Skip to main content Skip to secondary navigation

Jianyu Wang

Main content start

“Application of an Immersed Boundary Method to Generate Boundary Layer Turbulence and Non-uniform Wind Fields”

Advised by Prof. Parviz Moin & Prof. Catherine Gorle

Abstract: Large-eddy simulations of wind engineering problems frequently rely on a combination of artificial turbulence generation and a rough wall function on the ground surface to generate a neutral surface layer flow. This approach may fall short when the aim is to capture non-standard wind conditions. Examples range from modeling the roughness sublayer for simulations of low-rise buildings to modeling profiles that deviate from the typical log-law shape or modeling unsteady events such as downbursts. To address these challenges, this study explores the use of an Immersed Boundary Method (IBM) to simulate the interaction between the wind flow and a combination of roughness elements, spires, or louvers positioned in the flow development section of the computational domain. Similar to wind tunnel experiments, the upstream configuration of these elements can then be optimized to reproduce a specific target wind field in the test section of the domain. We employ a direct forcing IBM, implemented in an otherwise body-fitted computational fluid dynamics (CFD) code. The implementation is tested on two set-ups: one with roughness elements that will generate a roughness sublayer for low rise building applications, and one with louvres that generate a non-standard mean velocity profile. For both cases, the IBM flow predictions in the test section downstream of the flow development section are compared against results obtained using body-fitted meshes for the same configurations. For the cases where data is available, comparative analysis with wind tunnel experiments is used to further determine the method’s capability to accurately simulate a range of ABL wind conditions. The findings demonstrate the method’s promising capabilities for simulating boundary layer flows with a range of mean velocity profiles and turbulence characteristics, simplifying the set-up of numerical analysis of a wide range of wind engineering flows without compromising on computational efficiency.