Improving Mesh Adaption with Fidelity CFD

The aerospace and defense industry continues to grapple with the challenge of ensuring accurate and efficient simulations, which are essential for expediting the design process. Various techniques have been developed to automate the mesh generation process in order to enhance flow-field predictions. Traditionally, creating meshes for computational fluid dynamics (CFD) necessitated a deep understanding of physics and adherence to complex best practices, often resulting in inconsistent outcomes. This resulted in poorly refined grids that either overperformed or underperformed, especially in crucial flow areas. A potential solution to this problem is output-based adaptive remeshing, which improves accuracy and keeps the cell count manageable.

In this blog post, we'll explore the concept of mesh adaptation, introduce the latest mesh adaptation features in Fidelity 2024.1, discuss its benefits, share best practices, and highlight important considerations when using mesh adaptation in Fidelity 2024.1.

About Mesh Adaptation

Mesh adaptation involves altering an existing mesh to capture essential flow features more accurately, ultimately improving simulation results without significantly escalating computational costs. This meshing process automates the mesh refinement process by targeting specific areas with vital flow features that necessitate smaller cell sizes. Leveraging this optimization enhances accuracy and maintains a manageable cell count, thereby improving the overall CFD workflow turnaround time.

Mesh Adaption in Fidelity 2024.1

The latest advancement in Fidelity 2024.1 enhances the capability of the density-based solver (DBS) in Fidelity CFD to conduct adaptive remeshing with Fidelity Pointwise and Fidelity Hexpress meshes. This mesh refinement approach relies on truncation error estimates to improve accuracy and performance. This is achieved by exporting a Point Cloud Data file (pcd file) that guides the mesher on where to refine, enabling an adaptation workflow. Additionally, a new Python script has been developed to automate the remeshing process.

When setting up the simulation, the user can enable the mesh adaptation feature with specific parameters. Once selected, a pcd file will be generated at the end of the simulation. Fidelity CFD includes a Python script that automates the remeshing process, streamlining the mesh adaptation workflow. Manually performing the remeshing and resetting the simulation setup with adapted meshes can be time-consuming. However, the Python script simplifies this by automating all mesh adaptation steps—from simulation to remeshing in Fidelity Pointwise or Fidelity Hexpress. There are two ways of running this script: in batch or from the GUI.

The user can specify the total number of remeshings to be performed and run the workflow for the specified number of cycles.

Advantages

The implemented adaptation method is based on reducing truncation errors and has been proven to be more efficient than adaptation based on gradient or curvature values (Roy, 2009). Here are some advantages of this method:

• The user does not have to specify the minimum and maximum gradient or curvature values.
• It takes into account the local cell size and addresses coarser regions first, ensuring better grid convergence properties during the adaptation process.
• It helps to avoid over-refinement of strong discontinuities compared to other methods. The multiscale parameter gives the user control over the adapted cell clustering.

Best Practices

The mesh adaptation in Fidelity CFD does not change the anisotropic layer mesh settings (boundary layer, anisotropic blunt surface, or edge refinement). For meshes with anisotropic boundary layer, leading and trailing edge refinements, it is important to ensure the settings meet the following requirements in the initial setup:

• Leading and trailing edges are sufficiently refined.
• First cell size and expansion ratio in the boundary layer are sufficiently converged.

It's best to use tetra-core meshes and avoid voxel-core meshes when using Fidelity Pointwise meshes. When using Fidelity Hexpress meshes, the volume-to-surface meshing approach is more reliable in the adaptation process than the surface-to-volume approach.

The multiscale parameter allows the user to control the adapted cell clustering. It is recommended that the value be kept between 2.5 and 4 for 3D meshes. Lower values result in more local clustering of the cells, while higher values result in the cells being more spread out. In practical terms, lower values will increase clustering in the shock regions, while higher values will increase clustering in the wake and vortex regions if the flow field features both a shock and a secondary flow, such as wakes or vortices.

Key Points to Remember

• Capture complex flow features accurately and efficiently through adaptive remeshing
• Automate meshing best practices to avoid over and under-refined regions
• Leverage efficient adaptation method based on truncation error estimates

Reference

Roy, Christopher (2009). Strategies for Driving Mesh Adaptation in CFD (Invited). 47th AIAA Aerospace Sciences Meeting, including the New Horizons Forum and Aerospace Exposition. 10.2514/6.2009-1302.

Learn more about adaptive remeshing workflow using the Fidelity Pointwise and Hexpress meshers alongside the Fidelity Flow solver in the online webinar "Enhancing CFD Accuracy and Efficiency with Adaptive Remeshing."

Date: Thursday, September 19, 2024

Time: 8:00am PT

Register Today!