Speed Optimization Of Laser Ultrasonic SAFT And Its Application In Detection Of Bulk And Surface Defects

In industrial production and applications,complex metal products such as cables,rails and engines may have minor defects that affect product quality.In order to achieve high precision non-destructive detection of defects,an emerging detection technology,laser ultrasonic non-destructive detection technology,has received wide attention from researchers due to the advantages of non-contact long-range excitation and wide signal band.In this paper,considering the long time consumption of laser ultrasonic synthetic aperture focusing technique(SAFT),CPU+GPU is used to improve the detection speed.For the problem of difficult localization of free-form surface defects,an algorithm combining geodesic and ellipse imaging is proposed.Firstly,to address the problem of long computation time of SAFT,a heterogeneous parallel computation method that uses CPU+GPU is proposed.Taking metal buried defect detection as an example,the computation speed is improved.SAFT-based laser ultrasonic detection technology is used to realize the localization of metal buried defects by extracting ultrasonic body wave signals that carries the defect information.Besides,the defect imaging effect of singleexcitation point and multi-excitation point is compared,and the detection error is limited in the order of tens of microns.The results show that the computation time of CPU+GPU heterogeneous parallel computation method is one hundredth of that of traditional SAFT,which significantly improves the detection speed.Then,the detection of surface defects using laser ultrasonic SAFT is realized.A defect detection technique based on laser ultrasonic ellipse imaging algorithm is proposed for the shortcomings of SAFT in detecting surface defects.This technique extends SAFT to twodimensional planes and achieves the localization of sub-millimeter scale defects on metal planar surfaces.Afterwards,the detection of the free-form surface defects is studied,a method of using geodesic distance to represent the propagation path of ultrasound along a free-form surface is proposed,and the laser ultrasound signal that propagates on the free-form surface is investigated and processed using ellipse imaging algorithm.The results show that the proposed algorithm that combines geodesic and laser ultrasonic ellipse imaging can achieve the localization of submillimeter scale defects on free-form metal surfaces.Finally,based on the principle of laser ultrasound SAFT and ellipse imaging algorithms,an experimental setup that can detect both metal body defects and surface defects is built,which verifies the feasibility of the laser ultrasound SAFT algorithm and ellipse imaging algorithm for detecting defects.