Research On The Defect Identification Of The Compressor Blades Local Weight Gain Based On Vibration Signal

With economic development,Compressor plays an increasingly important role in the petrochemical industry.The design and manufacture of compressor technology is also put forward newer and higher requirements.Blade as a core component of compressor,its safety and reliability is particularly important.The main reason for failure of the blades is fatigue fracture caused by vibration.After the blade damaged,it will lead to change in the structure parameters.Therefore,this paper mainly through vibration experimental methods to research the dynamic characteristics of the blade,and then further research on the dynamic properties of the blade fracture to judge the position and mass quantity of damage,for the purpose of decreasing the mechanical failure in maximum limit,reducing maintenance time and avoid catastrophic accidents.This paper was based on the blade local weight gain as the research object simulation defects of the blades.Using the methods of correlation analysis,wavelet analysis,modal analysis to judge the defect location identification and defect degree of the preliminary quantitative of the defective blade.The main work is as follows:This papper was finding the relevant literature,several methods for the above theoretical research and analysis,experimental step plan and assess the feasibility.The excitation signal and the response signal acquisition,cross-correlation analysis and wavelet transform method,fitting a function of time and distance,find the position of the defect.Using the method of modal analysis of blades damage detection,although using the method of modal analysis on the method of damage detection of structure has been proposed for many years,in the experimental aspect,especially in the aspect of the modal analysis experiments of the sheet metal structure there is little work has done.This paper is based on the method of modal analysis of the defect blade modal experimental analysis,selecting the first four frequencies and mode shapes of leaves modal parameters,analyzing with the defect location and size changing,the change of modal parameters.Compared wavelet transform method,correlation analysis,natural frequency,mode shape and curvature mode ctural damage identification,and proposed damage detection methods.