Strength through defects: A novel Bayesian approach for the optimization of architected materials
Abstract
We use a previously unexplored Bayesian optimization framework, "evolutionary Monte Carlo sampling," to systematically design the arrangement of defects in an architected microlattice to maximize its strain energy density before undergoing catastrophic failure. Our algorithm searches a design space with billions of 4 × 4 × 5 3D lattices, yet it finds the global optimum with only 250 cost function evaluations. Our optimum has a normalized strain energy density 12,464 times greater than its commonly studied defect-free counterpart. Traditional optimization is inefficient for this microlattice because (i) the design space has discrete, qualitative parameter states as input variables, (ii) the cost function is computationally expensive, and (iii) the design space is large. Our proposed framework is useful for architected materials and for many optimization problems in science and elucidates how defects can enhance the mechanical performance of architected materials.
Additional Information
© 2021 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. Distributed under a Creative Commons Attribution NonCommercial License 4.0 (CC BY-NC). Submitted 2 July 2021; Accepted 17 August 2021; Published 8 October 2021. We thank K. Komvopoulos (Department of Mechanical Engineering) and U. Seljak [Department of Physics, University of California at Berkeley (UCB)] for insightful discussions, P. Hosemann [Department of Nuclear Engineering (UCB)] for the use of the nanoindentation apparatus, F. Allen [Department of Materials Science and Engineering (UCB)] for training to use the helium ion microscope, and S. Govindjee and F. Armero [Department of Civil and Environmental Engineering (UCB)] and D. J. Steigmann [Department of Mechanical Engineering (UCB)] for fruitful discussions regarding the mechanical performance of the structures. This work is suppoted by the Extreme Science and Engineering Discovery Environment (XSEDE), supported by NSF grant number ACI-1548562 through allocation TG-CTS190047 (H.M.S. and P.S.M.); NSF Scalable Nanomanufacturing Program, award no. 1449305 (C.P.G. and Z.V.); The California Institute of Quantitative Bioscience, QB3 Lab (C.P.G. and Z.V.); and Lotusland Investment Holdings Inc. and Bohn Valley Inc. (financial gift; H.M.S. and P.S.M.). Authors contributions: Conceptualization: Z.V. and H.M.S. Methodology: H.M.S., Z.V., P.S.M., and C.P.G. Investigation: Z.V., H.M.S., V.Z.L., and G.F. Visualization: Z.V., H.M.S., P.S.M., and C.P.G. Supervision: C.P.G., P.S.M., and M.F. Writing (original draft): Z.V., H.M.S., and P.S.M. Writing (review and editing): Z.V., H.M.S., P.S.M., and C.P.G. The authors declare that they have no competing interests. Data and materials availability: The data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials.Attached Files
Published - sciadv.abk2218.pdf
Supplemental Material - sciadv.abk2218_movies_s1_to_s4.zip
Supplemental Material - sciadv.abk2218_sm.pdf
Files
Additional details
- PMCID
- PMC8500519
- Eprint ID
- 111353
- Resolver ID
- CaltechAUTHORS:20211011-175305271
- ACI-1548562
- NSF
- TG-CTS190047
- NSF
- EEC-1449305
- NSF
- California Institute of Quantitative Bioscience
- Lotusland Investment Holdings Inc.
- Bohn Valley Inc.
- Created
-
2021-10-11Created from EPrint's datestamp field
- Updated
-
2021-10-18Created from EPrint's last_modified field