Patented in 1960, ultrasonic welding for metals, has received significant research and there have been advances in the technology. Because of these advances, the process has been developed into a practical production tool. Although the ultrasonic metal welding has many advantages including speed, efficiency, long tool life, there are several issues. For example, tool/part adhesion (sticking), part marking and the lack of the consistency and predictability of the weld strength are issues that industry currently encounters with ultrasonic metal welding. To resolve these issues of the ultrasonic metal welding process, various experiments have been conducted on aluminum 5754 alloy samples. These experiments included replacing a constant amplitude with amplitude profiling and placing buffer sheets of copper and zinc between the tool (horn) and the top part prior to ultrasonic welding. In addition, experiments were conducted to compare the consistency of weld strength for the three control modes: energy, height (post height), and time. Their results were analyzed and compared in terms of weld strength, weld consistency and weld quality.

It was seen that amplitude profiling produced an increase in weld strength, however the part marking and the tool/part adhesion were not reduced. By matching the amplitude to the various weld phases, the weld cycle was optimized to produce relatively strong welds and weld cycle time was reduced. For example, by initiating the weld cycle with a relatively high amplitude, the asperity peaks are sheared, heating and softening of the faying surfaces are completed quickly and efficiently. Then by reducing the amplitude at the end of the weld cycle, shearing of the weld is reduced and weld damage is minimized resulting in relatively strong welds.

Very simple models based on frictional heating were able to predict power dissipation reasonably well. When these models were coupled with a one dimensional heat flow solution, the initial bondline temperatures were accurately predicted. However, the predicted temperatures diverged from the measured temperatures as a function of cycle time. It is believed that this deviation was due to the over simplified assumptions used to model the process. For example, the assumption that the process could be modeled with one dimensional heat flow in a semi-infinite body was seen to be false shortly after the weld process was initiated.

The study involving the placement of buffer sheets (copper or zinc) between the tool (horn) and the top part prior to the ultrasonic welding process reduced the tool/part adherence and the part marking. However, it was seen the use of buffer sheets slightly lowered the weld strength and this was especially seen for the thicker (3-mm) parts and copper buffer sheets. It was also seen that zinc buffer sheets did not lower weld strength as much as the use of copper buffer sheets.

Also, it was concluded that the weld strength variance for those welds made with the time mode was smaller than that for the welds made in the height and energy mode.