Experimental design using the Taguchi method for the development of conductive textiles used in flexible thermoelectric generators

This paper presents several aspects of the robust experimental design methodology using the Taguchi method to develop electrically conductive textiles. These conductive textiles will be used to make thermoelectric generators based on the Seebeck effect. Since the experimental development involves the use of the magnetron sputtering method with more than three variables, the Taguchi method was selected to observe how different parameters (5 independent variables such as argon flow, power, pressure, deposition surface area and metal type used for solid targets) influence the mean and variance of the process performance defined by the thickness of the metal film deposited (dependent variable). In the experimental design framework, the analysis using the Taguchi method, followed by optimisation, helped select the optimal experiments from the set of possible experiments. This methodology reduced the number of experiments by 21–42% and minimised resource consumption (e.g., metal targets, argon, energy).