Scholarlink

<p>Lattice materials bear similarity to the crystalline materials as both have periodic structures. Recent studies have demonstrated the feasibility of imitating the microscopic strengthening mechanisms of crystals in the design of heterogeneous lattice materials with improved mechanical properties. In this paper, inspired by the structures of high-entropy alloys (HEA), a lattice material with distorted lattice based on body-centered cubic (BCC) cell is proposed. Through a multi-scale analysis, a theoretical model regarding to the distortion parameters and material stiffness is established. Based on this, a database including material geometric parameters, elastic modulus, and shear modulus is constructed. A deep neural network (DNN) model with geometric parameters as input and Young’s modulus and shear modulus as output is established, and multi-objective optimization is carried out. The results show that the introduction of lattice distortion can increase Young’s modulus and shear modulus by up to 60.7% and 7%, respectively. Then, inverse design is carried out for the large deformation behavior of HEA lattice materials. Based on the database constructed by the finite element simulations, a hierarchical neural network is proposed via the K-means method and convolutional neural network (CNN) model, which can effectively generate the lattice material with the target performance. This inverse design method paves a new pathway for the effective and efficient design of lattice materials with improved mechanical properties.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Inverse design and optimization of a distorted-cell lattice material inspired by high-entropy alloys

  • 期刊论文

  • 发表日期：

    2025年9月27日

查看全文
  • Puhao Li,
  • Keng Lin,
  • Qiang Luo,
  • Fan Yang,
  • Yi Chen,
  • Jiacheng Wu,
  • Qingcheng Yang,
  • Lihua Wang

摘要

Lattice materials bear similarity to the crystalline materials as both have periodic structures. Recent studies have demonstrated the feasibility of imitating the microscopic strengthening mechanisms of crystals in the design of heterogeneous lattice materials with improved mechanical properties. In this paper, inspired by the structures of high-entropy alloys (HEA), a lattice material with distorted lattice based on body-centered cubic (BCC) cell is proposed. Through a multi-scale analysis, a theoretical model regarding to the distortion parameters and material stiffness is established. Based on this, a database including material geometric parameters, elastic modulus, and shear modulus is constructed. A deep neural network (DNN) model with geometric parameters as input and Young’s modulus and shear modulus as output is established, and multi-objective optimization is carried out. The results show that the introduction of lattice distortion can increase Young’s modulus and shear modulus by up to 60.7% and 7%, respectively. Then, inverse design is carried out for the large deformation behavior of HEA lattice materials. Based on the database constructed by the finite element simulations, a hierarchical neural network is proposed via the K-means method and convolutional neural network (CNN) model, which can effectively generate the lattice material with the target performance. This inverse design method paves a new pathway for the effective and efficient design of lattice materials with improved mechanical properties.