SAR evaluation reports "grid too coarse", but works after cropping

[Zoe]

Hello,

I am having an issue when evaluating SAR results in Sim4Life.

When I try to view the SAR result (e.g., psSAR IEEE/IEC 62704-1 Avg.SAR(x,y,z,f0)) over the original simulation domain, Sim4Life reports an error indicating that the grid is too coarse, and I cannot obtain the SAR result.

However, if I first use Crop to reduce the analysis region to a smaller area around the region of interest, the SAR evaluation works and I can successfully visualize the SAR distribution.

The underlying simulation results and grid remain unchanged; I am only changing the spatial region used for post-processing.

Could you please clarify why cropping the analysis region changes whether the SAR evaluator considers the grid sufficiently fine?

In particular:

Could a coarse-grid region somewhere outside my region of interest cause the SAR evaluation for the whole domain to fail?
If the SAR result becomes available after cropping out those regions, can the SAR values within the cropped region be considered valid, or should I refine the simulation grid and rerun the simulation?

Thank you for your help.

[brown]

Even if your region of interest is well-resolved, running psSAR over the entire sensor/domain means the evaluator tries to process all relevant lossy voxels in that domain. If there is a region somewhere in the simulation domain where voxels are too large for the desired averaging mass, then the evaluator may report the “grid too coarse” error.

Cropping away those problematic regions can therefore allow the evaluation to complete for the remaining volume.

Values inside a cropped region are generally meaningful as long as the important features are not intersected by or too close to the crop boundaries. The exact definition of "too close" will depend on the averaging mass used. If you enable the "Compute Cube Information" option in the psSAR evaluator, you can inspect the averaging cubes that were used by the algorithm.