摘要
考虑目前多数计算机辅助公差工具仅能针对具有理想几何表面的CAD模型,无法从物理几何角度真正反映制造误差,本文研究了非理想表面的多尺度表征,提出了一种能够表征产品表面及其截面轮廓宏观及微观多尺度形貌误差的非理想表面模型。首先,提出了一种利用离散小波实现形貌误差多尺度仿真的方法;其次,针对实际工件,利用离散小波对其表面及截面轮廓采样数据进行形貌误差多尺度仿真;最后,对形貌误差各尺度成分进行合成,得到具有多尺度形貌误差成分的工件三维表面及其二维截面轮廓的非理想表面模型。仿真及实验结果表明:利用提出方法可以实现具有多尺度形貌误差的非理想表面模型表征;仿真所得粗糙度R_a值与白光干涉仪测量所得值的平均相对误差不超过4%。得到的结果证明了提出方法的正确性和可行性,为更加全面地表征产品的非理想表面模型提供了有效途径。
As most of the Computer Aided Tolerancing tools can only deal with the CAD models with an ideal surface and can not reflect manufacturing errors in physics and geometry,this paper explores a multi-scale representation method for skin model shapes in the Geometrical Product Specification.A discrete data modeling method was proposed for the simulation of topographic errors of surfaces and section profiles of a product in macroscopic and microcosmic scales based on new-generation geometrical product specification.Firstly,a simulation method for multi-scale surface topographic errors based on discrete wavelet was presented.Then,discrete wavelet was used to simulate multiscale surface topography errors for sampling data of the surfaces and section profiles of a part.Finally,multi-scale surface topography errors were composed and skin models for two-dimentional profile and three-dimentional surface were acquired.The simulation and experiment results show that the proposed method represents skin models with multi-scale surface topographic errors and the average relevant error between the results of Ra obtained by a white-light interferometer and the proposed simulation method is less than 4%.The results verify the correctness and applicability of theproposed method,and provide a valid way for more comprehensive representation of skin model in new-generation geometrical product specification.
出处
《光学精密工程》
EI
CAS
CSCD
北大核心
2016年第7期1647-1654,共8页
Optics and Precision Engineering
基金
国家教育部博士点基金资助项目(No.20131103110001)
国家自然科学基金资助项目(No.51305006)
北京市教委项目(No.JC001013201402)
关键词
产品几何技术规范
非理想表面模型
离散几何
小波分析
多尺度表征
geometrical product specification
skin model
discrete geometry
wavelet analysis
multiscale representation
