Part 2: Flow Around Obstacles: Geometry, Mesh, and Boundary Conditions

Part 2: Flow Around Obstacles: Geometry, Mesh, and Boundary Conditions

The flow around an obstacle is involved in many engineering applications, such as wind engineering, aerodynamics in automotive, airfoils, and other fields. It refers to the behavior of fluid dynamics that occurs when a fluid flows around an obstacle: cylinder, square, triangle, NACA, or angle member. To explore more about the key aerodynamic parameters, please check our previous Newsletter's article.

  1. Geometry

To accurately describe the flow around an obstacle, it is essential to provide optimal geometry dimensions. For example, when using a small computational domain height around the obstacle, this can lead to a loss of information on fluid detachment. A wider domain with optimal height prevents the blockage effect due to the upper and lower boundaries. An optimal distance should be chosen ahead of the obstacle to ensure a smooth and continuous flow of fluid and to accurately capture fluid blockage upon impact with the obstacle. Similarly, the distance after the obstacle should be sufficient to capture the fluid wake. For 3D geometries, it is recommended to use at least π *D for the domain width. Hereafter is an example of a computational domain chosen by Trias et al. to describe a flow around a square. They explored different dimensions, and it was observed that the size of the domain in the cross-stream direction has a relevant influence on the results probably due to some blockage effects.

Aucun texte alternatif pour cette image
Figure 1: Schema of the square cylinder problem of Trias et al. 2015

2. Mesh

In order to obtain accurate results, a high-quality mesh is essential. For flow around obstacles, the literature explores both structured and unstructured meshes. While a structured mesh is recommended for optimal CFD simulations, it is only feasible for simple geometries such as squares and cylinders. For more complex geometries like motorbikes or cars, a structured mesh can be challenging to construct and may not be the best option. Hybrid grids present an alternative option as they can combine an unstructured grid in complex zones with a structured grid in the remaining domain. This approach can offer a good compromise between accuracy and feasibility for complex geometries like motorbikes or cars. Huang et al. conducted a study to investigate the impact of grid type on simulation results, using different grids (see Figure 2 and Figure 3). The resulting aerodynamic coefficients and flow behavior were found to be similar for both grids.

Aucun texte alternatif pour cette image
Figure 2: structured grid around an angle member
Aucun texte alternatif pour cette image
Figure 3: Hybrid mesh around an angle member

3. Boundary conditions

Accurate and appropriate boundary conditions are essential for accurately describing the conditions applied to the computational domain boundaries. In their study of flow around an angle member, Huang et al. employed the boundary conditions shown in Figure 4. The upper and lower walls of the computational domain were assigned as symmetry boundaries to prevent any blockage effect of the fluid at these boundaries and so to indicate that the domain is open from the top and bottom. A velocity inlet was chosen to specify the fluid flow velocity and direction, while an outlet boundary condition was used to drive the fluid out of the computational domain by pressure. The validation of boundary conditions is essential to ensure that they accurately reflect the physical conditions of the system and result in reliable and meaningful simulation results.

Aucun texte alternatif pour cette image
Figure 4: Boundary conditions used by Huang et al.

4. Conclusion

Flow around an obstacle has a wide range of applications. In order to accurately simulate this flow and obtain reliable and meaningful simulation results, it is essential to carefully select the computational domain geometry and mesh, as well as appropriate boundary conditions. The literature provides numerous recommendations for optimizing these factors. Additionally, the specificities of the physical problem being studied should be taken into consideration when selecting boundary conditions.

5. References

  • F.X. Trias, A. Gorobets, A. Oliva, Turbulent flow around a square cylinder at Reynolds number 22,000: A DNS study, Computers & Fluids, Volume 123, 2015, Pages 87-98, ISSN 0045-7930, https://doi.org/10.1016/j.compfluid.2015.09.013.
  • Guo Huang, Fengli Yang, Hongjie Zhang, Fei Wang, Junke Han; Study on effect of grid meshing on simulation of flow field around an equal-leg angle member in transmission tower. AIP Conference Proceedings 26 July 2018; 1995 (1): 020007. https://doi.org/10.1063/1.5048738

要查看或添加评论,请登录

Asmaa Hadane的更多文章

社区洞察

其他会员也浏览了