Back to search

Cornell University

Multiscale Modeling and Machine Learning Design Optimization of Nanomaterials for Sustainability

Abstract

dc:description.abstract

The development of machine learning (ML) techniques is one of the biggest breakthroughs in the 21st century, which may revolutionize many diverse industries. In this thesis, I apply ML and design optimization to solve key problems in multiscale materials modeling. First, I aim to alleviate the biofouling of surfaces by rapidly designing antimicrobial nanosurfaces. I proposed a framework coupling individual-based modeling and Bayesian optimization for digital topological optimization. My framework successfully optimized computer-modeled functional nanosurfaces that improved biofilm removal through applied shear and vibration. My results provide further insights into various engineering fields that require surface-mediated biofilm control. Second, I determined the multiscale mechanisms of graphene fracture under thermal gradients using non-equilibrium molecular dynamics (MD) simulations. I studied the fracture of graphene by applying a fixed strain rate under different thermal gradients while employing different interatomic potential fields for comparisons. The temperature gradients did not significantly influence fracture stresses and crack propagation dynamics. I used quantized fracture mechanics to verify these observations. Transverse bond forces shared the loading to account for the nonlinear increase of fracture stress with shorter crack length. The fractures were more ”brittle-liked” when the machine learning interatomic potential (MLIP) was used while being incapable of simulating post-fracture dynamical behaviors. Third, I benchmarked 12 different inverse optimization algorithms for molecular materials design. I proposed a Python platform that implemented these different optimization methods to design molecular materials given defined targeted properties. Under the preliminary stage of limited material evaluations, by analyzing the evaluated materials counted in a defined targeted design space bounded by properties of silicon, Runge Kutta Optimization is found to be outstanding for both single and multi-element de-sign optimizations. By estimating the mean value of the objectives it is found that Ant Colony Optimization, Simulated Annealing, and Runge Kutta Optimization outperformed the other algorithms for single-element optimizations, and Genetic Algorithm and Runge Kutta Optimization outperformed for multi-element optimizations. In summary, this thesis contributes results and deeper insights toward designing and understanding the complex behaviors of diverse engineering materials.

Degree

thesis:*
Name thesis:degree_name
M.S., Mechanical Engineering
Level thesis:degree_level
Master of Science
Discipline thesis:degree_discipline
Mechanical Engineering
Grantor
Cornell University
Year dc:date.issued
2023

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Zhai, Hanfeng
Chair dc:contributor.chair
  • Yeo, Jingjie
Committee member dc:contributor.committeemember
  • Warner, Derek

Subjects

dc:subject × 6

Rights

dc:rights
Statement dc:rights
  • Attribution 4.0 International
Language dc:language.iso
en

Identifiers

dc:identifier.*
Dc Identifier Bibid
16176698
Dc Identifier Other
ProQuest Submission ID: 11708
ProQuest Publication ID: 30426510

Chain of custody

source
Harvested from
Cornell University
Base URL
ecommons.cornell.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Zhai, Hanfeng. Multiscale Modeling and Machine Learning Design Optimization of Nanomaterials for Sustainability. Master of Science thesis, Cornell University, 2023. https://hdl.handle.net/1813/113974