Coupled Thermo-Fluid Models for Heat Transfer Optimization in Microelectronic Cooling Systems

Abstract

The ever-increasing miniaturization and scaling of performance of micro-electronic devices have resulted in unprecedented levels of heat flux, requiring innovative and dependable thermal management solutions. The traditional single-domain thermal models lack the ability to explain the intense interaction between fluid convection and solid conduction, and hence is not predictive of the next-generation cooling systems. This paper constructs a unified thermo-fluid modeling platform incorporating incompressible Navier Stokes equations, the transient heat conduction and heat transfer equations to represent the multi-physics interactions between solid and fluid interfaces. The discretized coupled system is solved with a hybrid finite element / finite volume method that allows both the complex geometry and nonlinear interactions between thermal and fluid forces to be effectively addressed. The framework is used on representative representative microelectronic cooling systems, such as, microchannel heat sinks, nanofluid-imbued channels, and phase-change-aided hybrid systems. Multi-objective optimization scheme is used to reduce maximum device temperature, thermal resistance and temperature non-uniformity and balance hydraulic performance. The simulation findings indicate that coupled modeling is more effective in the prediction of temperatures by greater than 12 percent relative to decoupled modeling models, whereas optimized designs will lead to a decrease in peak temperature of up to 15 percent and temperature evenness of as much as 22 percent. Moreover, phase-change-aided cooling is shown to be better at transient thermal control with time-varying loads, and nanofluid-enhanced channels provide a higher steady-state heat transfer rate. The suggested hybrid framework does not only promote mathematical modeling of thermo-fluid systems, but offers useful design guidelines to optimized cooling structures in high-power-density microelectronics, which is within the journal focus on mathematical modeling, computational methods and engineering applications.