Defect Detection And Three-dimensional Visualization Characteriztion Of Complex Lattice Structure In 3D Printing

Lattice structure is a new type of porous material with ordered periodic arrangement.It has small volume density,large specific area,light weight and other structural characteristics and fire resistance,energy absorption,heat and other functional characteristics.At present,it is widely used in aerospace,medical and other fields.Because complex lattice structures prepared by 3D printing are prone to defects such as not fusion,cracks,holes and so on,the performance and reliability of structural parts will be greatly reduced and they cannot meet the stringent performance requirements and urgent needs of current advanced structural systems.Therefore,it is necessary to carry out in-depth research on defects.Master the types,sizes,locations,distribution and other characteristics of defects,the influence of internal defects on the structure-function performance is further studied and analyzed.In order to solve this problem,a 3D visualization method for defects in complex lattice structures is studied in this paper.The main research content is as follows:Firstly,Cone-beam CT scanning technology was used to detect complex lattice structures in 3D printing,due to the influence of objective factors such as X-ray quantum,measurement system and reconstruction algorithm,noise and artifacts appear in the obtained grayscale tomography images,which affect the imaging quality.In order to ensure the image recognition accuracy,this paper adopts two different methods to preprocess the grayscale tomography image.Median filtering combined with image grayscale enhancement and Channel_shift_range function in deep learning image preprocessing were used to preprocess grayscale tomography,and the results of the two methods were compared.Secondly,based on the grayscale tomography obtained by CT scan,Through the analysis of the structural information features in the image,it is concluded that there is a grain rational feature difference between the defect position and the non-defect position.According to this characteristic difference,a method based on gaussian mixture entropy is proposed to automatically identify defects.By positioning the coordinates of the defect location,and using the method of binarization combined with Otsu to accurately segment the defect.Finally,in view of the problem that the two-dimensional defects segmented are not enough to reflect the specific morphology characteristics and distribution of defects in lattice structure.Based on the image sequence of segmented defects,a 3d viewable view of internal defects of lattice structure is reconstructed by ray-projection method,so as to realize the research method of 3d segmentation of internal defects of complex lattice structure.The feasibility of the proposed method is verified by calibrating the position coordinates of the segmented defect and the actual structural parts.And the defects are characterized and analyzed to provide strong data support for the health assessment of structural parts.