[wolterj]

Hey CFD Experts,
I'm simulatation a turbulent flow of water throught a nozzel in a large airfilled volume. Point of interest is the spraycontour after the nozzel. Contourunstructured.jpg
The picture shows the contour(3D) of the flow. In this case the mesh is unstructured and the viscous model transition-SST is used.

Other specifications of the Simulation:
Multiphase: VOF- Level-set-Coupled with implicit-body-force and interfacial anti-diffusion
Method: PISO with skewnesscorrection 2 and N-S-coupling
as inlet-conditions a velocityprofile is used, which were calculated in a previous Simulation.


my Problem is:
if i change the viscous model to Transition SST-SAS or simply SAS i get peaks in the Scaled Residuals of the Continuity which result in rise of the Courant number over 250. If i now lower the timestep to lower the courant-number the Spraycontour changes and is no longer comparable to the previous simulations.
Anybody got an idea how i can use SAS without these peaks?
I tried it with a structured mesh, but the results were even worse although the calculation ran through quite smooth.
StrukturiertGroßerInnererKörperTransitionSSTSAS.jpg
Picture shows a 2D contour of the same nozzel as in the first picture. In this case a structured symmetric mesh were used combined with the viscous-model TransitionSST-SAS. As u can see, it differs quite much....


Thnak you for your help!

[rupak504]

For spray modelling you can try coupled VOF DPM model.