Research On Crack Detection Of Hydraulic Pipeline Based On Fiber Bragg Grating Sensing

As an important part of the hydraulic system,hydraulic pipeline's safe operation plays a significant role in ensuring the normal operation of the hydraulic system.However,there exist a large number of hydraulic pipeline failures in actual work,and one of the main faults is crack.The emergence of cracks and its expansion are likely to threaten the safety of the pipeline,even the entire hydraulic system.Therefore,it is necessary to conduct crack detection of hydraulic pipeline.Compared to the traditional detection methods,Fiber Bragg Grating sensing with its strong environmental adaptability,distributed multi-point measurement and other prominent advantages presents a better detection effect on hydraulic pipeline crack.In addition,the keys and difficulties in crack detection are how to identify the crack of hydraulic pipeline efficiently from the acquired FBG signal,including the degree of crack,the trend of the crack propagation,location of crack and so on.In this paper,the hydraulic pipelines are taken as the research objects respectively.After acquiring the effective strain signals of hydraulic pipelines on different work condition through Fiber Bragg Grating sensing,multifractal and its improved algorithm as well as some other methods are used to do crack detection to reach the goal of judging position,degree,trend and so on.The main research work is as follows:(1)Hydraulic pipeline crack simulation based on ANSYS.Hydraulic pipeline crack simulation is carried out through ANSYS software based on the theory of hydraulic pipeline crack generation,and the pipeline strain signals under different conditions are obtained.Through researching and comparing the advantages and disadvantages of multifractal method,wavelet analysis and empirical mode decomposition methods,the simulated crack signals are taken as the original signals to extract effective characteristic parameters which can present the crack states,crack detection performance simulation based on different methods are compared and analyzed.The results provide the theoretical basis for the actual hydraulic pipeline crack detection.(2)Dynamic crack detection research of hydraulic pipeline based on FBGsensing.The signals in the whole process hydraulic pipeline from healthy state to penetrating crack fault state are acquired by FBG strain sensor Then effect hydraulic pipeline strain signals are acquired after the pretreatments are conducted,such as temperature and strain separation and denoise.Moreover,the crack detection is carried out via multifractal and empirical mode decomposition methods,the fault characteristics are extracted,the crack propagation trend is studied,and the result can provide the basis for judging the hydraulic pipeline leakage.The result demonstrates that the multifractal methods based on multifractal detrended fluctuation analysis can conduct a more accurately and efficiently detection on hydraulic pipeline crack compared with other methods.(3)Crack detection research of hydraulic pipeline based on distributed FBG sensing.The multi point complex hydraulic pipeline signals under healthy state and several fault states are acquired by distributed FBG strain sensors.Combined with the characteristics of distributed FBG,the multifractal detrended fluctuation analysis algorithm is improved,it can transfer multi-point signals processing into image processing.After the effective pretreatments,hydraulic pipeline crack detection is conducted to distinguish the degree of cracks,judge the location of cracks.Then the results of crack detection based on improved multifractal methods are compared to the results based on traditional multifractal methods.It demonstrates that the improved multifractal methods have better crack detection performance and more obvious recognition in degree and position compared with the traditional multifractal methods.