Study Of Damage Location And Feature Recognition Algorithm In A Plate Structure Based On Ultrasonic Guided Wave Using Waveform Centroid

Structural stability and safety are critical in industrial production.Thin-walled structures are widely used in various industries and daily life,e.g.ships,aerospace and other fields.However,structures are prone to be damaged due to the natural or human factors,such as corrosion,cracks,holes,and even fractures.If these damages are not found in time,serious security incidents may occur.Structural health monitoring(SHM)based on active Lamb wave can realize real-time and on-line detection of damage in thin-walled structures and reduce safety risks.In the field of structural damage detection,the common damage detection algorithms rely on sensor array to achieve damage location which have the following disadvantages:it needs a large number of sensors and a large amount of signal processing,which is not suitable for industrial applications.Therefore,the implementation of new algorithms and signal processing technology to reduce the number of sensors is of great significance to accelerate the development of structural health monitoring technology.In this thesis,the study of damage monitoring based on active Lamb wave in thin aluminum alloy plate is mainly focused on the algorithm development in damage location and imaging.The thesis is written as follows:Firstly,the SHM technology is briefly introduced,and the related theories of SHM based on active Lamb wave and the method of damage location and imaging technology are mainly studied.Then,the simulation model is established by using the finite element analysis software ABAQUS for theoretical analysis and simulation.Based on the simulation results,the time of flight(TOF)calculation by different methods is compared,and finally the peak to peak interval of the waveform is adopted to determine the TOF.To reduce the number of sensors and to improve the accuracy of damage location,a waveform centroid algorithm is proposed.By dividing the structure into multiple sub-regions comparing the relative positions of each group of detected signal waveform centroids,the damaged area is gradually narrowed and finally the damage sub-regions are determined.With the aforementioned method,four sensors can be used to locate the damage location to 1/8 of the detection area,and then three sensors adjacent to the sub-area can be selected to accurately determine the damage location.Damage localization algorithm is validated with experiment.After locating the damage location,based on the characteristics of ultrasonic propagation and attenuation in the plate structure,the damage size is analyzed by the damage reflection wave.Through successive iteration,both the number of sensors and the detection cost are reduced,and the location accuracy is improved.