Closed moprak-nrel closed 2 months ago
Comparing the old/fixed AMD behavior against Smagorisnky: amd_line_plots.pdf. There is still a significant contribution from the buoyancy term in \nu_t
near sharp \theta
changes, but it's less pathological than the previous behavior.
This PR has a few different things:
fvm::gradient
instead of hand computing it, this allows the model calculations to also look just like the tensor expressionFieldPlaneAveraging
, currently the only access to the derivative is through single point functions.TODO: