Civil Engineering Reference
In-Depth Information
Unfortunately, full-scale measurements of such events are few in number, and there are
no reliable analytical models for the surface wind structures in these storms. However,
the few full-scale measurements, and some meso-scale numerical models, have enabled
qualitative characteristics of the winds to be determined.
Figure 7.4
The barrier-roughness technique.
Tropical cyclones, known also as 'hurricanes' and 'typhoons' in some parts of the world,
are circulating systems with a complex three-dimensional wind structure near their centre
(Section 1.3.2). At the outer radii, where the wind speeds are lower, a boundary-layer
structure should exist and conventional boundary-layer wind tunnels should be quite
adequate for flow modelling. However, the region of maximum horizontal winds occurs
just outside the eye wall. Here the winds near the surface turn towards the low-pressure
centre and in a spiralling upward direction at greater heights. Measurements have
indicated a steeper mean velocity profile than would be expected for gales, for the surface
roughness conditions around the site, up to the height of about 100 m. Above that height,
the mean wind velocity is approximately constant up to the top of the tower (Section
3.2.5). Measurements of turbulence intensities in typhoons have shown higher values
than occurring at the same site in non-cyclonic conditions (Section 3.3.1). As most
structures do not exceed 100 m in height, a reasonable approximation to the tropical
cyclone flow can be obtained by using a boundary-layer flow generated for urban terrain
conditions, even for directions with lower roughness lengths, such as off-water winds for
coastal sites.
The laboratory modelling of thunderstorm winds is a more difficult problem for a
number of reasons. First, there are a number of different types of local wind storms
associated with thunderstorms, although some of these have similar characteristics.
Second, these storms are individually transient, although a number of them may occur
sequentially in the same day. The length of an individual storm rarely exceeds 30min.
Third, thunderstorm winds are driven by thermodynamic processes which probably
cannot be reproduced in a laboratory simulation.
The velocity profile in a thunderstorm downdraft is quite similar to a wall jet. The
latter has been proposed as a laboratory model of the flow in a downdraft, and some
studies have been conducted using the outlet jet from a wind tunnel impinging on a
vertical board, as shown in Figure 7.5. Measurements can be carried out at various radial
positions from the centre of the board. This system gives velocity profiles which are quite
similar to those measured by radar in microbursts, but the transient characteristics of the
real downdraft flow are not reproduced and the turbulence characteristics in the two
flows could be quite different.
