Analysis of Correlations Between Joint Physical Parameters and Laser Welding with Filer Wire by the Adaptive-network-based Fuzzy Inference Systems Technique
S. Cvetkovic
There is considerable challenge to make uniform joint gap mismatch in any machine and mechanism during welding process. To overcome the mismatch there is need to select the optimal welding parameters for the different gaps and mismatches. Laser welding with filler wire is used for experimental study in order to collect data for further analysing. The collected data are then used in adaptive-network-based fuzzy inference systems (ANFIS) technique in order to find and analyse the correlations between input and output parameters of the laser welding process. As the input parameters gap, mismatch, bead width and reinforcement are used. As the output parameter laser power and wire feed rate are used. Obtained results show correlations between the joint geometrical parameters and the welding parameters and it could be of practical importance in laser welding engineering.
Keywords: Fibre laser, A709 alloy steel, ER506 filler wire, laser welding, adaptive-network-based fuzzy inference systems (ANFIS), root mean square error (RMSE), correlation