BibTex format
@article{Thillaithevan:2025:10.1007/s00158-025-04088-7,
author = {Thillaithevan, D and Hewson, R and Murphy, R and Santer, M and Calver, A and Nikiteas, G and Raske, N},
doi = {10.1007/s00158-025-04088-7},
journal = {Structural and Multidisciplinary Optimization: computer-aided optimal design of stressed solids and multidisciplinary systems},
title = {A subspace method for 3D multiscale heat sink modelling and optimization},
url = {http://dx.doi.org/10.1007/s00158-025-04088-7},
volume = {68},
year = {2025}
}
RIS format (EndNote, RefMan)
TY - JOUR
AB - The increasing computational demands of modern microprocessors require efficient thermal man-agement solutions. To address this design challenge, a novel 3D multiscale heat sink modelling andoptimization framework is presented. The approach combines a multiscale momentum model withan iterative temperature-flux projection scheme that accurately resolves heat transport across theentire macroscale domain without relying on traditional homogenization methods. Unlike conven-tional fixed-grid topology optimization approaches, the framework maintains solution accuracy nearsolid/fluid interfaces while achieving superior computational efficiency, reducing memory requirementsand computation time relative to equivalent explicit single scale simulations. Bayesian optimizationis utilised to demonstrate the framework’s practical utility by designing multiscale heat sinks withhundreds of unit cells subject to homogeneous, and more accurate in-homogeneous surface heat flux,achieving significantly larger heat transfer in both cases, while maintaining pressure drop constraints.This framework enables the practical optimization of complex 3D heat sink designs at multiscaleresolutions previously intractable with traditional explicit modelling approaches.
AU - Thillaithevan,D
AU - Hewson,R
AU - Murphy,R
AU - Santer,M
AU - Calver,A
AU - Nikiteas,G
AU - Raske,N
DO - 10.1007/s00158-025-04088-7
PY - 2025///
SN - 1615-147X
TI - A subspace method for 3D multiscale heat sink modelling and optimization
T2 - Structural and Multidisciplinary Optimization: computer-aided optimal design of stressed solids and multidisciplinary systems
UR - http://dx.doi.org/10.1007/s00158-025-04088-7
VL - 68
ER -