Modelling the natural wind : wind protection by fences
Thesis DisciplineMechanical Engineering
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
This thesis describes the development of a simulated atmospheric boundary layer, and its application to the specific problem of windbreak aerodynamics. A review is made of the state of the art in wind tunnel simulation of atmospheric boundary layers. An investigation of the use of grids to simulate atmospheric wind velocity profiles and turbulence in a short tunnel working section, is used to justify the construction of an atmospheric boundary layer wind tunnel in the Department of Mechanical Engineering. The design and construction of this wind tunnel is described. The main working section of the new facility is 4 ft x 4 ft in cross-section and 40 ft long. The stepwise development of an accelerated growth, neutrally stable, simulated rural atmospheric boundary layer of approximately 1:300 linear scale is described. The boundary layer is grown in a distance nine times its final depth by means of an initial coarse grid, followed by trip fences of successively decreasing height and a baseboard of uniform surface roughness. The linear scale of the model flow is larger than in most earlier simulations of the accelerated boundary layer growth type reviewed by the author. An examination of the leeward flow behind model fence windbreaks was chosen as a first application for the simulated boundary layer. The object of this work was to give a clearer picture of the leeward flow field, and quantitatively relate mean velocity and turbulent intensity behind windbreaks. Measurements of mean and fluctuating velocities and energy spectra were carried out in the lee of model shelter fences of 0%, 20%, 34% and 50% geometric permeability. Results of the tests are comparee with existing field and wind tunnel data, with due regard for the uncertainty of hot wire anemometer measurements in fence wakes. The mean velocity reduction data provide an extensive verification for Jensen’s (1958) Model Law. The 20% permeable fence was found to give optimum mean wind reduction. The turbulence measurements have identified the regions dominated, respectively, by the bleed flow and by the displacement flow, and simple empirical equations are suggested to relate turbulent intensity and mean velocity in these two regions. Before a windbreak design manual can be compiled, further tests are required at a variety of ‘fence height’ to ‘upstream surface roughness length’ ratios. A basic format for a design manual is suggested.