M.R Forouzan - Assistant Professor, Mechanical Engineering
M.R Niroomand - PhD, Student of Mechanical Engineering
A Heidari - Faculty, Islamic Azad University, Khomeinishar Branch
S.J Golestaneh - MSc, Sadid Pipe and Equipment Co., Tehran

Submerged arc welding (SAW) is used extensively in industry to join metals in the manufacture of pipes of different diameters and lengths. The main problem faced in the manufacture of pipes by the SAW process is the selection of the optimum combination of input variables for achieving the required qualities of weld. This problem can be solved by the development of mathematical models through effective and strategic planning and the execution of experiments. The main aim of this study is to optimize the welding process to reduce the residual stresses of the welded pipes. So a new combined method which is named TAANGA (Taguchi‐ANn‐GA) is used for optimization. This method is combination of taguchi method in design of experiment (DOE), artificial neural network (ANN) and genetic algorithm (GA). Taguchi method was used for better training of neural network.The orthogonal array L' 32 from taguchi standard arrays was chosen. Then according to this array by use of finite element model training data was produced. A neural network is used to construct the relationships between welding process parameters and residual stresses in submerged arc welding. An optimization algorithm called genetic algorithm is then applied to the network for searching theprocess parameters with an optimal residual stresses. Results show very much reduction in maximum welding residual stress using this new method.

Submerged Arc Welding, Optimization, TAANGA, Taguchi, Artificial Neural Network,Genetic Algorithm