【作者】 吴朝生；

【导师】 陈树勋；

【作者基本信息】 广西大学 ， 机械设计及理论， 2004， 硕士

【摘要】 本文基于前人结构优化理论与方法研究成果，从工程应用的实际出发，将已有雏形的以ANSYS为分析器的结构优化软件SOGA的功能从只能优化桁架结构进一步扩充到可以优化杆-梁组合结构，使得该软件完全能够处理各种类型的杆系结构，实用性得到了进一步提高。 本文主要研究了ANSYS的各种数据接口文件，实现了五种典型截面梁(圆截面梁，矩形截面梁，工字截面梁，盒式截面梁和圆环截面梁)的截面细节尺寸变量与任意截面梁的截面特性变量的解析灵敏度计算以及优化模块与分析模块的数据交流，并将ANSYS分析器与结构优化导重法有机集成，形成了一个通用高效的杆系结构优化软件。本文研究重点解决了利用ANSYS作为分析器求解杆系结构解析灵敏度的难点，实现了杆系结构构件尺寸的动力、静力、刚度、强度优化。通过对几种不同截面形状的梁结构进行多约束最小重量优化设计计算，结果表明该软件研制正确、可行，充分了发挥ANSYS与导重法两者的长处，具有优化效率高、效果好、实用性强的优点。本文采用了变量与函数归一化、同类性态函数包络化等处理方法，减少了优化约束的数目，大大提高了计算效率。更多还原

【Abstract】 Based on previous research works of structure optimization design, from the practical need of engineering application, this paper extended functions of the software SOGA which can only optimize truss structure by using ANSYS software as a analysis tool, in order to further optimize the combination structure of truss and frame .So the practicability of SOGA software has greatly been improved.The paper mainly studied some ANSYS data files, realized the analytical sensitivity analysis of five typical section beam (circle section beam, rectangle section beam, GONGZI section beam, box section beam and cirque section beam) and data interchange between optimization module and analysis module, integrated ANSYS with Guide-weight criteria method and formed a universal and efficient optimization software of combination structure of truss and frame. It emphatically discussed the analytical sensitivity analysis of two kinds of detail measurement variables of section and feature variables of section and successfully solved the difficulty that is how to get analytical sensitivity of combination structure of truss and frame and fulfilled optimum design of dynamics and static and intensity and stiffness for link system structure. Through successfully optimizing. structures of several section beams, the results proved the development of the software is correct and feasible, fully used the advantages of ANSYS and Guide-weight method which is high effective, convenient and practical. In order to decrease the scale of optimal constraints, this paper adopted the method of enveloping to normalize variables and envelope objective function and constraint functions, so the number of constraints greatly decreases and the efficiency of computing has greatly improved.更多还原