Functional estimation of the junction temperature with Monte Carlo method for electronic module cooling - IMT Mines Albi-Carmaux Access content directly
Preprints, Working Papers, ... Year : 2024

Functional estimation of the junction temperature with Monte Carlo method for electronic module cooling

Abstract

In power electronics modules, accurately estimating the junction temperature is crucial for ensuring proper cooling and preventing device failure. Recent breakthroughs on the integral formulation of thermal couplings allows the estimation of physical quantities as transfer functions with Monte Carlo algorithms. The transfer function is estimated with a single Monte Carlo algorithm in which random walks are sampled to solve the coupled thermal problem involving conduction and convection in potential complex geometries. In the context of estimating junction temperature in power electronics modules, this involves simulating the thermal behavior of the module under various operating conditions, such as different levels of power dissipation and ambient temperatures. This paper addresses the junction temperature estimation as a function depending on a decisive parameter for the electronics power components cooling: the convective heat transfer coefficient. The different steps leading to the estimation of the transfer function are presented as well as its numerical validation against a results from a reference multiphysics electronics cooling simulation software (Celsius EC Solver). The results from the transfer functions are exposed for typical configurations of power modules and in both situations, the functions demonstrates the accuracy and reliability of the approach without degrading the geometrical or multiphysics complexities of the problem. Moreover, due to the computational time gain when estimating the junction temperature for different cooling configurations, the function obtained with Monte Carlo is prove being a relevant candidate for assessing thermal peformance of power electronics modules, accounting for uncertainties and variations inherent in operating conditions.
Embargoed file
Embargoed file
0 5 5
Year Month Jours
Avant la publication
Friday, August 9, 2024
Embargoed file
Friday, August 9, 2024
Please log in to request access to the document

Dates and versions

hal-04448075 , version 1 (09-02-2024)

Identifiers

  • HAL Id : hal-04448075 , version 1

Cite

L Éa Penazzi, Anne Castelan, Stéphane Blanco, Cyril Caliot, Christophe Coustet, et al.. Functional estimation of the junction temperature with Monte Carlo method for electronic module cooling. 2024. ⟨hal-04448075⟩
39 View
1 Download

Share

Gmail Facebook X LinkedIn More