| Low-speed Model Testing for High Pressure Compressor is an effective way to obtain the internal flow field of the compressor,reduce the cost of the test,and improve the design level and efficiency of the compressor.The low speed condition of the test bed with relatively low strength requirements makes it possible to replace the metal blades with resin matrix composite blades.Influenced by many factors such as injection process and mold cavity,the resin matrix composite blades are inevitably deformed which affecting the aerodynamic performance of the blades.Process parameters and mold cavity are the main factors affecting the blade dimensional accuracy.Therefore,it is of great theoretical and practical engineering value to carry out in-depth research on key technologies of blade precision injection molding for improving their dimensional accuracy and verifying the design scheme of compressor.Based on the analysis of the batch blade size characteristics and the causes of blade oversize,injection molding process parameters optimization and reverse adjustment of mold cavity are studied in this paper to solve the problem of large dimension variation and dimension deviation of blade.First,robust optimization of blade injection process parameters is carried out to improve blade dimension consistency.Secondly,the optimal design of the blade mold cavity is carried out to reduce blade dimension error.That is the following key technologies in the optimization design process are analyzed and studied: method of blade measurement planning oriented to model reconstruction,dimension deviation estimation oriented to deformation calculation,and reverse deformation methodology for blade mold cavity.On this basis,a prototype system is developed for optimization design of the blade mold cavity,which can effectively shorten the development cycle of the blade mold and improve the accuracy of the blade dimension.The main works of this study are as follows:(1)Robust optimization of blade injection molding process parameters.Combined with CAE analysis and injection experiments,deformation difference between the simulation results and the injection experiment is analyzed.Based on the maximum value of blade deformation,the simulation results are analyzed and evaluated.Then,based on CAE analysis results,a mapping model between process parameters and blade deformation is established.Finally,robust optimization design of blade parameters is carried out based on process optimization theory and Monte Carlo simulation technology.(2)Blade measurement planning oriented to model reconstruction.First,based on the geometric characteristics of the blade model,the curvature of the blade surface is reflected bythe curvature of the equal parameter curves of the mean camber surface along the blade stacking axis.The minimum common data points of the equal parameter curves are selected and the blade measurement sections are planned.Then,the cubic B-spline curve is used to approximate the blade profile.The points that needed to construct the B-spline curve are sampled as the measuring points of the profile.Compared with the existing measurement methods,the proposed method can significantly reduce the workload under the premise of modeling accuracy.(3)Blade sampling method oriented to dimension deviation estimation.Dimension variation is inevitable for batch parts.It may lead to large errors for optimization of blade mold cavity by the inspection data of a single blade.In order to solve this problem,the necessity of optimizing of blade mold cavity by calculating dimension deviation of batch blades is analyzed.A sequential sampling method for estimating dimension deviation of batch blades is proposed Firstly,displacements of the discrete points on the blade surface are employed to represent dimension error.Then,the problem of blade sampling is successfully transformed into multivariate statistical analysis.Secondly,Bonferroni simultaneous confidence intervals are employed to estimate the confidence interval of the blade surface dimension deviation.Moreover,by increasing samples,estimation of surface dimension deviation is more and more accurate.Therefore,a novel sequential sampling method is proposed based on the theory of sequential sampling,and the experiments is carried out.(4)Optimal design of blade mold cavity based on the technology of reverse deformation.A statistical processing method for blade multi-sample inspection data is studied.Cubic B-spline curves which constructed based on the statistical processing point coordinates and blade CAD information are used as the the blade profile.Registration of blade profile and corresponding design profile.The rotation angle and displacement of the registration are regarded as the twist deformation and bending deformation of the blade.Based on this,the section of the mold cavity is constructed based on the principle of reverse deformation,and the optimized mold cavity is obtained based on the surface lofting.The experimental results show that the optimized cavity can significantly improve the blade dimension accuracy.(5)Development of prototype system for blade injection mold cavity optimization.Based on UG/Open API and VC,the prototype system of blade mold cavity optimization system was developed.The system architecture and workflow were introduced.Finally,the mold cavity optimization design method was demonstrated by specific Compressor Rotor Experimental Blade. |
PDF Full Text Request