With continued development of the electronic industry, the demand for highly efficient heat removal solutions requires innovative cooling technologies. A computational fluid dynamic (CFD) study, including heat transfer, is performed for an axisymmetric, confined jet impingement experiencing boiling and coupled with vapor extraction. Boiling occurs at the target surface while extraction occurs at the wall confining the radial flow. The region between the target and confining wall is defined as a confined gap. Extraction is employed to enhance heat transfer and to minimize the potential negative influence of flow instabilities resulting from two-phase flow within a confined region.
A three-dimensional sector of the confined jet is employed in the simulation. A single circular impinging jet with a constant jet diameter (4 mm) and variable gap height (0.5, 1.0 and 1.5 mm), also known as nozzle-to-target spacing, is considered. The effect of mass flux at the confined gap entrance is also investigated (200, 400 and 800 kg/m²-s) for a range of heat flux (5 to 50 W/cm²).
Fluid flow and heat transfer are simulated using the Volume of Fluid (VOF) model and the wall-boiling sub-model within the Multiphase Segregated Flow (MSF) model. The boiling sub-model in the VOF model applies the Rohsenow boiling correlation, while in the MSF model, the Kurul-Podowski boiling sub-model is used. Also, vapor extraction is realized by different mechanisms for these two models. For the VOF model, a specific phase "wall porosity" can be assigned to a wall to make it porous. Over a range of pressure differentials across this porous wall such that the inertial transport influence is negligible, vapor transport should agree with Darcy's law. For the MSF model, a wall can be made permeability to one substance or phase while remaining impermeable to the other substance or phase. However, a portion of the substance or phase reaching the boundary allowed to pass through the surface must be specified. A pressure drop cannot be applied across the wall, thereby prohibiting Darcy flow modeling. The solutions of both models are at steady state.
The boiling curves without vapor extraction from both models are provided and compared to experiments. Simulations matching experimental wall temperatures under-predict theoretical vapor generation and those matching vapor generation over-estimate wall superheat. For cases with no extraction, local temperature and velocity profiles from the VOF model are provided at several radial locations within the confined gap. Scalar temperature and pressure distributions and velocity vectors are presented to explain observations in profiles.