Skip to content

Improve adaptive refinement method #8

@EvertBunschoten

Description

@EvertBunschoten

Make adaptive mesh refinement method more intuitive. Curvature-based adaptive mesh refinement is not intuitive and buggy. Also, making a good guess on how large the table is going to be a-priori is very difficult, resulting in unintentionally large tables. Proposed methods:

  1. Automated method based on regression error: Construct KD-tree algorithm with table nodes -> evaluate flamelet manifold data -> interpolate errors onto table levels -> apply refinement in areas with highest errors.
  2. User-based: apply refinement based on (relative) values of thermochemical state variables, e.g. "apply refinement where the temperature is between 50% and 100% of the maximum value on each table level"
  3. User-defined number of table nodes: Mesh resolution is adapted according to a specified approximate number of nodes per table level to avoid unintentionally large tables.

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't workingenhancementNew feature or request

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions