Topshape: an attempt to create design proposals including manufacturing constraints
Topshape: an attempt to create design proposals including manufacturing constraints
复制标题
DOI：
10.1504/ijvd.2002.001997
复制
发表时间：
2002
影响因子：
0.5
通讯作者：
L. Harzheim;G. Graf
中科院分区：
文献类型：
--
作者：
L. Harzheim;G. Graf
拓扑优化已被证明是开发过程早期阶段非常有效的工具，特别是对于铸造零件。然而，设计方案的解释通常非常困难，因为优化算法中没有包括制造限制。特别地，产生具有许多孔的中空方案的趋势使得难以产生可制造的部件。程序TopShape已被开发以尝试将制造约束考虑在内。它是形状优化（CAO）和拓扑优化（SKO）程序的综合。组件设计通过拓扑优化中的密度分布来描述，而定义如何修改形状的算法取自形状优化区域。此外，还添加了算法，以确保模型的连通性并防止产生孔洞。因此，拓扑结构在优化过程中保持不变，并且抑制了产生中空设计的趋势。由此产生的方法是一种非常灵活的形状优化方法，其结果不是一个非常详细的设计，而是一个粗略的建议，现在更容易解释。在本文中，结果将显示，以证明该方法。此外，还将讨论最小厚度控制、规定的滑动方向和频率问题的扩展等选项。
Topology optimisation has been proved to be a very effective tool in the early stage of the development process, in particular for cast parts. Nevertheless, the interpretation of the design proposals is often very difficult because there are no manufacturing restrictions included in the optimisation algorithms. In particular, the tendency to create hollow proposals with a lot of holes makes it difficult to create components which can be manufactured. The program TopShape has been developed to attempt to take manufacturing constraints into account. It is a synthesis of programs for shape optimisation (CAO) and topology optimisation (SKO). The component design is described by means of a density distribution as in topology optimisation, whereas the algorithm defining how to modify the shape is taken from the shape optimisation area. Furthermore, algorithms have been added to ensure the connectivity of the model and to prevent the creation of holes. Hence, the topology remains unchanged during the optimisation process and the tendency to create a hollow design is suppressed. The resulting method is a very flexible shape optimisation method, in which the result is not a very detailed design but a rough proposal, which is now easier to interpret. In this paper, results will be shown to demonstrate the method. In addition, options such as minimum thickness control, prescribed slide directions and the extension to frequency problems will be discussed.