Never miss a story from Computational Fluid Dynamics. Subscribe for in-depth analysis and articles.
Nick Wyman, Software Engineering Director at Cadence and an Associate Fellow of the American Institute of Aeronautics and Astronautics (AIAA), presented a paper where he talks about the different meshing technologies to capture and refine the critical locations to overcome the bottlenecks in preparing complex geometries for numerous applications. The video presentation titled Pointwise – The Choice for CFD Meshing is available at CadenceLIVE Europe 2021.
CFD practitioners often resort to simplifying complex geometries to ease the mesh generation process or in some cases for the grid-discretization to be possible at all. This reduces the accuracy of the CFD solution and can impact the performance of the application. To capture the leading edge of an airfoil or the fluid-structure interaction in a compressor or blood flow in a vascular bifurcation, the intricate features of the geometry need to be preserved. In the recent years, tremendous efforts have been devoted to developing algorithms that can automate mesh generation for improved accuracy and reduced human intervention. For example, the accurate meshing of healthcare devices such as an artificial tricuspid valve is mandatory for reliable prediction of heart dynamics. In such cases, the effort for accurate results is substantial as human lives are involved. An adaptive and automatic meshing technology would reduce the time spent on the repetitive meshing of similar equipment or components.
In this presentation, the author talks about the different automated meshing tools that enable surface meshing, off-body near-body meshing, mesh refinement, moving meshes, and high-order meshing for faster computation delivering accurate simulation results without the extra effort to manually guide the meshing process.
The precision and accuracy of the grid discretization technology is a precursor for reliable fluid dynamic studies of complex geometries. Several challenges need to be overcome before presenting CFD solutions as the primary tool for real-life system predictions.
Mesh generation is an inevitable process in CFD or FEA. It takes years of experience to generate a high-quality mesh that can capture stagnation points, flow separation, and shear layers. A mesh generation software that respects the geometry as designed and requires minimal user interaction can deliver a faster, flexible, and cheaper mesh solution.
A combination of structured and unstructured grids is favored because of its flexibility in handling complex geometry. Both meshing types have their pros and cons in terms of computation time, numerical accuracy, and aptness for the application. Structured meshes are easier to compute and are memory-efficient but they fall short in meshing complex geometries and may resort to approximating curved geometries. This is where unstructured meshing technologies are handy despite their longer hours for solving and greater memory requirements.
In the automotive domain, streamlining the body of a car for better performance and lower flow noise requires high-resolution mesh to study the wake layers and recirculation at both the rear and front end. To capture key flow physics, refinement of the mesh near the boundary layer, at the inlet and the outlet, near the wall, and at off-body locations is highly recommended.
Figure 1. High-resolution near-body and off-body mesh generation for a car.
It is impractical to use traditional meshing technologies for analyzing models involving high Reynold’s number and large Eddies due to the high cell counts required. High-order meshing technologies that utilize high-performance computing can provide high accuracy results with reduced memory requirements. But the generation of fewer cells of high quality comes at the expense of a higher cost.
In turbomachinery applications, the domain occupied by the fluid fluctuates with time and space. Under such circumstances, it is essential to have a moving mesh algorithm embedded within the meshing software to accommodate the need for re-meshing and interpolation so that the flow physics and resolution of the solution do not degrade.
Cadence Pointwise is a mesh generation software solution that has tools for geometry creation, geometry clean-up, and grid preparation. While generating a mesh, the lower-level entities glue the higher-level entities together to form a contiguous mesh which allows a lot of flexibility in mesh construction techniques and mesh styles. This flexibility is the meshing philosophy of the Cadence Pointwise product and allows it to be applied to a wide range of workflows. The mesh topology is independent of the CAD geometry and offers flexibility. The different meshing technologies in Cadence Pointwise can address the grid discretization challenges in varied applications.
Here, the user needs to provide bare minimum inputs for the geometry resolution such as mesh aspect ratio and mesh curvature resolution and a surface mesh is generated automatically. The meshing process begins with the geometry model (typically a B-Rep NURBS model) followed by the quilting technology embedded in Cadence Pointwise. Using the Cadence Pointwise Flashpoint tool, high aspect ratio cells are generated at the leading edge of a wing to effectively capture the geometry curvature.
Another important tool in the Flashpoint automation suite is the ability to refine the mesh. In this case, the mesh character remains unchanged, but the mesh size changes. This is a powerful and flexible technique for automating meshing. The non-dimensional, goal-based technique allows this meshing method to be applied to a wide variety of geometries and simulations. For thermal analysis of a drone in Figure 1, a mesh is created for both external parts of the drone and the internal printed circuit boards.
Figure 2. Grid discretization of a drone.
This tool is predominantly used for near body or boundary layer meshing with special handling of symmetry boundaries, sharp edges (in multiple marching directions), and baffles or thin surfaces. T-Rex generates layers of prisms and/or hexahedra to resolve the near-wall flow. In scenarios where the extruding layers are approaching each other; this method will automatically detect and stop the stacks locally to avoid collisions. This type of method applies to the mixed type of surface grids where certain regions are meshed using structured mapping while other areas are being meshed using Delaunay triangulation allowing flexibility in the meshing process. In Figure 2, the boundary layer mesh around the aircraft is created using the T-Rex meshing tool.
Figure 3. Boundary layer mesh around an aircraft.
High-quality, uniform cells are created using Voxel meshing offering excellent off-body resolution. This method operates in parallel and removes voxels intersecting the geometry for high-quality meshing. There is advanced handling for both internal and external flow geometry, symmetry boundaries, and transition layers. In Figure 3, the body of the Sedan is meshed using Flashpoint meshing, a robust boundary layer mesh is obtained using T-Rex meshing and the off-body features are defined using Voxel meshing.
Figure 4. Off-body meshing using Voxels.
This automatic mesh refinement tool is used only in those regions where the mesh is deficient. It starts by creating a baseline flow solution and using this flow solution, an estimate of the error corresponding to the deficiencies in the mesh size is determined. This step is repeated quite a few times to get a better hold of the mesh discretization error. For high-quality CFD meshing, this method can also be used on off-body voxel meshing for uniform and excellent resolution of the off-body features specially to capture the wake region. In Figure 4, the wake shear layer mesh for the Sedan is refined using the Mesh Adaption tool.
Figure 5. Mesh refinement to define off-body features.
This latest Cadence Pointwise meshing technology allows mesh generation using High-Performance Computing (HPC) with fewer elements, high accuracy, and reduced memory requirements. Previously, impractical analysis involving high Reynolds number and large eddies can be resolved using this meshing method. Even today, anisotropic boundary layer mesh with coarse transverse spacing, high element curvature, and mesh tangling are meshing challenges and these challenges are responsibly handled by the High-Order Curved Meshing technology. Figure 5 shows an airplane wing where High-Order Curved Meshing is used.
Figure 6. High-Order Curved Meshing in and around an aircraft wing.
Overset meshing methods are suitable for moving body applications. The mesh connectivity is recomputed in every iteration for the changing background of the fluid mesh. This meshing type is highly appreciated in turbomachinery applications. For example, the propulsion of the ship’s hull in Figure 6 uses Overset meshing to capture the moving body physics and Voxel meshing for the efficient discretization of the off-body volume and local refinement around the propeller region.
Figure 7. Overset and Voxel meshing around a ship’s propeller.
Cadence Pointwise is more than a collection of unrelated meshing tools. It offers flexibility in automating the grid discretization workflow for complex geometries without compromising the accuracy of the solution. Enough emphasis cannot be put on the importance of accurate meshing as human lives largely depend on it for various applications. The different meshing technologies embedded within Cadence Pointwise meshing software makes it suitable for various domains ranging from turbomachinery to medical applications.
Now that you've seen what Pointwise can do, why not request a free trial today?