APPENDIX C (continued)
Although this method has not gone through rigorous testing to date, it may
provide an alternative to wind tunnel testing for small scale buildings.
Model Requirements. The model, supported by a plywood base the
same size as the building's footprint, is mounted on a threaded flange fitted
on a threaded pipe. It can then be freely rotated for various wind
incidences. A wind vane is attached above the model to indicate relative wind
direction. The model can be built out of plexiglass for ease in viewing
during testing, although solar heating must be avoided. Aluminum foil on the
roof is recommended for this. A scale of 1 : 24 is recommended. The model
must be mounted and tested so that the height of the windows in the model is
the same as the height of the windows of the proposed building at full scale.
Applicable Buildings. There is no simulation of the ground plane
nor any match of the approach flow roughness length to the model height as in
wind tunnel testing. Instead, the model encounters a uniform vertical
velocity gradient with turbulent flow. The fluctuations of velocity and flow
direction give a useful if qualitative assessment of the ventilation in the
building. This type of testing is limited to residential scale buildings and
is not recommended for buildings taller than one story. In taller buildings
this type of testing will overestimate the effects of surrounding objects
since the surroundings are not matched to model height.
Types of Tests.
There are several possible approaches:
a) Smoke can be introduced into the model for flow visualization
(as is commonly done in wind tunnel testing).
b) A laser can be used with a glass rod to produce a planar light
source. The flow of smoke along a single plane may then be observed and
recorded with a low light level video camera. This must be done at night.
c) An omnidirectional temperature-compensated thermistor-type
airspeed probe can be used to measure interior velocities. Cloudy-sky
conditions are recommended to minimize radiation errors.
For examples of these, refer to Chandra et al. (1983), pp. 45-53.