Physics-informed machine learning-driven hierarchical structures for thin-film cooling

Abstract

Efficient cooling structures play a critical role in the advancement of power generation, electronics/photonics, and even battery mo,dules for electric vehicles. In addition, high-performance cooling structures will contribute significantly toward the carbon-neutra,l mission of the world in the next few decades. Extensive experimental efforts are conducted on the fabrication of cooling structure,s in the last several decades. the vast domain of available materials, dimensions, and morphologies make it impossible to explore al,l the variations through experimental efforts and to find optimal structures for high heat fluxes. Furthermore, substantial chemical, consumptions and time-consuming and costly fabrication and test procedures add more hurdles to these efforts.In this program, we pl,an to address this long-standing challenge by providing a new platform to discover high-performance micro/nano/molecular structures,for thin-film evaporation and to understand the physics behind these high-performance structures. Through a comprehensive physic-inf,ormed machine-learning (ML) platform, we aim to predict heat transfer characteristics of a material structure before fabrication and, experimentation. The governing variables on heat flux including geometrical dimensions of the material structure and properties of,the working fluid will be determined. Through optimization, we will find the optimal material structures for each working fluid. The,se structures will be fabricated and experimentally tested to validate this new platform. Furthermore, this platform enables us to r,eveal new fundamentals of liquid-vapor phase change at nano/molecular scales including the role of interface curvature on mass flux,, wetting characteristic and momentum transport in confined geometries.

Document Details

Document Type
DoD Grant Award
Publication Date
Dec 06, 2022
Source ID
N000142312034

Entities

People

  • Hadi Ghasemi

Organizations

  • Office of Naval Research
  • United States Navy
  • University of Houston System

Tags

Fields of Study

  • Engineering

Readers

  • Combustion and Flow Dynamics.
  • Distributed Systems and Data Platform Development
  • Nanofabrication and Microfabrication.

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks
  • Microelectronics