5 Attachment(s)
Hi, I've made a larger domain and a finer mesh and run the 2D case again but this time full-scale.
I've used k-ep RNG instead of realizable. Simulation starts well, no need to modify under-relazation factors, but it end up in divergence (picture 1) due to turbulence viscosity ratio (TVR) limited to 1e5 in too many cells. I've taken a picture of TVR distribution just before divergence (picture 2) so I guess the problem is TVR limitation is reached in that area and then it's transported into the wake. I guess I need an finer mesh in the wake region to solve that, don't I? I've checked y+ around the plate and it's always under 3 (picture 3). Static pressure and x-wise velocity are plotted in pictures 4 and 5, they both seem right to me, including the profile distribution for the velocity. Then I've tried a finer wake mesh but there must be something else wrong because divergence occurs just from the very begining. Thanks |
All times are GMT -4. The time now is 17:42. |