4.7 Article

Topology optimization of hyperelastic structures using a modified evolutionary topology optimization method

期刊

STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION
卷 62, 期 6, 页码 3071-3088

出版社

SPRINGER
DOI: 10.1007/s00158-020-02654-9

关键词

Topology optimization; Nonlinear; Modified evolutionary topology optimization; Hyperelastic

资金

  1. National Natural Science Foundation of China [51675525, 11725211]

向作者/读者索取更多资源

Soft materials are finding widespread implementation in a variety of applications, and it is necessary for the structural design of such soft materials to consider the large nonlinear deformations and hyperelastic material models to accurately predict their mechanical behavior. In this paper, we present an effective modified evolutionary topology optimization (M-ETO) method for the design of hyperelastic structures that undergo large deformations. The proposed M-ETO method is implemented by introducing the projection scheme into the evolutionary topology optimization (ETO) method. This improvement allows nonlinear topology optimization problems to be solved with a relatively big evolution rate, which significantly enhances the robustness. The minimal length scale is achieved as well. Numerical examples show that the proposed M-ETO method can stably obtain a series of optimized structures under different volume fractions with smooth boundaries. Moreover, compared with other smooth boundary methods, another merit of M-ETO is that the problem of the dependency on initial layout can be eliminated naturally due to the inherent characteristic of ETO.

作者

我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。

评论

主要评分

4.7
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

Article Engineering, Multidisciplinary

TONR: An exploration for a novel way combining neural network with topology optimization

Zeyu Zhang, Yu Li, Weien Zhou, Xiaoqian Chen, Wen Yao, Yong Zhao

Summary: The paper introduces a method for topology optimization using neural networks, the TONR framework, which allows flexible design variable updates and sensitivity analysis. With this approach, optimized structures for different optimization problems can be obtained stably without the need to construct a dataset beforehand.

COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING (2021)

Article Engineering, Multidisciplinary

Topology optimization via implicit neural representations

Zeyu Zhang, Wen Yao, Yu Li, Weien Zhou, Xiaoqian Chen

Summary: With the rapid development of artificial intelligence (AI) technology, scientific research has entered a new era of AI. The cross development between topology optimization (TO) and AI technology has been receiving continuous attention. This paper introduces the concept of Implicit Neural Representations from AI into the TO field and establishes a novel TO framework called TOINR.

COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING (2023)

Article Thermodynamics

A general differentiable layout optimization framework for heat transfer problems

Xianqi Chen, Wen Yao, Weien Zhou, Zeyu Zhang, Yu Li

Summary: In this work, a general and differentiable heat source layout optimization framework based on parameterized level set functions is proposed. The framework incorporates Heaviside projection for an analytical description of the heat source intensity function and automatic differentiation technique for sensitivity analysis. An adaptive multiresolution FEA method is introduced to eliminate gradient oscillations caused by finite element discretization. Numerical experiments demonstrate the positive effects of the adaptive multiresolution strategy and the effectiveness of the proposed approach in heat conduction problems.

INTERNATIONAL JOURNAL OF HEAT AND MASS TRANSFER (2023)

Article Engineering, Multidisciplinary

An Improved Data-Driven Topology Optimization Method Using Feature Pyramid Networks with Physical Constraints

Jiaxiang Luo, Yu Li, Weien Zhou, Zhiqiang Gong, Zeyu Zhang, Wen Yao

Summary: An improved deep learning model is proposed in this study, which learns the physical laws of topology optimization through a feature pyramid network, integrating pixel-wise errors and physical constraints. By adjusting the time when physical constraints are added, the balance between training cost and effect is achieved, ultimately verifying the effectiveness of the method.

CMES-COMPUTER MODELING IN ENGINEERING & SCIENCES (2021)

暂无数据