3353

3201

كارشناسي ارشد

طراحي كاربردي

اصفهان: دانشگاه صنعتي اصفهان، دانشكده مكانيك

1385

ده، 108،[II]ص.: مصور، جدول

ص. ع. به فارسي و انگليسي

مهندسي مكانيك

ID3201

به فارسي و انگليسي: قابل رويت در نسخه ديجيتال

AbstractNowadays the use of spiral pipes in oil and gas pipelines is increased The pipes oftenweld by submerged arc welding SAW technique Despite of many benefits of thiswelding method producing high residual stress which leads to stress corrosion cracking SCC at pipe wall is big drawback for SAW welding process The SCC phenomenoncaused numerous explosion and hence many damages The main aim of this study is to optimize the welding process to reduce the residualstresses on the heat affected zone of the welded pipe So after investigating the mainparameters in reducing the residual stress a neural network model of welding is built forcalculating residual stress Inputs of this model are welding speed and inner and outerheat inputs and the output is residual stress Taguchi method in design of experiment DOE is used for better training of neuralnetwork The orthogonal array L32 from taguchi standard arrays is chosen Then weldingprocess parameters are optimized with genetic algorithm More than 60 reduce inmaximum welding residual stress is achieved using this method

محمدرضا فروزان

محمدرضا احمدزاده

حسن نحوي، احمد كرمانپور