Research On Measuring And Modeling Method For Free-Form Surface Of Large Thin-Walled Spherical Crown Workpiece

With the rapid development of China’s aerospace industry,the processing and manufacturing requirements of some large and complex workpieces in major engineering fields such as aerospace are becoming more and more strict.This kind of workpiece has the characteristics of large space size and thin shell structure.As a large key component in the last stage propellant tank of launch vehicle,large thin-walled spherical crown workpiece needs to be sprayed with thermal insulation layer on the workpiece surface and processed by profiling.However,the large thin-walled spherical crown workpiece is easy to deform in the process of processing and transportation,so the thermal insulation layer cannot be processed according to the original theoretical design.Therefore,this kind of workpiece needs to be regarded as a free-form surface.With the help of the key technology of reverse engineering,the workpiece is reverse formed first,and then processed according to the forming data.Aiming at the key problems in aerospace,such as the difficulty of ensuring the profiling accuracy of the thermal insulation layer of large thin-walled spherical crown workpiece,this paper studies the complex free-form surface measurement trajectory planning algorithm based on reverse engineering,the point cloud data processing algorithm,and the point cloud data three-dimensional reconstruction and machining trajectory generation method.The research of this paper has important theoretical significance and engineering application value.In order to realize the profiling processing of thermal insulation layer of large complex spherical crown thin-walled workpiece,the technical requirements of profiling processing of such large thin-walled workpiece are analyzed,and the technical process of reverse forming and profiling processing of large complex free-form surface workpiece is studied.A machining scheme of scanning measurement machining integration for large complex free-form surface workpiece is proposed.In the scanning measurement machining integration scheme,according to the profile characteristics of the measured workpiece,two different non-contact measurement schemes are proposed.The measurement methods and data characteristics of the two schemes are analyzed.According to the advantages and disadvantages of different schemes,the non-contact point by point scanning measurement scheme is determined.Feature extraction is the most critical part of complex surface point cloud data processing.Aiming at the problem that the existing algorithms can not recognize the subtle features in the actual measurement point cloud data of complex workpieces,a subtle feature extraction algorithm is proposed.The covariance matrix of the point neighborhood is established by constructing the L₁ median points in the point cloud neighborhood,so as to improve the resistance of the central point to outliers.The quotient of curvature change parameters between k+1 neighborhood and k neighborhood is defined as the feature descriptor of point data to amplify the feature difference of feature points.An adaptive local threshold detection operator based on Poisson distribution is constructed.The selected seed points are grown to mark the feature points.Through simulation experiments,the feature recognition performance of the algorithm to the general model and the sensitivity to the neighborhood are verified.Finally,taking the point cloud data of large spherical crown workpiece as the input,it is verified that the algorithm can effectively recognize the subtle features in the surface model.The scanning point cloud data of large complex spherical crown thin-walled workpiece contains rich detail features.In practical engineering applications,some details are unnecessary and need to be smoothed or erased.Aiming at the disadvantages and shortcomings of traditional point cloud denoising and filtering methods in erasing the detailed features of point cloud,a point cloud data feature smoothing algorithm based on convolutional neural network is proposed.The algorithm constructs samples on the basis of feature recognition,and constructs input samples from non feature data around feature points.At the same time,a multi-channel convolution neural network structure is proposed,which can fit the local surface highly fitting the original model by using the data information of adjacent non feature points around the feature points,and then realize the smooth operation of the feature points.The algorithm simulates the weld characteristics of spherical crown workpiece point cloud data.Experimental results show that the proposed method has excellent feature smoothing performance and will not affect other non feature regions of the model.The triangular mesh reconstruction method and tool path planning algorithm for free-form surface point cloud data of large complex spherical crown thin-walled workpiece are studied.The vertex offset method of triangular mesh model is studied,and the defects in the existing vertex offset operation are analyzed.Aiming at the offset error of triangular mesh model based on vertex normal vector,a correction algorithm based on adjacent mesh vertex normal vector is proposed.The mathematical model of distance compensation weighted by offset error of adjacent vertices is established.Experiments show that the proposed method effectively improves the offset effect of triangular mesh model.Taking the triangular mesh surface model of large complex spherical crown thin-walled workpiece as the research object,a tool location generation algorithm based on the intersection of section planes is compiled.The position relationship between the cutting plane and different triangular patches is analyzed,and the coordinate calculation of the tool locus is realized.Finally,the measurement method,point cloud data processing algorithm and tool locus generation algorithm are successfully applied to the integrated equipment of digital measurement and processing of large common bottom components.The equal thickness processing of the thermal insulation layer of large complex spherical crown thin-walled workpiece is realized.This study has made a positive contribution to the vigorous development of China’s aerospace industry.