Bug in computing hemodynamic indicies

[keiyamamo]

Same problem as KVSlab/VaSP#157

[hkjeldsberg]

@keiyamamo Any updates on this? A fix for P1 elements would be of great interest!

[keiyamamo]

Hi @hkjeldsberg
Fixing the problem of wall shear stress magnitude having negative value should be easy.
You just need to replace tawss = project(inner(tau, tau) ** (1 / 2), U_b1) as follows.

https://github.com/KVSlab/VaSP/pull/160/files

This should work for P1 since we are just taking the magnitude of dofs instead of using project.

To me, the other problem is that VaSP and VaMPy produce different values. I tested VaSP implementation against analytical solution and it is integrated as pytest. Therefore, I’m pretty confident that the implantation in VaSP is correct, but I might be wrong. Additionally, the old (see below) and current implementation of wall shear stress seems different in VaMPy . Maybe they produce the same result anyway, but it is unclear why or how these changes are made.