Numerical and Experimental Investigations on Damage Detection in Joints Based on Statistical Energy Analysis Like Approach

Abstract

Engineering structures have to be regularly monitored to avert catastrophic failure and for newlinemaintenance, etc. The extent of damage has frequently been used for low-frequency vibration newlinebased health monitoring problems. However, there are incipient damage effects on newlineconnections of structures with spot welds or bolted joints, etc., which affects mainly on newlinehighest modes, rather than on lowest. Energy based approaches like statistical energy analysis newline(SEA) is one of the widely used methods in such conditions. The present study focuses on the newlinenumerical and experimental investigation on damage detection in plates with lap joint newlineconfigurations viz. spot welded, bolted and adhesive bonded joints using statistical energy newlineanalysis like (SEAL) approach. Two materials mild steel and acrylic have been used in the newlineinvestigation. In the first phase, studies have been carried out to investigate the effects of newlineinternal loss factor on the estimation of coupling factors of mild steel plates and the finite newlineelement models for spot welds, bolted joints and adhesive bonds. In the second phase, forced newlineharmonic analysis is performed experimentally and numerically using commercially available newlinefinite element tool (ANSYS V13) to obtain the velocity responses, total energies and coupling newlinefactors of the assembly of two plates with lap joint for the healthy and damaged newlineconfigurations. Further, the velocity and acceleration responses have been simulated by FEA newlineand predicted by the SEAL approach for an assembly of three plates with lap joint newlineconfigurations and compared with experiments for healthy and damaged configurations. newlineResults have revealed that at low damping, the coupling factors computed by the analytical newlineapproach are overestimated and the coupling factors computed by finite element analysis and newlinethe experimental SEA is observed to be more accurate. Responses predicted at low newlinefrequencies are found to be not accurate due to the reduction in modal density, modal overlap newlineand violation of assumptions in SEA like approach.