Prediction of angular distortion in gas tungsten arc welded 202 grade stainless steel plates using artificial neural networks - An experimental approach

Abstract

GTAW process is appropriate for joining slim and medium thickness materials like stainless steel sheets. Stainless steel 202 grade has wide applications in making consistent stainless steel tubes, super warmer cylinders, cook products and so forth., Angular distortion is one such imperfection that makes the work piece twist in angular ways around the weld interface. After welding treatment is required to take out the distortion so the work piece is defect free and acknowledged. The level of angular distortion is specifically affected by the welding input parameters amid the welding procedure; along these lines, welding can be considered as a multi input procedure. This paper presents advancement of an Artificial Neural Network (ANN) demonstrate to anticipate angular distortion for different input process parameters in 202 steel gas tungsten arc welded plates The picked info process parameters were welding gun angle, welding current, plate thickness, welding speed and gas flow rate. The picked yield parameter is angular distortion. The analyses were directed dependent on five factor five dimension central composite rotatable structures with full replication system. Utilizing the test information multi layer feed forward neural system display was created and it was prepared by utilizing back propagation technique. The created model is then contrasted with the experimental results and it is discovered that the outcome acquired from neural system demonstrate is accurate in foreseeing the angular distortion.