Introduction: Multi-material constructs are considered a key method for achieving the simultaneous goals of mass reduction along with improved safety in NVH. Given the common materials used for body in white construction aluminum\steel structures are receiving considerable interest. Structures containing both aluminum and steel are most commonly joined using mechanical fastening or adhesive bonding methods, however direct welding offers both economic and mechanical performance advantages. Recent work has focused on applying resistance spot welding for this application. Resistance spot welding of aluminum and galvanized steels has been successfully demonstrated and integrated into one commercial vehicle. However, using common resistance spot welding practices, joints are plagued by intermetallic formation and inconsistent mechanical properties. This paper addresses an alternate method of creating such resistance spot welds, that using a capacitor bank as a power source. Capacitor discharge power supplies offer significantly shorter cycle times, and can be advantageous for minimizing the deleterious effects associated with welding. This paper preliminary efforts at creating aluminum to steel joints using this process. This work includes assessments of processing requirements, resulting mechanical properties, and underlying microstructures. Simplified modeling tools are also used to assist in characterizing process\microstructural relationships.
Experimental Procedures: Experimental evaluations were done between a low carbon hot-dip galvanized steel and two grades of aluminum. This included alloys AA5182 and AA6451. All materials were nominally 1-mm thick. All welding was done using a 4-kJ capacitor discharge power supply coupled with a pedestal welding frame. Welding forces and electrode sizes were based on AWS recommended practices for resistance welding. Selected processing conditions were defined based on the thickness of the aluminum sheet. Current waveforms for each trial were measured using a Rogowski coil, active integrator, and digital oscilloscope. Basic weldability assessments were assessed through current range testing. Current range tests show the relationships between both the underlying nugget diameters and the joint strengths at failure to the peak waveform currents. Best practice welds were then subject to metallurgical assessment. Metallography was conducted to both understand developed nugget morphologies as well as to characterize the bond line microstructures. Etching of these samples was done in a two-step process. This included first etching with a picric acid saturated in water solution (to reveal the steel nuggets) then mixed acids (to reveal that in the aluminum sheet). Sample macrostructures as well as details of the aluminum nugget and joint interface were assessed using optical microscopy. Additional mechanical testing was done for samples welded in both the tensile shear or cross tension configurations. Tensile shear and cross tension testing for each attached aluminum alloy was performed with a sample size of 33 replicates. This was done to provide assessments for stability of the measured mechanical performance.
Results and Discussion: Current range testing for both materials showed considerable process robustness. For both materials, acceptable button sizes could be obtained from currents ranging from 30-kA to 60-kA. Joint strengths were much more sensitive to the applied current. Joints made using the AA5182 material achieved strengths of roughly 3.0-kN, while those employing AA6451 were limited to 2.5-kN. These strengths are similar to those obtained when welding the individual aluminum sheet materials to themselves. Metallographic inspection showed that for both aluminum sheet types, the joint morphology consisted of two separate weld nuggets. This included one fully contained within the steel sheet, and a second that appeared to form epitaxially in the aluminum at the faying surface. Further, aluminum weld nuggets appeared largely featureless, and were limited to less than 100-Ám in thickness. In addition, the observed thickness of any intermetallic compounds at the bond line appeared to be limited to less than 1-Ám. Characteristics of the resolidified aluminum were related to the rapid cooling rates inherent in the process. Simple thermal modeling suggested these cooling rates were on the order of 105-║C/s, and sufficient to suppress chemical segregation during solidification. Observed intermetallic thicknesses were consistent with that resulting from the steel hot-dipping process. An analysis of intermetallic growth kinetics suggested that little or no thickening of these phases could result as part of the welding thermal cycle. Finally, replicate mechanical testing of these joints in both tensile shear and cross tension configurations showed considerable consistency in measured strengths. Of interest, those made with the AA5182 material showed higher strengths but consistently failed interfacially. Those made with the AA6451 material showed slightly lower strengths but consistently with a button failure mode. This was related to heat affected zone softening in the precipitation hardened AA6451 joints, providing a preferred failure path during testing.
Conclusions: Capacitor discharge based resistance spot welding appears to offer considerable potential for creating aluminum to steel joints in automotive assembly. Welding itself shows considerable process robustness, and joints appear to have strengths consistent with conventional aluminum spot welding. Microstructural analysis shows that the resulting joints were created by locally melting the aluminum sheet adjacent to the steel, creating highly localized melt zones. The short thermal cycles and implied rapid cooling rates appear to result both solidification with minimal compositional segregation and secondary growth of aluminum-steel intermetallics. Finally, replicate mechanical testing for both attached aluminum sheet types and in both tensile shear and cross tension configurations shows that consistent mechanical properties can be readily achieved.