specified pressure drop

subhra

In many practical situations where the flow develops to have only a single nonzero component (and usually a linear pressure drop) the pressure drop across the duct can be used to calculate (theoretically)the flow rate and fully developed profile.

The most common numerical approach in primitive variables of specifying an arbitrary (usually flat)inlet flow profile and placing outflow conditions at a sufficiently long distance seems roundabout in such cases.

I am trying to solve numerically (with a SIMPLER code in Fortran) the full Navier stokes equation (for a flow which is known to reach a developed state) with the inlet pressure and exit pressure specified at the ends of my computational domain, with the implementation as outlined in page 203-205 in the finite volume book by Versteeg and Malalasekera.

(1) It would be a great favor to me, if any of you with similar experiece can provide some tips, identify potential problems,strategies etc. As of now my SIMPLER code is taking exceedingly large number of iterations to give a reasonable constancy of flow across different cross sections. Is there any remedy?

(2) Another problem, some researchers say in general this boundary setup is well-posed only in Stokes flow. I have Re~1 and get reasonable results.