A smart prediction tool for estimating the impact strength of welded joints prepared by vibratory welding process

Abstract

Previous researches have been devoted to development of vibratory setup for inducing mechanical vibrations into the weld pool during welding process. The designed vibratory setup produces the required frequency with suitable amplitude and acceleration in terms of voltages. This helps in producing uniform and fine grain structure in the welded joints which results in an improvement in the mechanical properties of the weld pieces at heat affected zone. This paper presents the development of a smart prediction tool by implementing generalized regression neural network to establish a relation between vibration parameters such as input voltage to the vibromotor, time of vibration and impact strength of vibratory weld joints. In order to validate the feasibility of the developed prediction tool, a comparison is made with the experimental results.