Abstract:Efficient and robust anisotropic mesh adaptation is crucial for Computational Fluid Dynamics (CFD) simulations. The CFD Vision 2030 Study highlights the pressing need for this technology, particularly for simulations targeting supercomputers. This work applies a fine-grained speculative approach to anisotropic mesh operations. Our implementation exhibits more than 90% parallel efficiency on a multi-core node. Additionally, we evaluate our method within an adaptive pipeline for a spectrum of publicly available test-cases that includes both analytically derived and error-based fields. For all test-cases, our results are in accordance with published results in the literature. Support for CAD-based data is introduced, and its effectiveness is demonstrated on one of NASA's High-Lift prediction workshop cases.

What problem does this paper attempt to address?

This paper attempts to address the problem of achieving efficient and robust anisotropic mesh adaptation techniques in computational fluid dynamics (CFD) simulations. Specifically, the authors propose a fine-grained speculative approach to handle anisotropic mesh operations on shared memory architectures, aiming to improve parallel efficiency on multi-core nodes and evaluate the effectiveness of this approach in a series of publicly available test cases. Additionally, the paper introduces support for CAD data and demonstrates its effectiveness through a case from the NASA High Lift Prediction Workshop. ### Main Contributions: 1. **Application of Fine-Grained Speculative Scheme**: The fine-grained speculative scheme is applied to a set of metric-related mesh operations, suitable for both volume and boundary adaptation. This method provides the highest strong scaling speedup in existing scalability-prioritized methods and better weak scaling speedup in functionality-prioritized methods. 2. **Support for Geometric (CAD-Based) Models**: Extending the above operations to support geometric models by accessing the geometric kernel to obtain information about the curvature and local feature size of the domain, thereby recovering this information in the coarse mesh representation. 3. **Evaluation in Adaptive Pipeline**: The method is evaluated in an adaptive pipeline containing a CFD solver and a real geometric case, comparing the computational results with values in the literature, focusing on the main goal of mesh adaptation: capturing the features of the underlying simulation. ### Background and Motivation: - **CFD 2030 Vision**: Emphasizes the importance of efficient and robust mesh adaptation techniques for CFD simulations on future supercomputers. - **Limitations of Existing Methods**: Existing adaptive techniques need optimization to run effectively on current and future systems, especially in the face of large-scale parallel computing environments. ### Technical Details: - **Fine-Grained Speculative Method**: Concurrently executes multiple mesh operation kernels at a fine-grained (element) level, capturing runtime dependencies, and if conflicts arise, releases dependencies and continues processing other elements. - **Parallel Point Creation**: Each thread iterates over a subset of elements that do not meet the local length criteria, generates candidate points, and performs proximity checks to avoid creating points that are too close. - **Parallel Edge Collapse**: Used to suppress edges smaller than the target value, parallelized by traversing vertices and utilizing OpenMP constructs. - **Vertex Smoothing**: Improves mesh quality by optimizing the line segments connecting vertices to the centers of their cavity faces, ensuring no element inversion. ### Conclusion: The paper details the implementation of the fine-grained speculative method in the CDT3D library and evaluates the generated meshes through a series of CFD applications. The results indicate that this method performs excellently in terms of parallel efficiency and mesh quality, especially when handling complex geometries.

Speculative anisotropic mesh adaptation on shared memory for CFD applications

Parallel Adaptive Anisotropic Meshing on cc-NUMA Machines

Towards Distributed Semi-speculative Adaptive Anisotropic Parallel Mesh Generation

Towards Real-time Adaptive Anisotropic Image-to-mesh Conversion for Vascular Flow Simulations

Tasking framework for Adaptive Speculative Parallel Mesh Generation

Verification of Anisotropic Mesh Adaptation for Unsteady Mixing and Reacting Flow

Feature-based and goal-oriented anisotropic mesh adaptation for RANS applications in aeronautics and aerospace

Parallel CFD-Algorithm on Unstructured Adaptive Meshes

Mean Mesh Adaptation for Efficient CFD Simulations with Operating Conditions Variability

Monte Carlo multiphysics simulation on adaptive unstructured mesh geometry

Adjoint Based Anisotropic Mesh Adaptation for the CPR Method

Anisotropic unstructured mesh adaption for supersonic flow simulation

Some progress on CFD high lift prediction using metric-based anisotropic mesh adaptation

Anisotropic Mesh Adaptation For Steady Convection-Dominated Problems Based On Bubble-Type Local Mesh Generation

Parallel Hybrid Mesh Adaptation by Refinement and Coarsening

Conservative solution transfer between anisotropic meshes for time‐accurate hybridized discontinuous Galerkin methods

Anisotropic mesh adaptation for 3D anisotropic diffusion problems with application to fractured reservoir simulation

Flow2Mesh: A flow-guided data-driven mesh adaptation framework

Consistent Multi-Threaded Mesh Refinement with Adaptive Length Scale Estimation for Moving Boundary Problems.

Experience and Analysis of Scalable High-Fidelity Computational Fluid Dynamics on Modular Supercomputing Architectures

An Anisotropic hp-mesh Adaptation Method for Time-Dependent Problems Based on Interpolation Error Control