. 1) and changing the geometry of the specimen [10]. The test is essentially a dynamic impact test. The maximum kinetic energy of the impactor is 300 J. Figure 1 shows the test layout, the specimen fixed in the threaded seat of the "double-headed" hammer, and the impactor is clearly visible. The hammer performs the dynamic rupture in the same direction as the longitudinal axis of the specimen [10].
The emphasis on safety aspects became increasingly important (such as compliance with increasingly stringent fracture tests), and the automotive industry needed to dynamically test welded joints in thin sheets in addition to testing sheet materials, so research has continued in this direction. In the mid-1990s, Kaplan's team in Paris further developed the Grumbach test design. That new method was already suitable for the dynamic testing of spot-welded joints from 0.5-to 5-mm-thin plates, and the maximum kinetic energy was increased to 450 J [13]. The new method was called the impact tensile crash test. The aim of the developers was to perform the test under industrial conditions and to get as close as possible to the processes that take place during real crash tests [10]. There is no notch in the thin-plate specimen containing the spot-welded joint. This is because, during the design of the test specimen, the developers assumed that the spot-welded joint would serve as a stress concentration point, so no separate notch was required. However, with this specimen design, the dynamic behavior of the larger material around the joint is also investigated. This means the dynamic characteristics of the whole structural element containing the welded joint can be determined, thus simulating the processes that take place during an actual collision [10] [14]. The developed specimen design is shown in Fig. 2.
The principle of the test has not changed, and the specimen is still subjected to dynamic impact tensile loading. Figure 3 illustrates the fixation of the specimen, the direction of loading, and the test setup.
The 1995 version was upgraded in 2003; the specimen size was increased, the hammer was redesigned, and the impact energy was increased to 750-800 J [13]. The new developments are shown in Figs. 4 and 5.
The results of the performed dynamic tests and fracture tests showed that the failure mode of RSW joints of advanced high-strength steels is very often interfacial or is only partial weld button. This is presumably due to the fact that the welding parameters to be used during production are chosen on the basis of the results of conventional, quasi-static tests (shear-tensile, cross-tensile, peel, etc.) [10]. During the quasi-static test, the joints are still buttoned, but they already show a less favorable failure mode due to dynamic loads. Beside these, the developed impact test methods for RSW joints are investigating the impact-shear or impact-tensile properties, which not perfectly show the behavior of spot welds in case of real-impact effect. Typically, the side parts of a car chassis during a car crash suffer impact bending. The impact velocity range can be very variable; in case of lowspeed car crash simulations, it is less than 10 m/s [16]. The stress condition near the weld nugget is three dimensional and very complex in a spot-welded specimen. Furthermore,  Developed impact-tensile specimen for resistance spot welded joints [15] the strain rate near the spot weld could be very high under ordinary vehicle collision speed, for example, the local strain rate can reach 3500 1/s at 6 m/s impact velocity [17]. For higher impact velocity, more and more researchers are using a ballistic test of spot welds, where the impact velocity is more than 100 m/s [18]. Nowadays, several data sets have been published about the strain rates and their complexity [19,20].
According to the literature overview, there are several trials to find the good method to determine the behavior of resistance spot-welded joint during dynamic loading. In this paper a developed instrumented impact-bending test was used to investigate these properties, where the loading direction is similar like in case of a car crash. RSW joints are subjected to dynamic loads during their lifetime, either randomly or as a result of an extraordinary event (e.g., crash), and thus should be welded with welding parameters which are also optimized for dynamic loads, so in this investigation, different RSW joints were compared on DP600 and DP800 steels.  The redesigned hammer and seat [10] 1 3

Testing method
The typical loading of spot-welded joints in the automotive industry is not static, cyclic and dynamic loadings are especially likely to occur. The design of the joints requires knowledge of the loading limits of each spot-welded joint. Some vehicle parts are subjected to a wide range of dynamic loadings in varying sizes, frequencies, and types during the operation of cars. When installing these components, it would be risky to rely only on the generally good quality of spot-welded joints and previous favorable experience, as well as the results of conventional quasi-static tests. It is important to know the absorbed impact energy, and also need to know the area under the curve of the force-time diagrams, as well as the shape of the curve, in order to determine the deformation behavior of spot-welded joints made of high-strength steels. However, this requires an instrumental impact test of the spot-welded joints on the overlapped thin sheets. In order to get a more accurate and thorough picture of the dynamic behavior of spot-welded joints in automotive steel sheets, a special instrumented testing method was developed for spot-welded joints.
The usual impact velocity of the Charpy test is 3-6 m/s [21][22][23]. The kinetic energy is used for the forming of the specimens and the breaking of the joints during the dynamic (three-point) bending that causes the specimens to fail. The pendulum hammer with m = 19.794 kg falling weight and a height of fall h = 1.545 m used for conventional standard tests has a maximum impact energy of 300 J with an impact velocity of 5.5 m/s. However, the experience of the preliminary experiments showed that even for 1 mm + 1 mm overlapped joints of high strength DP steels, this falling weight and the maximum impact energy of 300 J are too high, so the Charpy impact test machine was modified: in order to register the force, a force cell was placed in the impact edge of the hammer and the weight of the hammer was reduced; thus, the maximum impact energy was changed to 118.25 J. In Fig. 6 The testing equipment and the modified hammer Fig. 7 The prepared specimen and its geometry order to examine the spot-welded joints as accurately as possible, the design is also modified as shown in Fig. 6.
During the development of the geometry of the specimen, an optimal bridge width was determined, which can provide an overall view of the dynamic behavior of the whole welded joint (nugget + heat-affected zone + base material), considering the avoidance of base material failure around the fixing points. The results of the initial experiments showed that the effect of the base material on the test result is still quite negligible at a bridge width of 20 mm. The specimen geometry and a prepared specimen are shown in Fig. 7.
Due to the specimen geometry, it was also necessary to design and manufacture a new seat that would be suitable for holding the spot-welded specimen. The new seat developed for the test is shown in Fig. 8. Figure 9 shows a schematic drawing of how this method works and how the failure happens.

Base materials
In our study, sheets of 1 mm sheets thickness of two different grades of SSAB Docol dual-phase (DP) steels (DP600 and DP800) were used as a base material in RSW joints. The chemical and mechanical properties provided by the manufacturer in material certificates are shown in Tables 1 and 2.

Welding circumstances and technologies
The resistance spot welding was carried out using a TECNA 8007 resistance spot welder using AC (50 Hz) and controlled by a TE550 microprocessor-based welding control unit. The experiments were performed in constant current mode.
Copper chromium zirconium welding electrodes were used with a 5 mm spherical head diameter. The radius of the electrodes with spherical head must be chosen depending on the sheet thickness to be welded. For a sheet thickness of 1 mm, R = 50 mm radius was used, which according to the literature [24,25] and previous experiments can contribute to the stability of the process. The design of the electrodes is presented in Fig. 10.
One of the electrodes is fixed to a piston connected to a pneumatic system to move the electrode vertically toward the other electrode as a result of pneumatic pressure. The pressure generated by the pneumatic system becomes a pressure force applied on the samples to be welded. A specified electric current passes through the sheets pressed by the electrodes to generate Joule heat at the faying surfaces as a result of electrical resistance.  In this experiments, three different welding technologies were applied on the two steel grades. After preliminary experiments, long-time welding with high welding current and welding force (no. 1), two-pulse welding with lower welding force (no. 2), and short-time welding with lower welding force (no. 3) were used. The cooling time means the time between pulses in the case of two-pulse welding. Table 3 shows the exact parameters of these technologies and the weld nugget diameter from one sample each.

Results and discussions
For the impact-bending test, the previously described method was used. For the impact testing, eleven specimens were made for the investigation of the same welding technology in one base material, which means a total of 66 tested specimens at all. The tests were performed at 25 °C room temperature. This test is very fast, so a high sample rate was necessary for the determination of force-time diagrams during the fracture. The sample time was 5 ms and 2500 samples were made within this time. This high sampling rate gave exact results. Figure 11 shows the three typical failures of the spot-welded joints. In the case of long-time welding, base material failure occurs instead of button pull failure, while in the other cases, the spots buttoned with different sizes. The partial weld button and button pull failure modes are typical failures on resistance spot welded joints and have been reported in several publications [26][27][28][29].
In Table 4 the impact-bending test results are summarized and statistically evaluated. In every case the impact energy and the maximum impact force were measured. Standard deviations (SD) and the coefficient of standard deviations (SD c ) were calculated. The plug diameters (after the test) were measured on the fractured specimens, perpendicularly to the hammer edge.
The results of resistance spot-welded DP600 show that there are big differences between the long-time welding welding and the other two technologies for both impact energy and maximum impact forces. The short-time welding shows the worst results but the deviation is the lowest. There are slight differences in impact energy and maximum impact force between short-time welding and two-pulse welding; however, the plug diameter differs significantly. Generally, the standard deviations of all results are acceptable. Figure 12 shows the impact force-time curves of these tests.
The explanation of the curve is as follows: -The first phase from 0 to 0.7 ms shows nothing important because the hammer has not yet reached the specimen yet. -The second phase from 0.7 to 1.5 ms: the hammer reaches the specimen, the base material forming, but no fracture. -The third phase from 1.5 to the end of wave: weld nugget fracture (in the case of long-time welding half weld nugget, half base material). -The last phase from end of wave to 5 ms: base material forming after weld nugget fracture.
The curves clearly show the differences in impact energies especially in the case of long-time welding. The curve of long-time welding (gray) shows two waves of maximum forces: the first wave shows the half-broken spot, the bigger second wave is the base material failure. The difference between short-time welding (blue) and two-pulse welding (orange) are clearly seen; the two-pulse welding shows better maximum impact force and better elongation, too. Table 5 shows the results for the DP800 base material. Both impact energy and maximum impact force show clear differences between the welding technologies. Short-time welding is the worst, with low values for both impact energy and impact force. Two-pulse welding is much better, and the long-time welding is again the best. The standard deviations are good in all cases for both impact energy and impact force too. Figure 13 shows the impact force-time curves of the three technologies for the DP800 base material.
The curves clearly show the differences in maximum impact forces as well as time differences. The shape of long-time welding curve is somewhat different than that of DP600. The long-time welding curve shows two well separated waves, and it can be assumed that the first wave is caused by the failure of the half joint, while the second wave occurs because the base material has broken. As expected, the short-time welding shows the worst results. From the impact-bending tests, three absolutely different failure modes can be distinguished, which clearly depend on the applied welding technology. To check where the failure occurs in the joints, crosssectional macroscopic tests were made. Figures 14, 15,   and 16 show the cross-sections and hardness curves of RSW joints made by, respectively, long-time, twopulse and short-time welding, respectively. Below the cross-sections and hardness curves, the cross-sections of the impact-bending tested specimens can be seen. All cross-sections and hardness measurements were made on DP800 base material. The Vickers hardness   measurements were made by a Mitutoyo microhardness tester, the loading was 200 g, and the distance between measurements was 0.2 mm.
In the case of long-time welding, the fracture occurred in the outer part of the heat-affected zone (which is called the sub-critical heat-affected zone), and it continued on the base material. Because of this, only one side fracture can be seen in Fig. 14. Based on hardness distribution values, this is the most softened area, so the resistance against dynamic loading is the smallest here (252 HV). Basically, the value of softening and the plug diameter determine the maximum impact force in this case. Figure 15 shows the hardness distribution and the failure of two-pulse welding specimens. The hardness distribution is similar to that of the long-time welding technology, but in the sub-critical heat-affected zone, the hardness is slightly higher. The fracture started in same place on the right side of the photo; the hardness is a little bit higher here than for long-time welding (276 HV). The left side shows the end of the fracture. There is a clear difference between the two sides; on the left side, the crack is not perpendicular to the sheet surface. This is because the direction of loading changed during the test and shearing happened on the left side. Here the crack goes from the heat-affected  zone to the weld nugget, so the crack direction is determined by the loading direction, and the hardness has no effect on this.
In the case of short-time welding technology, the hardness distribution is similar to that two-pulse welding, but the failure started from the edge of weld nugget and finished in the weld nugget, too, so it is a partial weld button. This can be caused by imperfect fusion on the edge of the weld nugget. The shape of the left side is similar to that of two-pulse welding, but the plugged diameter is the smallest, so it causes the worst result of the impact-bending test. 14 Cross-section of long-time welding technology on DP800 before (above, with hardness curve) and after (below) impact-bending test 1 3

Conclusion
Based on the experimental results, it can be stated that the developed impact-bending test method can be applied to compare different resistance spot welding technologies and/or different base materials.
The impact test specimen geometry is acceptable for the resistance spot weld impact-bending test if the plug diameter is smaller than 5.1 mm. In the case of 7.2 plug diameter, the base material broke and the weld nugget just partially broke. The conclusion is that the bigger diameter nugget requires bigger width of the base material in the impact zone. Therefore, the Fig. 15 Cross-section of two-pulse welding technology on DP800 before (above, with hardness curve) and after (below) impact-bending test Fig. 16 Cross-sections of short-time welding technology on DP800 before (above, with hardness curve) and after (below) impact-bending test data for 7.2 mm nugget diameter do not show the exact results; however, the best results come from this type of weld. Probably the impact-bend loading resistance of these welds is better.
In terms of impact energy, there are significant differences between the three technological combinations on DP800, but DP600 does not show any major differences between short-time welding and two-pulse welding.
In the case of DP600, the worst results come from the shorttime welding, next is the two-pulse welding, and the long-time welding gave the best results. The plugged weld diameter is not correlated with the impact energy and impact force for two-pulse welding. The plugged diameter is significantly smaller in shorttime welding, but this does not cause so much impact force and impact energy loss in case of DP600.
The resistance spot-welded joints of DP800 show bigger differences between technological parameter combinations. It clearly identified the short-time welding has the worst resistance against impact loading, two-pulse welding is next, and the long-time welding has the best results. The impact energies and impact forces are almost perfectly correlated with the weld nugget diameters.
Three typical failure modes occur during the tests. In case of two-pulse welding and long-time welding of DP800, the failures start from the heat-affected zones in the most softened part. In the case of short-time welding, the failure starts from the weld nugget.
Funding Open access funding provided by University of Miskolc.

Conflict of Interest
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.