Numerical investigation of GaN MMIC PA thermal management system and multi-objective genetic algorithm optimization of heat sink parameters

•Instead of simplifying the device as a single heat source applied directly to the heat sink, a comprehensive GaN MMIC PA model was developed.•The heat generated by the finger gates of HEMTs on the device is simplified as a surface heat source.•Thermal resistance analysis includes the influence of t...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Applied thermal engineering Ročník 272; s. 126446
Hlavní autori: Li, Yunhao, Zheng, Jinquan, Kong, Lili, Chang, Long, Kong, Lingyue, Kou, Guiyue, Sun, Jiang, Mu, Mingfei
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Elsevier Ltd 01.08.2025
Predmet:
GaN devices
Genetic algorithm
Heat sinks
Multi-objective optimization
Thermal management
Thermal resistance
GaN devices
Thermal management
Multi-objective optimization
Thermal resistance
Genetic algorithm
Heat sinks
ISSN:1359-4311
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:•Instead of simplifying the device as a single heat source applied directly to the heat sink, a comprehensive GaN MMIC PA model was developed.•The heat generated by the finger gates of HEMTs on the device is simplified as a surface heat source.•Thermal resistance analysis includes the influence of thermal boundary resistance at the GaN-SiC interface on heat transfer process.•Fitting equations for Nu and f of Si heat sink were proposed.•The NSGA-II multi-objective optimization method was applied to the thermal management system of GaN MMIC PA. The local hot spots generated in the gate region will significantly affect the performance of GaN monolithic microwave integrated circuit power amplifier (GaN MMIC PA). Reasonable thermal management methods are significant for reducing the hot spot temperature. The thermal management system of GaN MMIC PA with heat flux greater than 500 W/cm2 and local hot spot heat flux greater than 30 kW/cm2 is numerically analyzed, and the parameters of its heat sink are optimized by NASA-II multi-objective optimization algorithm. The sandwich structure of five layers consisting of GaN, GaN-SiC interface, SiC, AuSn and Si heat sink was constructed, and rectangular microchannels were arranged directly below the hot spots. By parametric design, 65 rectangular microchannels were defined with aspect ratio (α) and length ratio (β), and the influence of parameters α, β, Reynolds number (Re), and RTBR on the heat transfer process were analyzed. The final optimization takes the Nusselt number (Nu) and the friction factor (f) as the objective functions. The optimization results obtain the pareto optimal solution, which can reduce the f by 11.30 % under the condition that the Nu remains unchanged, or increase the Nu by 7.41 % under the condition that the f remains unchanged compared to two simulation data. Three sets of pareto optimal solutions (cases 1, 2, 3) selected according to the demand of the working conditions are verified by simulation, which shows that the average mean absolute percentage error (MAPE) between the predicted values of Nu and the simulation values is 2.14 %, while that for f is 1.22 %. Additionally, a fitting equation for the pareto optimal solutions was developed, achieving R2 of 0.998, providing a useful reference for selecting heat sink parameters in GaN MMIC PA thermal management systems.
ISSN:1359-4311
DOI:10.1016/j.applthermaleng.2025.126446