Air flow patterns in the lee of model windbreaks

Abstract

From a consideration of the nature of air flow about bodies of high form drag it is shown that the existing range of experimental evidence enables windbreaks to be classified as aerodynamically bluff bodies when their permeability is below a critical level, which is about 35% in wind tunnel flow. Wind tunnel experiments with model windbreaks showed a transition from leeward flow which was independent of the Reynolds number and characterized by a turbulent wake, to flow dependent on Re and without eddying, at a level of permeability between 25% and 38%. With 50% permeability a considerable reduction in leeward velocity was maintained without appreciable disturbance of the flow. The direction and fluctuation of the flow in the 2-dimensional wakes of windbreak models of zero, 25%, 38% and 50% permeability was observed with a freely rotating vane. These observations confirmed the conclusions about flow in the lee of windbreaks reached from various considerations and will enable the qualitative characteristics of such flow to be predicted for a wide range of conditions. The location, extent and intensity of the disturbed flow behind the three densest models varied systematically with the permeability of the barriers and was noticeably influenced by the normal vertical distribution of velocity. In view of this and of the known dependence of the flow in the lee of windbreaks on the vertical velocity profile of the incident wind, it was concluded that the theoretical treatment of the leeward flow could be approached by formulating it as a function of the vertical profile of the free wind. The visual observation of the flow patterns left no doubt about the presence of intense eddying in the turbulent wakes at those locations where conventional ‘shelter’ curves show relative velocity to be at a minimum. Such curves are therefore misleading indicators of shelter effect in proportion to the degree that the wake flow fluctuates from the horizontal. This is due to the limitations of the uni-directional anemometers usually used in shelter research. For progress to be made in the quantitative prediction of shelter effect more sophisticated instrumentation and analysis of the flow than hitherto employed will be required.