Development of fuzzy rule based system for prediction of weld bead geometry

Abstract

Welding bead is a deposit of filler metal. The weld bead quality can be determined by the welding input parameters. The factors which influence the joint quality of metals are weld bead geometry, distortion and mechanical properties, by experimental setup of Ravikumar et al., [1]. The weld bead quality was predicted using mathematical model by comparing actual value with predicted value. They acquired error of 4.316. The input parameters considered for the model is welding current, welding speed, nozzle to plate distance and the output parameters are depth of penetration, weld bead width, reinforcement and dilution. In this paper, we have applied fuzzy rule based system as it’s efficient to find the relationship between the input and output parameters to predict the bead geometries and dilution with adequate result. The purposes of the experiment is to predict the error percentage of output parameters with actual values from experimental setup thus the comparison between actual value and fuzzy rule based predicted value has been made and found that fuzzy rule based is efficient.