Fuzzy Logic Temperature Control for the Continuous Resistance Welding of Fiberglass/Polypropylene Laminates

Presented in this thesis is the approach taken to implement a real-time temperature control system into the continuous resistance welding process (CRW) for thermoplastic composite materials. To accomplish this, the first step was to create a numerical model of the process. A 3D FEM heat transfer model was created in COMSOLRTM Multiphysics that coupled the electrostatics of the process to the heat generation. The result was a heat transfer model that predicted the weld temperature within less than 10% error of experimental values. Temperature data from this model was extracted and used to identify a second order linear model for use online in estimating the weld temperature. A fuzzy logic controller was designed and implemented in real-time with the experimental setup using the temperature estimate as the feedback signal. The result was a controller capable of controlling the weld temperature within 10% of the reference temperature.