The current work examines the CFD analysis of a square-shaped structure that has undergone several aerodynamic alterations. The study is conducted for modifications like setback, corner cut, void\opening and their combinations such as corner cut with opening, setback with corner cut, setback with void\opening.
2.1 MODELLING
The analysis work in this paper is based on the outcomes of CFD simulations. SolidWorks was used to produce the building's 3D models, as seen in the square model in Fig. 2.2.1. For this study's tunnel modelling, ANSYS Design Modeler is utilized. The simulation of wind tunnel testing was done with ANSYS FLUENT. The drag coefficients of the building were calculated using these simulated wind loads.
2.2 MODEL
According to Fig. 2.2.2, the model employed in the study has a square-shaped plan. The building has a B/H ratio of 1/8, with a breadth of 50 m and a height of 400 m. To make it easier for further wind tunnel experiments, the model has been scaled down to a 1:1000 scale. The models of the building were made with three aerodynamic modifications including setback, corner cut, void\opening and their combinations such as corner cut with opening, setback with corner cut, setback with void\opening as shown in Fig. 2.2.2.
2.3 DETAILS OF COMPUTATIONAL DOMAIN
According to Fig. 2.3, the wind tunnel is a rectangular box. It is designed to have a downstream fetch of 15B and an upstream fetch, side clearance, and height of 5B, where H is the height of the building. These domain dimensions were determined by Y. Meng and K. Hibi[13]. In order to prevent the recommended inflow profiles along the empty computational domain of the building's windward side from deteriorating, the windward distance was set only a little bit below the suggestions by the AIJ guideline [14].
Both the domain's inlet and output have a velocity inlet and an outflow condition applied as their respective boundary conditions. Zero Pa is assumed to be the outlet's relative pressure. The building's walls are given a no-slip wall condition, while the free slip wall condition is applied to the domain's other four rectangular sides.
2.4 BOUNDARY CONDITIONS AND NUMERICAL METHODS
The wind environment in this study is produced using the well-known k-epsilon (k-) Reynold's-averaged Navier-Stokes turbulence model. The variance of velocity fluctuations is known as the turbulent kinetic energy, or k. The turbulence eddy dissipation, or, calculates the rate at which velocity fluctuations dissipate. The turbulence kinetic energy and turbulence dissipation rate's differential transport equations provide the values for k and ɛ.
The fluid's density and velocity are represented by ρ and U, respectively. Pk is the kinetic energy of the generated turbulence that results from mean velocity gradients, Pb is the generation that results from buoyancy, and YM is the contribution of fluctuation dilatation incompressible turbulence to the overall dissipation rate. C1 and C2 are constants. Prandtl number for k and ε are σk and σε. The values of C1ε, σk and σε are 1.44, 1 and 1.3. Atmospheric boundary layer (ABL) wind profile is used, is governed by the power law equation, which is given as follows:
yref is the height of the computational domain, uref is the free stream velocity, and the power law coefficient is 0.27, where u is the velocity at heights above ground,(for sub-urban area).
2.5 VALIDATION OF WIND PROFILE
To confirm the accuracy of the work, the velocity profile that was developed must be evaluated. The wind tunnel data utilized as illustrated in Fig. 5 and the computed profile are compared. Since the wind environments employed in both situations are similar, Fig. 2.5 demonstrates that the simulated profile reflects the profiles from wind tunnel studies and the power law equation with sufficient accuracy. As a result, it may be said that the study's flow characteristics are acceptable.
The square Model is used for validation purpose. The result obtained like Drag coefficient values are 0.48. The corresponding value obtained from this study are in close range as compare with Yukio Tamura, Hideyuki Tanaka, Kazuo Ohtake, Masayoshi Nakai, Yongchul Kim [2010].
2.6 MESHING
The software ANSYS FLUENT is used for meshing. To ensure that the solution variables decrease to acceptable values, mesh is modified. Tetrahedron meshing is applied. The domain face and edge has been appropriately adjusted. Figure 2.6 illustrates the meshing of the base model. For each model, grid convergence studies were done. To obtain convergent values of drag coefficients, the mesh conditions are modified while increasing the number of elements.