Convergent Science

True Geometry Representation

The user does not supply a grid as an input to CONVERGE™. Instead, the user must only supply the surface of the geometry with the boundaries marked. The grid is generated by CONVERGE™ at run-time. There is an inherent, albeit not obvious, advantage to providing the surface as an input to the CFD code. When the grid resolution is changed during the simulation due to Adaptive Mesh Refinement (AMR) or other embedding, the refined cells near the geometry walls are re-cut using the original surface definition.

In traditional CFD codes that provide the full grid as an input, the actual surface geometry is no longer available. The accuracy of the surface location is limited by the resolution in the original grid. Any grid refinement performed during the simulation will cut the existing cells. Any resolution added near walls will not result in a better representation of the actual geometry in other codes.

Only by including the original surface information, as is done in CONVERGE™, can the additional embedding result in improved accuracy in the near-wall flow predictions.

Share This Page