Effect of Constituent Materials on Composite Performance: Exploring Design Strategies via Machine Learning

被引:56
作者
Chen, Chun-Teh [1 ]
Gu, Grace X. [2 ]
机构
[1] Univ Calif Berkeley, Dept Mat Sci & Engn, Berkeley, CA 94720 USA
[2] Univ Calif Berkeley, Dept Mech Engn, Berkeley, CA 94720 USA
基金
美国国家科学基金会;
关键词
composites; finite elements; graphene; machine learning; molecular dynamics; MECHANICAL-PROPERTIES; GRAPHENE SHEETS;
D O I
10.1002/adts.201900056
中图分类号
O [数理科学和化学]; P [天文学、地球科学]; Q [生物科学]; N [自然科学总论];
学科分类号
07 ; 0710 ; 09 ;
摘要
Nature assembles a range of biological composites with remarkable mechanical properties despite being composed of relatively weak polymeric and ceramic components. However, the architectures of biomaterials cannot be considered as optimal designs for engineering applications since biomaterials are constantly evolving for multiple functions beyond carrying external loading. Here, it is aimed to develop an intelligent approach to design superior composites from scratch starting from constituent materials. A systematic computational investigation of the effect of constituent materials assumed to be perfectly brittle) on the behavior of composites using an integrated approach combining finite element method, molecular dynamics, and machine learning (ML) is reported. It is demonstrated that instead of using brute-force methods, machine learning is a much more efficient approach and can generate optimal designs with similar performance to those obtained from an exhaustive search. Furthermore, it is shown that the toughening and strengthening mechanism observed in composites at the continuum-scale by combining stiff and soft constituents is valid for nanomaterials as well. Results show that high-performing designs of graphene nanocomposites can be generated using our ML approach. This novel ML-based design framework can be applied to other material systems to study a variety of structure-property relationships over several length-scales.
引用
收藏
页数:12
相关论文
共 49 条
[1]  
Abadi M., 2016, 12 USENIX S OPERATIN, P265
[2]  
[Anonymous], SCI REP
[3]  
[Anonymous], 2011, SOLID FREEFORM FABR
[4]  
Bendsoe M. P., 2004, Topology optimization: theory, methods, and applications
[5]   Nacre-inspired design of graphene oxide-polydopamine nanocomposites for enhanced mechanical properties and multi-functionalities [J].
Chen, Chun-Teh ;
Martin-Martinez, Francisco J. ;
Ling, Shengjie ;
Qin, Zhao ;
Buehler, Markus J. .
NANO FUTURES, 2017, 1 (01)
[6]   Molecular mechanics of elastic and bendable caffeine co-crystals [J].
Chen, Chun-Teh ;
Ghosh, Soumyajit ;
Reddy, C. Malla ;
Buehler, Markus J. .
PHYSICAL CHEMISTRY CHEMICAL PHYSICS, 2014, 16 (26) :13165-13171
[7]   Biological materials: Functional adaptations and bioinspired designs [J].
Chen, Po-Yu ;
McKittrick, Joanna ;
Meyers, Marc Andre .
PROGRESS IN MATERIALS SCIENCE, 2012, 57 (08) :1492-1704
[8]   3D-Printing of Lightweight Cellular Composites [J].
Compton, Brett G. ;
Lewis, Jennifer A. .
ADVANCED MATERIALS, 2014, 26 (34) :5930-+
[9]   Metallic materials for 3D printing [J].
Das, Suman ;
Bourell, David L. ;
Babu, S. S. .
MRS BULLETIN, 2016, 41 (10) :729-741
[10]   Artful interfaces within biological materials [J].
Dunlop, John W. C. ;
Weinkamer, Richard ;
Fratzl, Peter .
MATERIALS TODAY, 2011, 14 (03) :70-78