Finite element model updating of microwave welded lap joint with direct updating algorithm

Abstract

Abstract Microwave joining is a non-conventional joining method that can be used to join bulk metals. Different types of joints such as butt-joint, lap joint, etc can be successfully joined with this novel green technique of joining. In this paper, a domestic microwave oven having a power output of 700W, frequency of 2.45 GHz, and capacity of 25L is used to weld the SS202-SS202 and SS304-SS304 bulk metals. To characterize the lap joints Vicker’s hardness test and SEM of the welded specimens are carried out. The experimental modal analysis is used to measure the experimental eigenvalues and eigenvectors of the welded lap joints. The finite element method has been used to develop the simulated finite element model of the microwave welded joints. The developed finite element model may be very beneficial to predict the dynamic characteristics of the welded structures. However, the finite element method is a numerical tool that gives approximate results. The finite element modeling of the structures depends upon various uncertain factors such as structural material properties, dimensions, boundary conditions, etc Due to these uncertainties in the simulated finite element model, there is always an error between the simulated and experimental observations. In this paper, a lap joint of bulk metals is fabricated by using microwave hybrid heating and a finite element model updating technique such as a direct updating algorithm is proposed to update its simulated finite element model. The objective is to update the simulated mass and the stiffness matrices of the microwave welded lap joint. The experimental modal analysis is used to measure the natural frequencies of the lap joint specimen experimentally. It is observed that the direct updating algorithm can successfully update the simulated finite element model of the welded joints and reduces the error between the simulated and experimental natural frequencies. It is found that the maximum error between the predicted and measured natural frequencies reduces to 0.56% by using the proposed algorithm for welded joints.