Machine learning-enabled constrained multi-objective design of architected materials
Abstract Architected materials that consist of multiple subelements arranged in particular orders can demonstrate a much broader range of properties than their constituent materials. However, the rational design of these materials generally relies on experts’ prior knowledge and requires painstaking...
Main Authors: | , , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Nature Portfolio
2023-10-01
|
Series: | Nature Communications |
Online Access: | https://doi.org/10.1038/s41467-023-42415-y |
_version_ | 1797558437393989632 |
---|---|
author | Bo Peng Ye Wei Yu Qin Jiabao Dai Yue Li Aobo Liu Yun Tian Liuliu Han Yufeng Zheng Peng Wen |
author_facet | Bo Peng Ye Wei Yu Qin Jiabao Dai Yue Li Aobo Liu Yun Tian Liuliu Han Yufeng Zheng Peng Wen |
author_sort | Bo Peng |
collection | DOAJ |
description | Abstract Architected materials that consist of multiple subelements arranged in particular orders can demonstrate a much broader range of properties than their constituent materials. However, the rational design of these materials generally relies on experts’ prior knowledge and requires painstaking effort. Here, we present a data-efficient method for the high-dimensional multi-property optimization of 3D-printed architected materials utilizing a machine learning (ML) cycle consisting of the finite element method (FEM) and 3D neural networks. Specifically, we apply our method to orthopedic implant design. Compared to uniform designs, our experience-free method designs microscale heterogeneous architectures with a biocompatible elastic modulus and higher strength. Furthermore, inspired by the knowledge learned from the neural networks, we develop machine-human synergy, adapting the ML-designed architecture to fix a macroscale, irregularly shaped animal bone defect. Such adaptation exhibits 20% higher experimental load-bearing capacity than the uniform design. Thus, our method provides a data-efficient paradigm for the fast and intelligent design of architected materials with tailored mechanical, physical, and chemical properties. |
first_indexed | 2024-03-10T17:31:25Z |
format | Article |
id | doaj.art-325d2539a52947b886fb6353a83f2801 |
institution | Directory Open Access Journal |
issn | 2041-1723 |
language | English |
last_indexed | 2024-03-10T17:31:25Z |
publishDate | 2023-10-01 |
publisher | Nature Portfolio |
record_format | Article |
series | Nature Communications |
spelling | doaj.art-325d2539a52947b886fb6353a83f28012023-11-20T10:00:13ZengNature PortfolioNature Communications2041-17232023-10-0114111210.1038/s41467-023-42415-yMachine learning-enabled constrained multi-objective design of architected materialsBo Peng0Ye Wei1Yu Qin2Jiabao Dai3Yue Li4Aobo Liu5Yun Tian6Liuliu Han7Yufeng Zheng8Peng Wen9State Key Laboratory of Tribology in Advanced Equipment, Tsinghua UniversityInstitute for Interdisciplinary Information Science, Tsinghua UniversityDepartment of Materials Science and Engineering, Peking UniversityState Key Laboratory of Tribology in Advanced Equipment, Tsinghua UniversityMax-Planck-Institut für EisenforschungState Key Laboratory of Tribology in Advanced Equipment, Tsinghua UniversityDepartment of Orthopaedics, Peking University Third HospitalMax-Planck-Institut für EisenforschungDepartment of Materials Science and Engineering, Peking UniversityState Key Laboratory of Tribology in Advanced Equipment, Tsinghua UniversityAbstract Architected materials that consist of multiple subelements arranged in particular orders can demonstrate a much broader range of properties than their constituent materials. However, the rational design of these materials generally relies on experts’ prior knowledge and requires painstaking effort. Here, we present a data-efficient method for the high-dimensional multi-property optimization of 3D-printed architected materials utilizing a machine learning (ML) cycle consisting of the finite element method (FEM) and 3D neural networks. Specifically, we apply our method to orthopedic implant design. Compared to uniform designs, our experience-free method designs microscale heterogeneous architectures with a biocompatible elastic modulus and higher strength. Furthermore, inspired by the knowledge learned from the neural networks, we develop machine-human synergy, adapting the ML-designed architecture to fix a macroscale, irregularly shaped animal bone defect. Such adaptation exhibits 20% higher experimental load-bearing capacity than the uniform design. Thus, our method provides a data-efficient paradigm for the fast and intelligent design of architected materials with tailored mechanical, physical, and chemical properties.https://doi.org/10.1038/s41467-023-42415-y |
spellingShingle | Bo Peng Ye Wei Yu Qin Jiabao Dai Yue Li Aobo Liu Yun Tian Liuliu Han Yufeng Zheng Peng Wen Machine learning-enabled constrained multi-objective design of architected materials Nature Communications |
title | Machine learning-enabled constrained multi-objective design of architected materials |
title_full | Machine learning-enabled constrained multi-objective design of architected materials |
title_fullStr | Machine learning-enabled constrained multi-objective design of architected materials |
title_full_unstemmed | Machine learning-enabled constrained multi-objective design of architected materials |
title_short | Machine learning-enabled constrained multi-objective design of architected materials |
title_sort | machine learning enabled constrained multi objective design of architected materials |
url | https://doi.org/10.1038/s41467-023-42415-y |
work_keys_str_mv | AT bopeng machinelearningenabledconstrainedmultiobjectivedesignofarchitectedmaterials AT yewei machinelearningenabledconstrainedmultiobjectivedesignofarchitectedmaterials AT yuqin machinelearningenabledconstrainedmultiobjectivedesignofarchitectedmaterials AT jiabaodai machinelearningenabledconstrainedmultiobjectivedesignofarchitectedmaterials AT yueli machinelearningenabledconstrainedmultiobjectivedesignofarchitectedmaterials AT aoboliu machinelearningenabledconstrainedmultiobjectivedesignofarchitectedmaterials AT yuntian machinelearningenabledconstrainedmultiobjectivedesignofarchitectedmaterials AT liuliuhan machinelearningenabledconstrainedmultiobjectivedesignofarchitectedmaterials AT yufengzheng machinelearningenabledconstrainedmultiobjectivedesignofarchitectedmaterials AT pengwen machinelearningenabledconstrainedmultiobjectivedesignofarchitectedmaterials |