Prediction of Weld Width of Shielded Metal Arc Weld under Magnetic Field using Artificial Neural Networks

I. Abstract: The prediction of the optimal weld bead width is an important aspect in shielded metal arc welding (SMAW) process as it is related to the strength of the weld. This paper focuses on investigation of the development of the simple and accurate model for prediction of weld bead width of butt joint of SMAW process. Artificial neural networks technique was used to train a program in C++ with the help of sufficient number of weld ing data sets having input variables current, voltage, speed of weld ing and external magnetic field and output variable weld bead width. These variables were obtained after welding mild steel plates using SMAW process. The welding set-up was mounted on a lathe machine. In this paper, the effect of a longitudinal magnetic field generated by bar magnets on the weld bead width was experimentally investigated. Using the experimental data a multi-layer feed forward art ificial neural network with back propagation algorithm was modeled to pred ict the effects of weld ing input process parameters on weld width accurately. It was found that welding voltage, arc current, weld ing speed and external magnetic field have the large significant effects on weld bead width. It has been realized that with the use of the properly trained program, the pred iction of optimal weld bead width becomes much simpler to even a novice user who has no prior knowledge of the SMAW process and optimizat ion techniques.