pyopensn.mesh.DistributedMeshGenerator

class pyopensn.mesh.DistributedMeshGenerator

Distributed mesh generator.

This class is responsible for generating a mesh, partitioning it, and distributing the individual partitions to different MPI locations. The mesh is generated on location 0, partitioned into multiple parts, serialized, and distributed to all other MPI ranks.

Wrapper of opensn::DistributedMeshGenerator.

Execute(self: pyopensn.mesh.MeshGenerator) → pyopensn.mesh.MeshContinuum

Final execution step.

__init__(self: pyopensn.mesh.DistributedMeshGenerator, **kwargs) → None

Construct a distributed mesh generator.

Parameters:

  • scale (float, default=1.0) – Uniform scale to apply to the mesh after reading.

  • inputs (List[pyopensn.mesh.MeshGenerator], default=[]) – A list of MeshGenerator objects.

  • partitioner (pyopensn.mesh.GraphPartitioner, default=None) – Handle to a GraphPartitioner object to use for parallel partitioning. This will default to PETScGraphPartitioner with a “parmetis” setting.

  • replicated_mesh (bool, default=False) – Flag, when set, makes the mesh appear in full fidelity on each process.

  • coord_sys ({'cartesian', 'cylindrical', 'spherical'}) – The coordinate system of the mesh.