ENH: Warning/Error Message for Reynolds Stress Tensor Input - DFSEM
DFSEM requires Reynolds stress tensor as input. Nonetheless, the tensor values need to be constrained in few ways in order to avoid mathematical domain errors, which is currently absent.
Steps to reproduce
Assume a user inputs this:
R uniform (1 1 1 0.1 1 0.1);
However, for example, R11*R22 - (R21)^2 > 0 should be held, which is negative with the particular setting above.
Consequently, the relevant case hangs at the outset of the first iteration without raising any warning/error message (Observed in v1712, v1806, and development-branch).
What is the expected correct behavior?
The user should be accordingly informed, and directed to the correct usage with a warning message; and computation should be ceased.