參數化建模、結構分析與最佳化之整合軟體系統

Abstract

本文發展一套機械結構參數化建模、結構分析與最佳化之整合軟體系統。首先，設計Excel試算表以儲存結構建模以及分析的參數。在模型建立的部分，所需要的設計參數皆取自於Excel試算表。開發AutoLISP程式來自動產生模型並且輸出為SAT檔案。在結構設計的部分，分析的參數以及材料性質可以從Excel試算表中取得。利用ANSYS APDL來編寫自動化結構分析的程式，像是靜態以及動態的響應，同時分析的結果也能儲存在Excel試算表中。最後，開發C++程式來呼叫AutoCAD以及ANSYS進行結構的最佳化迭代，將迭代的結果儲存於相同的試算表中。將此整合系統應用於一些結構範例和設計一台三維零阿貝誤差晶圓檢測平台的結構，此平台的行程範圍為300mm x 300mm x 5mm，適用於量測12吋的晶圓。藉由此整合系統，相較於傳統的繪圖方式可以有效減少模型繪製的時間，同時避免重複步驟可能導致的人為疏失。此外，設計以及分析結果的資料可以儲存於Excel試算表當中，更易於分析及數據處理，改善設計的方便性以及最佳化結果。

This thesis develops an integrated software system for parametric modeling, structural analysis and optimization of machine structures. First, Excel spreadsheets are designed to store both of the modeling and analyzing parameters of a structure. In model generation, design parameters are obtained from Excel spreadsheets. An AutoCAD LISP program is developed to automatically generate the solid model of the structure and then exported as a SAT file. In structural analysis, material properties and analysis parameters can also be obtained from Excel spreadsheets. A customized ANSYS APDL program is also developed to automatically analyze structural behaviors, such as static and dynamic responses. Analyzing results are then stored back to the Excel spreadsheets. Finally, a C++ program is developed to call AutoCAD LISP program and ANSYS APDL program iteratively for performing optimum design of the structure. Results of design iterations are also stored back to the same spreadsheets. The developed integrated software system is applied to some demo structures, and good results are obtained. This software system is also used in design of a large 3D Abbe-error-free wafer inspection stage structure. The stage has travelling range of 300mm x 300mm x 5mm, and suitable for measurement of 12 inch wafers. Through this system, it can significantly reduce the modeling time and avoid the human error that may be caused by repeated steps compared to conventional methods. In addition, design and analysis result data are stored in Excel spreadsheets, those can be analyzed and processed more easily, improving design convenience and optimization results.