Machine Learning Approach to Predict the Effect of Metal Foam Heat Sinks Discretely Placed in a Cavity on Surface Temperature

Aug 5, 2026

Upload by Liam

Metal foam heat sinks are widely used in electronic cooling due to strong heat-transfer performance, low weight, and the ability to promote coolant mixing. However, extensive experiments are costly and difficult to perform, and the complex foam structure can complicate numerical simulation setup. This study applies multiple machine learning methods to predict mean surface temperature for metal foams discretely placed in partially open cavities, using pore density, Reynolds number, modified Grashof number, and distance to aperture as inputs.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages11

Reading Time28 min

AccessFree to download

Abstract

Introduction

Background and motivation for cooling improvements

Prior research on metal foam heat sinks

Machine learning data and modeling approach

Input parameters and problem setup

Model evaluation and ranking

ISSN 1330-3651 (Print), ISSN 1848-6339 (Online) [https://doi.org/10.17559/TV-20240302001366](https://doi.org/10.17559/TV-20240302001366) Original scientific paper Machine Learning Approach to Predict the Effect of Metal Foam Heat Sinks Discretely Placed in a Cavity on Surface Temperature Oğuzhan ÖZBALCI, Mustafa ÇAKIR, Okan ORAL*, Ayla DOĞAN Abstract: Metal foam heat sinks are preferred in electronic cooling systems with their advantages such as superior properties in heat transfer, light weight and ability to mix the cooling fluid. It is very difficult to conduct extensive experimental studies with metal foam heat sinks due to the difficulty of production and high cost. In addition, due to the complex structure of metal foam heat sinks, difficulties may arise in the creation of numerical simulations. In the present study, various machine learning methods were used, taking into account the mean surface temperature values obtained by using metal foam heat sinks discretely placed in a partially open volume. The pore density of metal foam heat sink, Reynolds number, modified Grashof number and distance to aperture were taken as input parameters. When the results were examined, it was determined which of the inlet parameters were more effective on the mean surface temperature. It was determined that modified Grashof number was the most effective parameter on mean surface temperatures, but L was the weakest parameter. The models were ranked according to 3 different evaluation metrics. It was observed that the top three most successful machine learning algorithms were eXtreme gradient boosting, support vector machine and random forest. Keywords: artificial intelligence; electronic cooling; machine learning; metal foam heat sink; regression 1 INTRODUCTION The increasing performance of electronic systems with the developing technology causes the existing cooling systems to be insufficient and the surface temperatures to exceed the allowable limit values. This situation causes the performance of the electronic system to decrease and become unusable by being damaged. To eliminate this heating problem in electronic systems, researchers have turned to new cooling system designs and the use of innovative materials. Metal foam heat sinks, first produced by De Meller in 1925 and investigated in various fields by researchers, are very popular and innovative materials [1] . Metal foam heat sinks have advantages over existing materials due to their high area/volume ratio, flow mixing feature and light weight. In many different studies in the literature, heat transfer from metal foam heat sinks in which air was used as a refrigerant under free and forced convection conditions were investigated [2-6] . There were studies in which metal foams were examined as heat sinks with different number and size of open slots [7], and their effects on heat transfer and pressure drop were investigated by placing them discretely or along the channel [8, 9] . Studies were carried out with water [10-13], nanofluids or various refrigerants in the block or channel containing metal foam heat sinks [14-18] . Hsieh et al. [19] experimentally researched the effects of porosity, pore density and air velocity on the heat transfer characteristics of metal foam heat sinks. It was determined that increasing the pore density and the porosity increases the Nusselt number. Experimental investigation of the heat transfer characteristics of aluminium metal foam heat sinks with restricted flow outlet under impinging jet flow was conducted by Shih et al. [20] . It was found that the flow outlet height was more effective than the pore density, the porosity and aluminium heat sink height. Shih et al. [21] experimentally investigated its effect on heat transfer by placing an aluminium cylinder block in the center of a cylindrical aluminium metal foam heat sink. Depending on the increasing contact area (from 0 to 0.013), the Nusselt number first increased and then decr

FAQs

Why are metal foam heat sinks difficult to study experimentally and numerically?

Experimental work is challenging due to production difficulty and high cost. Numerically, the complex structure of metal foam heat sinks can make it difficult to construct reliable simulations.

Which parameters were used as inputs to the machine learning models?

The models used pore density, Reynolds number, modified Grashof number, and the distance to aperture as input parameters, with mean surface temperature as the target.

Which machine learning algorithms performed best and which parameter was most influential?

The top-performing algorithms were eXtreme gradient boosting, support vector machine, and random forest. The modified Grashof number was the most effective parameter for mean surface temperature, while L was the weakest.

Aug 5, 2026

Upload by Liam