Experimental setup

In this experiment, we want to infer temperature fields and heat transfer
from interferometric images.  For this purpose, a transparent wind tunnel
was designed. The experimental setup is presented schematically here.
The wind tunnel is positioned horizontally and consists of five main parts:
the intake and exit nozzles and the three tunnel sections. The channel
is H = 25 mm high and W = 200 mm wide.  The 549 mm long upstream section
allows the velocity profile to develop before reaching the test section.
The length of the upstream section is limited by the size of the optical
table and the dimensions of the optical arrangement.  The test section
is 508 mm long.  Ten equally spaced, electrically heated copper blocks
are attached to the bottom wall of the test section. The downstream
section is physically identical to the upstream section. The exit nozzle
converges to a hose with an inner diameter of 32 mm. The exit hose is
connected to a rotameter and a draw-through blower. A control valve
together with the bypass valve allow adjustments to the desired flow rate.

Channel geometry

The geometry of our experimental setup was chosen to allow direct
comparison with numerical studies of  Ghaddar et al. (1986a, 1986b)
and  Patera and Mikic (1986). The aspect ratios common to these studies
relate various groove streamwise and heightwise geometrical attributes
to the half-height of the main channel,  . The characteristic dimensions
in our experiment correspond to those introduced by Ghaddar et al.
(1986a, 1986b) and  Patera and Mikic (1986) as their base geometry, and
with respect to the main channel half-height of 0.008 m they are:
periodicity length, 6.66; groove depth, 1.11; and space between the 
consecutive blocks, 2.22. Using these aspect ratios and the initially
determined main channel half-height, the following dimensions for the
experimental apparatus were derived: periodicity length, L = 50 mm,
height of the heated blocks, a = 9 mm, and interblock spacing,   mm.
The spanwise dimension of the test channel, W = 200 mm, was selected
to yield approximately two-dimensional flow and temperature fields
(which is a requirement for accurate measurements by holographic
interferometry) as well as the desired number of fringes for a
prescribed temperature difference between the heated block and the
adiabatic wall.

Heated blocks

The structure of the electrically heated copper blocks is shown in
Figure 3.  As indicated, each block is equipped with two Type-T
thermocouples that measure surface temperatures; one close to the
upstream edge and the other close to the downstream edge. A resistive
electric heater foil is bonded to the groove manufactured on the bottom
surface of the block. Thus the heated block is in direct thermal contact
with the bottom plane wall only through the narrow region around the
groove.  The ninth block, which is visible in the interferometric image,
holds six thermocouples necessary to provide accurate reference values
for interferometric measurements. Constant power is delivered to the
heater foil by an AC power supply.