This experiment, carried out at Nokia Research Center in Helsinki,
was designed to provide well-defined model for CFD simulations,
thereby eliminating eliminate potential uncertainties. The fin array
was isothermally powered to within 1°C by electrical cartridge heaters.
Temperature was monitored by a combination of PT 100 platinum resistance
sensors (DIN IEC 751) embedded in the cartridge heaters and thermocouples
attached to the baseplate. Temperature and power dissipation were
measured to within ± 0.2 °C and 1% respectively. To permit heat
transfer rate calculations, a double-sided symmetry design was employed.
The enclosure walls were water cooled to ensure isothermal surfaces
and permit definition of fin array adiabatic wall temperature. Fluid
velocity and temperature profiles were measured at the upper and
lower opening of a central u-channel passage using hot wire anemometry,
and salient features of enclosure fluid flow were observed using
smoke flow visualization.
Coolit simulation was run blind, i.e. without the knowledge of
experimental data. The flow pattern computed by Coolit and the fin
array (1/4th) are shown in the adjacent figure. A plot of heat flux
from the heat sink predicted by Coolit shows excellent agreement
with experimental data.