Estimation of submerged-arc welding design parameters using Taguchi method and fuzzy logic

Fuzzy logic with the Taguchi method is proposed for estimating submerged-arc welding design parameters. Design of experiments based on the submerged-arc welding simulation is applied to the proposed approach. The finite element analysis of a steel plate is carried out to simulate the submerged-arc welding process using thermomechanical and temperature-dependent material properties. Welding simulation results on thermal-mechanical coupled physical phenomena are compared with welding test. The proposed fuzzy model is generated with triangle membership functions and fuzzy if-then rules using training data obtained from design of experiments and the Taguchi method. The objective of the present study is to develop a Taguchi method-based fuzzy model to effectively approximate the optimized arc weld parameters. To validate the fuzzy model, an approximate model based on response surface method is generated from the Taguchi design data, and then the results are reviewed by the outcomes obtained from the fuzzy model. This proposed study facilitates a quantitative decision of weld factors such as speed and output of weld that are generally expressed by qualitative linguistic terms in the welding process.