It is now becoming generally recognised that wind tunnel wodel tests are necessary on every new design of offshore gas/oil production platform to identify any problems likely to arise due to the wind environment. Such tests should be conducted at the earliest possible stage in the design procedure in order that any modifications to the design suggested by the tests can be incorporated with minimum impact on time schedules.
Increasing experience with such model tests has resulted in wind tunnel testing procedures which are becoming increasingly sophisticated and comprehensive. Environmental problems on platforms can now be predicted with a reasonable degree of assurance from the results of model tests correlated with operational experience on existing platforms. However, such assurance in the validity of the model tests assumes that those tests have been performed with scientific integrity. Unfortunately some aspects of these types of wind tunnel model tests are still commonly performed using dubious techniques.
The object of this paper is to explain the author's view of how certain facets of wind tunnel model testing should be approached in the light of current knowledge.
The following topics are discussed:
simulation of the turbulent shear flow over the sea generated by surface roughness.
modelling of gas turbine exhaust plumes
measurement of wind velocities over helicopter flight paths
measurement of wind pressures on HVAC inlets and outlets.
The comments in this paper are based on the extensive wind tunnel model tests of offshore gas/oil production platforms performed by the Environmental Aerodynamics Group, Cranfield Institute of Technology, England.
The large scale movements of the air in the Earth's boundary layer result from pressure gradients which arise principally from thermal effects caused by differential heating within the boundary layer. These pressure gradients may be defined in terms of the separation of, constant pressure lines or isobars shown on the familiar weather map. The air velocityis a function of the pressure gradient, the greater the pressure gradient the higher the wind speed. Near the surface of the Earth, drag due to surface roughness retards the lower layers o[ the air to induce shear and turbulence (or vorticity) into the wind flow.
The magnitude of the wind shear may be expressed in the form dU/dz, where U is the mean velocity and z the height above the surface. T4e turbulent component of wind flow is random in amplitude and direction. Being a random variable it can only be described in statistical terms. The magnitude of the turbulence is usually expressed in terms of theroot mean square value (or variance) of the longitudinal fluctuating component (that is, the component in line with the mean wind direction), The magnitude of turbulence may also be expressed non-dimensionally as the ratio of variance to mean wind velocity at a given height above the surface. This ratio is referred to as the intensity of turbulence.
