February 13, 2010, 11:26
|
LRR Rstm pipe flow takes too long to get fully developed
|
#1
|
Senior Member
Join Date: Mar 2009
Location: Norway
Posts: 138
Rep Power: 17
|
Dear foamers,
I'm setting up a pipe flow and comparing this to Eggels data for pipe flow at Re = 5500. k-omega-SST yields "almost" satisfying results, and now I'm trying with LRR Rstm turbulence model instead. So far the plots for 'k' seems ok, but a thing that worries me is that it takes so long for the flow to get fully developed.
I include here a screenshot with a number of different plots, showing how axial velocity varies along the entire pipe length, and far beyond the "5 X diameter" rule.
Also included is the residuals plot. I feel it's clearly something fishy going on here, as the pressure and U_y and U_z never drops below 0.001, even though they they are set in fvSolutions (both tolerance and relTol) to limits far below.
Suggestions on to how to improve this?
|
|
|