- 學位論文
影像視覺之正齒輪熱變形量測與分析
Thermal Deformation Measurement and Analysis of Spur Gears using Image Vision
摘要
本研究為應用田口方法於影像視覺之正齒輪熱變形量測與分析，利用影像辨識軟體Matlab撰寫一套以正齒輪之幾何特徵為檢測基礎的運算程式，結合田口方法L9(34)直交表進行實驗設計規劃，以不同加熱溫度、材質、表面處理與冷卻方式等參數做為實驗設計的控制因子，每一因子各具3個實驗水準，主要探討正齒輪在極高的溫度場時因熱膨脹係數的不同而產生熱變形之差異性，並以熱變形量與齒冠直徑做為品質特性來討論，找出正齒輪之最佳製程參數水準。其研究結果顯示熱變形受材質的影響最為顯著，可確認材質選擇對品質的差異性為主要之因素。 而在正齒輪之幾何特徵尺寸的量測中，須運用影像處理技術將齒輪邊緣輪廓尋找出來，利用灰階化、二值化、閉合處理、影像旋轉、影像相減與影像相乘等影像形態運算，分別對齒冠圓與齒根圓辨識、橢圓孔長軸與短軸尺寸辨識、節點與節圓直徑之辨識、周長與面積之辨識等進行量測，其量測的誤差值可達±2%內。
並列摘要
This study invesitgated thermal deformation measurement and analysis of spur gear based on Tauchi method via image vision recognition technique. The measurement analysis utilizes Matlab software to develop a set of image algorithm computing for the geometric characteristics of the spur gear. The control factors of thermal deformation measurement for the spur gear includes heating temperatures, materials, surface treatments, and cooling types and the measurement analysis is used a L9(34) orthogonal array with 3 levels of Taguchi methods. The effect of coefficient of thermal expansion (CTE) at the different extreme temperatures was investigated for finding most suitable processing parameters for addendum circle of the spur gear. Experiment results showed that the material properties are the most influential factor in thermal deformation; thereby the selection of material is the main cause leading to different product quality levels. In regarding to the geometric characteristic measurement of spur gear, image processing technique is utilized to inspect the contour of the spur gear and is based on grayscale, binary threshold, closing, image rotating, image subtraction and image multiplication. The error value of the addendum and dedendum circle, the major and minor axis of ellipse, the pitch point and the diameter of pitch circle, and the radius and area of the spur gear is under ±2%.