Intelligent Algorithm-Driven Optimization of Water-Cooled Plate Structures for Enhanced Thermal Performance
Year: 2025
Type: article
Abstract: The study presents a systematic approach to optimizing heat sink performance in high-heat flux applications through topology optimization (TO). A computational framework was developed that combines computational fluid dynamics (CFD) simulations with a simplified two-dimensional thermo-fluidic model to reduce computational complexity while maintaining accuracy. The design domain was constructed to minimize pressure drop under specific thermal constraints, with material properties interpolated using a rational approximation of material properties (RAMP) method to ensure a smooth transition between fluid and solid regions during optimization. Validation through three-dimensional numerical simulations in ANSYS Fluent confirmed the reliability of the two-dimensional model, with turbulence modeling and mesh refinement ensuring high accuracy in capturing critical flow and thermal characteristics. The results indicate that the topology-optimized designs achieved significant improvements over conventional straight-channel heat sinks, including a 25% reduction in thermal resistance and up to a 30% increase in heat transfer efficiency under varying flow rates. Moreover, the study demonstrates the feasibility of integrating artificial intelligence algorithms to streamline design optimization processes and enhance adaptability to complex performance requirements. These findings offer valuable insights for advancing heat management solutions in high-performance electronics and related applications.
DOI: 10.71222/2xatmz14
Authors: Ethan L. Norwick, Adrian T. Caldwell, Vincent M. Roemer, Felix J. Hargrove
Language: English
Subfield: Civil and Structural Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4408642097