Multi-objective NSGA-II based shape optimisation of the cross-sectional shape of passively cooled heat sinks

Date published

2021-07-16

Free to read from

Supervisor/s

Journal Title

Journal ISSN

Volume Title

Publisher

Emerald

Department

Type

Article

ISSN

0961-5539

Format

Citation

Santhanakrishnan MS, Tilford T, Bailey C. (2022) Multi-objective NSGA-II based shape optimisation of the cross-sectional shape of passively cooled heat sinks. International Journal of Numerical Methods for Heat and Fluid Flow, Volume 32, Issue 3, July 2022, pp. 1025-1045

Abstract

Purpose The purpose of the study is to optimise the cross-sectional shape of passively cooled horizontally mounted pin-fin heat sink for higher cooling performance and lower material usage.

Design/methodology/approach Multi-objective shape optimisation technique is used to design the heat sink fins. Non-dominated sorting genetic algorithm (NSGA-II) is combined with a geometric module to develop the shape optimiser. High-fidelity computational fluid dynamics (CFD) is used to evaluate the design objectives. Separate optimisations are carried out to design the shape of bottom row fins and middle row fins of a pin-fin heat sink. Finally, a computational validation was conducted by generating a three-dimensional pin-fin heat sink using optimised fin cross sections and comparing its performance against the circular pin-fin heat sink with the same inter-fin spacing value.

Findings Heat sink with optimised fin cross sections has 1.6% higher cooling effectiveness than circular pin-fin heat sink of same material volume, and has 10.3% higher cooling effectiveness than the pin-fin heat sink of same characteristics fin dimension. The special geometric features of optimised fins that resulted in superior performance are highlighted. Further, Pareto-optimal fronts for this multi-objective optimisation problem are obtained for different fin design scenarios.

Originality/value For the first time, passively cooled heat sink’s cross-sectional shapes are optimised for different spatial arrangements, using NSGA-II-based shape optimiser, which makes use of CFD solver to evaluate the design objectives. The optimised, high-performance shapes will find direct application to cool power electronic equipment.

Description

Software Description

Software Language

Github

Keywords

Natural convection, Shape optimisation, Genetic Algorithm, Heat sink design, CFD

DOI

Rights

Attribution-NonCommercial 4.0 International

Relationships

Relationships

Supplements

Funder/s