I'm wondering about the validity of the flow velocity calculation for user-supplied conduit shape curve.
I simply compared the discharge/velocity/depth calculated with a rectangular closed shape and exactly the same curve but defined as a user-supplied shape curve and the results are significantly different.
Furthermore, in the case of the user-supplied curve, if I compute manually the discharge with the velocity and depth calculated it is completely different than the computed discharge. So I suppose there is a problem in the flow velocity calculation.
Therefore my question are:
- is there a way to correct this wrong velocity calculation?
- what is the influence of this wrong velocity on the resolution of the St-Venant equation. Indeed the frictional, inertial and losses term in the discrete form of the equation all need the velocity information. I'm not sure anymore which simulation results I can trust.
Thanks for your help,
N. de Ville