Vibration based detection of damage in fixed-fixed I-Beam using Particle Swarm Optimization

Abstract

The service life of any structure must be smooth and safe. A breakdown period is start because of damages on the structure. In the literature, various studies are found which deal with the safety of structural beams. The risk of damage development is assessed especially by structural health monitoring during the service. Transverse cracks are a common phenomenon in beams and detection of these cracks is a difficult task. The present study deals with damage detection in fixed-fixed I-section beams used for supporting of the structures. Modal analysis of the beam is carried out for thirty-nine crack locations from 10 mm to 390 mm at the interval of 10 mm length and for crack depth from 0.5 mm to 1.5 mm at the interval of 0.1 mm and the frequency data is obtained. The crack effect on beam natural frequency is analysed for the first three modes. The value of beam natural frequency is observed to decrease with presence of crack. The natural frequency of the cracked beam is observed same as of the healthy beam at specific locations. Experimental validation is carried out using FFT analyser, Photon pro+© software, impact hammer and accelerometer. To detect the damage, a mathematical model is developed and used as an inverse problem. For getting faster and accurate results Particle Swarm Optimization (PSO) algorithm is applied, which is nature inspired meta heuristic algorithm. The PSO algorithm is programmed in MATLAB© software. Algorithm uses the modal frequencies as an input and output of the algorithm is depth of crack and location of crack. The experimentation and simulation results show convergence with an error up to 10%. It is found that PSO gives satisfactory result for crack detection in fixed-fixed I- beam.