Intelligent Modeling Using Adaptive Neuro Fuzzy Inference System (ANFIS) for Predicting Weld Bead Shape Parameters During A-TIG Welding of Reduced Activation Ferritic-Martensitic (RAFM) Steel

Abstract

Reduced-activated ferritic-martensitic steels are considered to be the prime candidate for structural material of the fusion power plant reactor design. Tungsten inert gas (TIG) welding is preferred for welding of those structural materials. However, the depth of penetration achievable during autogenous TIG welding is very limited and hence productivity is poor. Therefore, activated-flux tungsten inert gas (A-TIG) welding, a new variant of TIG welding process has been developed in-house to increase the depth of penetration in single pass welding. In structural materials produced by A-TIG welding process, weld bead width, depth of penetration and HAZ width decide the mechanical properties and in turn the performance of the weld joints during service. To obtain the desired weld bead geometry, HAZ width and make a reliable quality weld, it becomes important to develop predictive tools using soft computing techniques. In this work, adaptive neuro fuzzy inference system is used to develop independent models correlating the welding parameters like current, voltage and torch speed with bead shape parameters like weld bead width, depth of penetration, and HAZ width. During ANFIS modeling, various membership functions were used. Triangular membership function provided the minimum RMS error for prediction and hence, ANFIS model with triangular membership functions were chosen for predicting for weld bead shape parameters as a function of welding process parameters.