Date of Award


Degree Type


Degree Name

Master of Science in Mechanical Engineering


Mechanical Engineering

First Advisor

Zhang, Wei

Subject Headings

Mechanical Engineering


Understanding and controlling atmospheric boundary-layer flows with engineered structures, such as porous wind fences or windbreaks, has been of great interest to the fluid mechanics and wind engineering community. Previous studies found that the regular mono-scale grid fence of 50% porosity and a bottom gap of 10% of the fence height are considered to be optimal over a flat surface. Significant differences in turbulent flow structure have recently been noted behind multi-scale fractal wind fences, even with the same porosity. In this study, wind-tunnel tests on the turbulent flow and the turbulence kinetic energy transport of 1D and 2D multi-scale fractal fences under an atmospheric boundary-layer flow condition were conducted. Velocity fields around the fractal fences were systematically measured using PIV to explore the turbulent flow around the fences at the Reynolds number of approximately 3.6x104 based on the free-stream speed and the fence height. The turbulent flow structures induced by specific 1D/2D multi-scale fractal wind fences were compared to those of a 2D conventional mono-scale grid fence. In addition, each wind fences performance on wind speed reduction and sheltering effect were evaluated to determine the effectiveness of the fractal versus mono-scale grid geometries. Among the three wind fences, leeward of the fence, the 2D fractal fence is the most effective in reducing the incoming wind speed showing a maximum wind speed reduction coefficient of 0.90. Also, the 2D fractal reveals the most impressive shelter zone consisting of a magnitude of 0.40 ranging from x = 1.5H to 4.5H. Near the surface, the Reynolds shear stresses are very high for the conventional and 2D fractal fence, which is an indicator of potential particle remobilization. In contrast, the 1D fractal fence has the lowest Reynolds shear stress near the surface, which suggests the 1D fractal fence may be better in preventing particle remobilization from excessive turbulent stresses. The present results can contribute to optimizing the design for new-generation wind fences to reduce oncoming wind velocities and help snow/sand particle deposition on critical infrastructure such as roads and bridges.