Introduction to DAGMC

DAGMC enables CAD-based geometry to be used in Monte Carlo particle transport simulations. This is achieved by converting CAD geometry into a form that can be used by Monte Carlo codes, such as MCNP, Geant4, and FLUKA. DAGMC is an open-source project that provides a toolkit for creating and manipulating CAD-based geometry for Monte Carlo simulations. It is built on top of the MOAB mesh database and provides a set of tools and libraries for importing, processing, and exporting CAD geometry for use in Monte Carlo simulations.

Most Monte Carlo codes natively support CSG geometry as it is proven to be robust for particle tracking applications. However, use of CAD (Computer-Aided Design) geometry offers several advantages, especially in the context of modern engineering and design practices. CAD geometry provides benefits in a number pieces in the engineering design chain:

Visualization and Realism

CAD geometry provides an interacctive visual representation of the design, making it easier for designers and engineers to conceptualize and iterate upon their ideas. Unlike CSG, which relies on a fixed set of surface objects, CAD geometry offers a more intuitive approach by allowing users to see the design in a realistic manner.

Flexibility and Iteration

CAD geometry offers unparalleled flexibility in design iteration. Designers can easily modify shapes, dimensions, and features with simple clicks and adjustments, enabling rapid prototyping and experimentation. In contrast, while most Monte Carlo codes CSG offer some option to visualize geometry, these tools are often limited (2D slices and laggy interaction) and require a specific native format. Iteration on CSG geometry is in turn inherently less intuitive.

Simulation and Analysis

CAD geometry enables integration with simulation and analysis tools for evaluating the performance and behavior of designs under various conditions. From FEA to fluid dynamics simulations, CAD software allows engineers to validate their designs before physical prototyping, saving time and resources.

One significant advantage of CAD-based geometry in simulation and analysis is its capability for multiphysics domain mapping. Engineers can model complex systems involving multiple physical phenomena, such as structural mechanics, heat transfer, and fluid flow, by seamlessly integrating different simulation modules within a CAD environment. This allows for a comprehensive understanding of how different aspects of the design interact with each other, leading to more accurate predictions and optimized designs.

CAD geometry also facilitates the mapping of simulation results back to the design, providing valuable insights for further refinement. Engineers can identify critical areas of stress, temperature gradients, or fluid flow restrictions directly on the CAD model, enabling targeted design improvements.

While CSG can theoretically support similar analyses, CAD geometry offers a more practical and integrated approach, simplifying the workflow and enhancing productivity.

Advantages and Disadvantages of Surface Mesh (Triangles) vs. Volumetric Mesh (Tetrahedra)

When using CAD geometry for Monte Carlo simulations, the choice between

Advantages of Surface Mesh (Triangles)

  1. Higher Fidelty Boundary Representations: Volumetric meshes are often limited in how well they can resolve the boundary between parts in a CAD model due to constriants on mesh quality of interior elements. This commonnly results in a more coarse approximation of boundaries than can be achieved with a surface mesh for the same number of triangle elements. They are also able to more accurately capture features of varying sizes, such as sharp corners, thin features, or regions of high curvature. This in turn translates to more accurate representation of surface area and volume.

  1. Robust Meshing: A surface mesh can be generated for nearly any manifold volume. It is very rare that a triangle mesh cannot be generated for a given volume.

  1. Fewer Boundary Crossings: For the same level of detail, surface meshes typically have fewer boundary crossings compared to volumetric meshes. This can reduce the computational overhead associated with tracking particles through volumetric mesh elements, making surface meshes more efficient for certain regions in which the particle's average path length is much larger than the local elements.

Disadvantages of Surface Mesh (Triangles)

  1. Limited Internal Representation: Surface meshes do not capture proprety and field variation within the volume of the object directly, which can limit their applicability for simulations involving volumetric phenomena. Extrapolating volume-based information from surface meshes can introduce uncertainties and approximation errors.

  2. Ray Tracing Operations: While particles are able to travel from one surface to another in a volume containing a low- or zero-density material, the ray tracing operations involved in this process can be computationally expensive compared to adjacency search when traversing volumetric elements for volumes with higher collision densities.

Advantages of Volumetric Mesh (Tetrahedra)

  1. Volume Representation: Volumetric meshes directly represent the internal volume of the object, allowing for a more accurate simulation of complex three-dimensional phenomena and property representation (density, temperature, etc.) during particle transport.

  2. Accurate Boundary Representation: With a volumetric mesh, the surface geometry is implicitly defined by the tetrahedral elements, ensuring accurate representation of the boundaries and interfaces between different materials or regions.

Disadvantages of Volumetric Mesh (Tetrahedra)

  1. Higher Computational Cost: Generating and solving volumetric meshes can be computationally expensive, especially for large and complex geometries. The presence of tetrahedral elements throughout the model increases the memory footprint and element traversal requires more computational resources, in particular with respect to memory footprints of the mesh and associated acceleration data structures.

  2. Mesh Quality Concerns: Ensuring high-quality tetrahedral meshes, such as avoiding element distortion or ensuring element aspect ratios, can be challenging, particularly for irregular or highly curved geometries. Poor mesh quality can cause particle tracking problems. Additionally, the set of models that can be represented by a volume mesh is more limited than the set of models that can be represented by a surface mesh.

In summary, the choice between volumetric mesh (tetrahedra) and surface mesh (triangles) depends on the specific requirements of the simulation or analysis, including the desired level of accuracy, computational resources, and the nature of the geometry being modeled. Volumetric meshes offer accurate property representation across a single volume/part and are suitable for tallying internal fields while surface meshes are generally able to better conserve volume and surface area and may be better suited for models involving complex geometry or large numbers of parts. Performance considerations between the two are often problem-dependent and should be evaluated on a case-by-case basis.