Research On Three-dimensional Visualization Method Of Insulation Layer Of Solid Rocket Motor

One of the key steps in the engine manufacturing process involves attaching insulation layers.The uniformity of the insulation thickness is an important index to measure whether the insulation paste meets the process requirements.Therefore,for the solid rocket motor with large aspect ratio structure,it is of great practical value to study the visual display method of the insulation thickness to quickly judge whether the insulation paste meets the process requirements.For the solid rocket motor with a large aspect ratio structure of 10 meters in length and 1meter in diameter,the maximum thickness of the insulation layer is only 20 mm,and the thickness change is not greater than 0.5mm.In such a structure,the slight thickness variation may be lost when the general appearance of the insulation layer surface is displayed by the conventional point cloud processing method.Therefore,on the basis of three-dimensional coordinate point cloud data on the surface of the insulation layer,this thesis studies the registration,simplification,interpolation,reconstruction and three-dimensional visualization of point cloud data of the insulation layer.The main work is as follows:(1)On the basis of describing the working principle of the insulation layer thickness detection system and the point cloud data acquisition method of partial lap scan,aiming at the dislocation of point cloud coordinates in the lap part,thesis studies the registration method of rough before fine,calculates the FPFH(Fast Point Feature Histograms)feature descriptor of the point cloud data in the lap area to complete rough registration,and then uses ICP(Iterative Closest Point)algorithm to perform fine registration.Then the registration of point cloud data on the surface of the insulation layer is completed.(2)As the point cloud data of the insulation layer are densely distributed in the axis direction and sparsely distributed in the circumferential direction,the traditional bounding box simplification algorithm may excessively remove the point cloud data in the circumferential direction.In this thesis,an adaptive bounding box reduction algorithm based on curvature and density is studied.This method can not only achieve the reduction effect in the non-characteristic region and the axis direction,but also retain the point cloud data of the insulation layer completely in the characteristic region and the circumference direction.In order to solve the problem of missing point cloud data of single scanning line,inverse distance weighting method is used to repair point cloud.At the same time,in view of the sparse distribution of scanning line point cloud data in the radial direction,which may lead to the loss of the variation of the surface thickness of the insulation layer,the point cloud interpolation method based on cubic B-spline was studied to realize the feature enhancement of the point cloud data of the insulation layer.(3)Aiming at the problem that small thickness changes cannot be clearly presented while displaying the overall picture of large size insulation layer,this thesis studies a 3D reconstruction method based on feature preservation.The point cloud data is divided into feature region and non-feature region.For the scan line point cloud in the non-feature region,the point cloud data is selected at equal intervals and projected onto a two-dimensional plane.In this plane,scan line triangulation algorithm is used to generate a large triangular grid,and the result of the split is projected into a three-dimensional space.Aiming at the scan line point cloud in the feature region,the region growth method is used to triangulate it to generate a dense triangular mesh.On this basis,a color mapping model was constructed,and the Euclidean distance from the point cloud to the engine axis and the curvature values of each point were combined to obtain the thickness variation information of the point cloud data,which was mapped as color information and rendered on the surface of the insulation layer.The experimental results show that the proposed method can accurately display the thickness variation of the insulation layer less than 0.5mm under the condition of displaying the overall contour of the insulation layer.