Predicting weld bead geometry and HAZ width in modified 9Cr-1Mo steel welds using ANFIS-based models

Abstract

Modified 9Cr-1Mo steels are used as structural materials for stream generator components in power plants. Generally, tungsten inert gas (TIG) welding is used in the fabrication of these steels. In TIG welding of modified 9Cr-1Mo steels, the depth of penetration achievable in single pass is limited. Thus, a novel technique called as activated flux tungsten inert gas (A-TIG) welding has been developed in which application of flux over the joint area increases depth of penetration. In these weld joints, weld bead shape parameters like depth of penetration, bead width and heat affected zone (HAZ) width influences the mechanical properties and also the performance of the weld joints during high temperature service. Since, the fuzzy system lacks learning ability, a hybrid neuro-fuzzy system called as adaptive neuro-fuzzy inference system (ANFIS) has been used to develop models correlating process parameters with weld bead shape parameters and the models are also validated experimentally.