Determination of proper parameters for ultrasonic welding of copper plate with copper wire strands

2.1. Design of experiments (DoE) - statistical analysis using Minitab software

Design of Experiments (DoE) is a practical and organized way for engineers to understand how different factors influence key outcomes. It is widely used in the manufacturing world and the automotive industry, within the Six Sigma approach, using software such as Minitab [9], [10].

Six Sigma is a data-driven methodology and philosophy used in the automotive industry to improve processes, reduce defects, enhance quality, and increase overall efficiency [11], [12]. It uses two process capability indices to measure process specifications (see Table 1).

The Sigma level is calculated by multiplying the Process Capability Index (Cpk) value with 3.

2.2. DoE Roadmap - determination of proper parameters for ultrasonic welding

After choosing the method, we have clarified the eight steps that are part of the process and defined the actions needed for each of them, tailored toward our specific needs. This resulted in a roadmap that guided us throughout the experiments.

Schematically, the DoE method can be represented in Fig. 2.

Fig. 2Design of experiments method

3.1. Conduct the experiment

To conduct the experiment, we used a TT7 welding machine from Telsonic.

The schematic below (Fig. 3) illustrates the basic principles of the welding process. As the sonotrode moves back and forth, termed the amplitude, the part being welded also moves over the lower part, which is being clamped rigidly in the anvil. The resultant welds provide very good electrical conductivity and strong welds with no structural changes in the parent material.

For the actual testing, we used fifteen sets of materials, each consisting of ten wire-to-copper-plate pairs (as seen in Fig. 4). The specimens used in this work are cleaned thoroughly using isopropyl to remove surface oxides and other impurities [13].

Based on the literature survey [14]-[16], the predominant process variables such as clamping pressure, amplitude of vibration of the sonotrode, weld time, and weld specimen part dimensions are considered for experimentation. In our experiment, we determined three parameters have the most influence on the quality of the weld: pressure, amplitude, and energy. Various values were considered according to the established design of the experiment (see Table 2).

Fig. 3The welding process diagram

Fig. 4Example of material sets used in the experiment

The resulting welds were evaluated visually and measured for welding height, welding time, absolute welding distance and differentiated welding distance.

Destructive tests were performed to determine the strength of the weld. Pull force testing (see Fig. 5) in ultrasonic welding is a method of measuring the strength of a weld by applying a pulling force to the joint. Peel tests are conducted by peeling one side of the sheet metal away from the other, often using a roller.

Fig. 5Example of pull test equipment

3.2. Data acquisition

The resulted dataset was collected and introduced in Minitab, then analyzed using DoE. We observed the best input parameters for the optimal welding process, but also working values for the input parameters that can be successfully used considering efficiency as a value-added element in the decisions making process. This means that we have interpreted the results considering not only the value levels at which the resulted weld is the closest to perfection, but also the value levels that will yield a good result, using less resources in terms of energy consumption.

Minitab can create Contour plots with the response variable on the $z$-axis and the two control variables on the $x$- and $y$-axes, used to identify optimal values of the predictor variables for a given response variable [15].

In Fig. 6. we can observe the best values for energy are situated between 13000 and 15000 W/s, but that values between 5500 and 13000 W/s could also be used with good results (light blue contour).

Fig. 6Contour plot of visual evaluation vs. energy and amplitude

Consequently, in Fig. 7 we observe that the best values for pressure are situated between 2.73 and 2.8, with an amplitude of 85 % to 93 %, while we will still get good results by using pressure between 2.55 to 2.73 if the amplitude is preserved:

Fig. 7Contour plot of pull test vs. pressure and amplitude

However, in Fig. 8 we can observe a case in which if the energy is dropped to values between 5500-6100 W/s or below the result is a bad welding.

In this case, the optimal energy level would be between 13000-15000 W/s, with working values situated at 9000-13000 W/s.