Spray cooling heat transfer — Test and CFD analysis

Spray cooling of high temperature surfaces subject to large internal heat generation is analyzed by computational fluid dynamics (CFD) methods to determine heat transfer coefficients and the micro-physical details of coolant droplet-heated surface interactions governed by evaporative processes. A high speed, high magnification digital camera (6000 frames/sec) is used to provide test data for micron scale spray droplet size distribution and droplet velocity from a spray nozzle for different supply pressures for HFE 7100 and water coolants. Droplet test data are then applied to construct FLOW-3D CFD models [1] of numerous translating spherical droplets impacting a heated surface with internal volume heat generation and the transient, free-surface fluid dynamics and heat transfer processes computed. Transient, expanding/collapsing, chaotic coolant vapor regions generated by evaporative processes during successive multiple droplet impacts on flat and roughened surfaces sustaining large heat fluxes (from 30 to 300 W/cm2) are generated from the CFD solutions and shown to reproduce qualitative phase transition features observed from test photography. A computer program is provided to calculate heat transfer coefficients for different combinations of coolant droplet size, droplet velocity, droplet spatial distribution in nozzle sprays, heat flux magnitude, evaporation temperature and coolant flow rate incorporating the thermophysical coolant and wall properties for both flat and surface roughness cases. CFD results for a wide variety of droplet sizes, translation velocities, magnitudes of heat flux for flat and surface roughness patterns, coolant flow rates, coolant types and prescribed wall surface temperatures are used to provide physical insights into best ways to achieve maximum spray cooling heat transfer coefficients and avoid surface flooding and dry spotting. Use of high speed photographic micro-details of droplet impingement and evaporation structures on heated walls is made to qualitatively substantiate the CFD methodology by comparison of computed to test observations.