| Lack of a quantitative theory to explain airflow near wind barriers in the atmospheric boundary layer has hindered experimental programs in barrier research and made optimum barrier design for practical applications difficult. The objectives of this investigation were to develop a quantitative, theoretical simulation of airflow normal to narrow wind barriers of various porosities, and, when possible, verify the results using experimental data.;Upwind, downwind, and top boundaries of the solution domain were selected so they would not be influenced by the barrier. At those locations, boundary conditions were based on the well-known logarithmic windspeed profile and neutral atmospheric stability. As a barrier boundary condition, the porous wind barriers were treated as sources of horizontal velocity. The source strength for 20, 40, and 60 percent porous slat-fence barriers was determined by measuring the leeward windspeed profiles at 0.5 to 1.0 H.;For experimental verification of the simulation model, windspeed reduction and surface shearing stress were measured leeward of 20, 40, and 60 percent porous barriers with H/z(,0) = 75, and compared to the simulated results. Windspeed reduction data in the literature also was compared to simulated windspeed reduction with H/z(,0) = 300. Finally, vertical profiles of turbulence energy were measured near a 40 percent porous wind barrier and compared with the simulated results.;Treating porous barriers as a source of horizontal velocity, appears to be a valid method to obtain useful simulation results, because the leeward simulated and measured windspeed patterns were generally in good agreement. The windspeed profiles measured at 0.5 to 1.0 H leeward of porous barriers provided a measure of the source strength. However, averaging several profiles was necessary to obtain an adequate estimate of course strength.;To simulate the airflow near wind barriers, five, linked, partial differential equations were used. The differential equations described the conservation of horizontal momentum, vertical momentum, mass, turbulence energy, and dissipation rate of turbulence energy. In addition, an algebraic turbulence model was used to relate the turbulent viscosity to the turbulent energy and turbulent energy dissipation rate. Finite difference methods which used a combination of upwind and central difference schemes were used to solve the equations.;The measured and simulated maximum turbulence energies near the 40 percent porous barrier were in good agreement. But, the simulated rate of diffusion of turbulent energy to the surface was larger than measured.;The simulation model predicted complex flow patterns near the porous barriers, such as the recirculation zone induced by the 20 percent porous barrier. The size and intensity of the predicted recirculation zone was in close agreement with that measured by the sharing stress meter. However, beyond 6 H lee, the predicted shearing stress was greater than that measured. Too rapid diffusion of turbulence energy to the surface in the simulation model was probably responsible for the latter result. |
