Faculty Awards for Women: Tolerance Evaluation Algorithms and Effective Inspection Procedures for Coordinate Measuring

过去的研究已经为边缘、平面、圆形和圆柱形特征开发了合适的参数化，结合线性和非线性最小二乘拟合，允许对许多几何公差（形状位置、方向、位置）以及传统的直接公差进行统计估计。此外，还开发了基于CAD数据库的二维检测任务离线编程能力。现在将扩大这项研究，以实现以下目标：制定连续抽样方法和确定零件可接受性的成本模型；制定正确估算零件基准的方法；纳入系统特征偏差和机器误差；将方法扩展到其他特征，如雕刻表面和复合材料公差；制定探头选择、零件夹具和定位方法；研究使用PDES/STEP标准来支持自动离线编程；并在制造质量实验室使用触觉和光学三坐标测量机实施和测试方法，并将其应用于零件检测和过程控制。

Past research has developed suitable parameterizations for edges, planes, circular and cylindrical features that, in combination with linear and nonlinear least squares fits, allow statistical estimates of many geometric tolerances (form position, orientation, location) as well as conventional direct tolerances. In addition, offline programming capability from a CAD data base for 2D inspection tasks has been developed. This research will now be extended to accomplish the following objectives: develop a sequential sampling methodology and a cost model for determining acceptability of a part; develop methods for correctly estimating part datums; incorporate systematic feature deviations and machine error; extend the methodologies to other features such as sculptured surfaces and composite tolerances; develop methods for probe selection, part fixturing and orientation; investigate the use of the PDES/STEP standards in support of automatic offline programming; and implement and test the methods utilizing the tactile and optical CMM's in the Manufacturing Quality Lab, and apply them to both part inspection and process control.