Resistance change modeling of sputtered thin films on flexible substrates under fatigue test

Flexible electronics are in the leading edge in the electronics manufacturing industry. In the Flexible Electronics industry, systems are built on flexible substrates - generally plastics- To produce devices that are not only flexible, but rugged and light in weight with a potential of high volume production at low cost. However, Electronic systems in general and flexible electronics in particular are susceptible to different types of stresses during processing and applications. Among these stresses, fatigue bending is common loadings that cause failure at very low stresses. In this work, high cyclic bending fatigue experiments were conducted on very thin films of sputtered copper on PET substrate. Design of Experiment tool, 2^k-1 fractional factorial design, were used to reduce the number of experimental runs. Five factors were considered including temperature, humidity, film thickness, bending radius, and frequency. The change in electrical resistance at certain number of cycles was chosen as the response for regression model to correlate the electrical resistance change with the independent variables. Furthermore, artificial neural network (ANN) was also used to model the change in electrical resistance of thin films on flexible substrate. The performance of these two models was compared.