【摘要】 针对基于测量数据的发动机薄壁叶片椭圆弧形前后缘建模问题,在最小二乘法椭圆拟合的基础上,提出一种基于误差控制的叶片椭圆弧形前后缘拟合重建方法.通过设定迭代控制误差对拟合过程进行控制,当拟合误差小于等于预先设定的迭代控制误差时迭代拟合终止,可有效地避免叶盆、叶背上的点参与拟合;提出一种考虑叶片纵向光顺要求的各截面椭圆参数调整方法,可实现前后缘椭圆参数的提取以及叶片前后缘区域光顺曲面的生成.在文中方法的基础上开发了叶片前后缘建模软件模块,并通过实测数据验证了该方法的有效性.更多还原

【Abstract】 On the basis of the method of direct least square ellipse fitting, an error control-based reconstruction method for ellipse leading edge and trailing edges of thin-walled blade is proposed. By setting the iterative fitting error threshold, the ellipse iterative fitting process can be controlled. When the average ellipse fitting error is equal to or less than the threshold, the iterative fitting process would be stopped, which can effectively prevent the non-edge points being involved in the fitting process. Furthermore, considering the requirement of the vertical smoothness of the blade, a parameter adjustment method was proposed, which could help to generate smooth surfaces of the blade edges. Finally, software modules based on the discussed methods were developed and verified based on real measurement data set.更多还原