Civil Engineering Reference
In-Depth Information
(Figure 7.3). Detailed measurements of mean velocity and turbulence intensity profiles at
various spanwise stations and of cross-correlations and spectra were made.
Figure 7.3
The Counihan method for short test
sections.
7.3.3
Simulation of the surface layer
For simulation of wind forces and other wind effects on low-rise buildings, say less than
10 m in height, geometric scaling ratios of 1/400 result in extremely small models and do
not allow any details on the building to be reproduced. The large differences in Reynolds
numbers between model and full scale may mean that the wind-tunnel test data is quite
unreliable. For this type of structure, no attempt should be made to model the complete
atmospheric boundary layer. Simulation of the inner or surface layer, which is
approximately 100 m thick in full scale, is sufficient for such tests. If this is done, larger
and more practical scaling ratios in the range of 1/50-1/200 can be used for the models.
Cook (1973) developed a method for simulation of the lower third of the atmospheric
boundary layer. This system consists of a castellated barrier, a mixing grid and surface
roughness. A simpler system consisting of a plain barrier, or wall, at the start of the test
section followed by several metres of uniform surface roughness has also been used
(Figure 7.4) (Holmes and Osonphasop, 1983). This system has the advantage that
simultaneous control of the longitudinal turbulence intensity and the longitudinal length
scale of turbulence, to match the model scaling ratio, is obtained by adjustment of the
height of the barrier. Larger scales of turbulence can be produced by this method than by
other approaches—large horizontal vortices with their axes normal to the flow are
generated in the wake of the barrier. Studies of the development of the flow in the wake
of the barrier (Holmes and Osonphasop, 1983) showed that a fetch length of at least 30
times the barrier height is required to obtain a stable and monotonically increasing mean
velocity profile. However, there is still a residual peak in the shear stress profile at the
height of the barrier at this downstream position; this shows that the flow is still
developing at the measurement position, but the effect of this on pressures on and flow
around single buildings should not be significant.
7.3.4
Simulation of tropical cyclone and thunderstorm winds
As discussed in Chapter 1, strong winds produced by tropical cyclones and thunderstorms
dominate the populations of extreme winds in most locations with latitudes less than 40°,
including many sites in the United States, Australia, India and South Africa.
