| Summary: | 碩士 === 國立臺灣科技大學 === 機械工程系 === 89 === Wear inspection and repair of a precision free-form die surface
Chang, Kuo-Hui
Abstract
A methodology used to determine the wear range and to repair the wear range of a precision free-form die surface has been proposed in this research. The study can be divided into three subjects. First, an application of 2D low-pass filters to smooth the digitized data points of different unknown free-form surfaces for reverse engineering, second, to search the wear range by using a coordinate measurement machine, third, repair of determined wear range and re-inspection.
Different 3x3 or 5x5 low-pass filters, such as average low-pass filter, median filter, and Gauss filter, were used to filter and smooth the digitized data points of unknown free-form surfaces by utilizing the spatial convolution operation concept of image processing technique for reverse engineering. Different carriers such as a New Taiwan ten-dollar, a U.S. coin, and a mouse mould cavity were first digitized by Renishaw Cyclone profile scanning machine. The digitized data points of these carriers were then filtered and smoothed by the developed software written with Visual C++. The calculated results of these carriers were loaded in the reverse engineering data processing software CopyCAD to determine the best 3D smoothing effect and section curves were also loaded in Matlab software to decide the best 2D smoothing effect. According to the tested results the 3x3 Gauss filter and 3x3 median filter have better smoothing effect then other filters.
Two different interpolation methods, namely the Composite Ferguson Spline method and the Shepard interpolation method, were used in this study to establish the CAD Model of a mouse mould cavity with uniform grid distance from the above smoothed scattered data points for the convenience of further data processing, such as the reconstruction of a surface and the configuration of the measuring path. According to the tested results the Composite Ferguson Spline method has better effect than the Shepard interpolation method. The CAD model of the mouse mould cavity was constructed with the interpolated data points. The mouse mould cavity of low carbon steel was manufactured by using Power Shape/Mill CAD/CAM system on a machining center.
Wear range was generated and simulated by little modifying the CAD Model of the carrier considering the economic reason. A two-stage automatic inspection process including rough inspection and fine inspection with a coordinate measuring machine (CMM) was proposed to search the wear range. The purpose of the rough inspection stage is to reduce the inspection time with a larger point-to-point inspection distance. In the second inspection stage a smaller point-to-point inspection distance was configured to determine the correct wear area.
The determined wear range was repaired by TIG welding process and fine re-machined on a machining center. The re-machined carrier was inspected again on a CMM to perform the error analysis. According to the experimental results the feasibility study of wear range automatic inspection and repair of wear range is proven to work.
|
|---|
