期刊
COMPOSITE STRUCTURES
卷 276, 期 -, 页码 -出版社
ELSEVIER SCI LTD
DOI: 10.1016/j.compstruct.2021.114374
关键词
Laminated composites; Structural Optimization; Hyperbolic function parametrization; Topology Optimization; Multi-material optimization; Composite sizing optimization
资金
- Swedish Energy Agency [P48175-1]
- Swedish Electromobility Center3 (SEC)
HFP introduces a novel way to parametrize candidate materials in optimization problems, using a filtering technique based on hyperbolic functions. Despite the additional non-linearity introduced, HFP has fewer optimization variables and constraints compared to DMTO and SFP, showing consistent performance in optimizing composite plates.
This paper presents a new discrete parametrization method for simultaneous topology and material optimization of composite laminate structures, referred to as Hyperbolic Function Parametrization (HFP). The novelty of HFP is the way the candidate materials are parametrized in the optimization problem. In HFP, a filtering technique based on hyperbolic functions is used, such that only one design variable is used for any given number of material candidates. Compared to state-of-the-art methods such Discrete Material and Topology Optimization (DMTO) and Shape Function with Penalization (SFP), HFP has much fewer optimization variables and constraints but introduces additional non-linearity in the optimization problems. A comparative analysis of HFP, DMTO and SFP are performed based on the problem of maximizing the stiffness of composite plates under a total volume constraint and multiple manufacturing constraints using various loads, boundary conditions and input parameters. The comparison shows that all three methods are highly sensitive to the choice of input parameters for the optimization problem, although the performance of HFP is overall more consistent. HFP method performs similarly to DMTO and SFP in terms of the designs obtained and computational cost. However, HFP obtains similar or better objective function values compared to the DMTO and SFP methods.
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。
推荐
暂无数据