The library does not compute WSS in the same way as the built-in FO

[timofeymukha]

Need to compare in detail what the difference is. Maybe we could just adopt the same approach as in the FO.

[timofeymukha]

See here for discussion of the FO implementation: https://develop.openfoam.com/Development/openfoam/-/issues/3011

However, I believe the implementation in the library is not fully correct either: the velocity is not projected before computing the gradient, and the stress is based solely on the wall-parallel velocity components, e.g. in the standard situation of y being wall-normal we have for the first stress component only nu du_x/dy and not the full nu (du_x/dy + du_y/dx). Both of these errors are pretty minor, since the velocity tends to align itself with the wall direction and the wall-parallel derivatives are typically negligable compared to wall-normal. So, previous simulation results are not invalidated but we should do things correctly.

My plan is to eventually align the implementation with that in the FO, so that it can be used interchangeably with the libraries own stress computation. But first we need to discuss the issue with the FO linked to above.

[timofeymukha]

Correction: The wall-normal gradient was in fact projected in the library. In the SampledWallGradU::sample() routine.