CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Mesh Validation - Inlet (

Daniel June 15, 2009 13:12

Mesh Validation - Inlet

i want to Validate several Meshes in order to know which Number of Cells is enough so that the solution doesn't change any more.

Therefore i created for my geometry several meshes and refined them step by step. After running CFX Solve, i pointed in CFX Post to the inlet of my gas flow and let cfx post readout the velocities in 1- 20 mm distance of the inlet (1mm steps). The problem is now, that my velocity is increasing in the beginning of the inlet. but There is no change in the crosssection in the first 5 mm, and after that it rapidly changes (becomes bigger). Its like a free jet flow. so the velocities should go down.

Any ideas ?

ghorrocks June 15, 2009 20:18


You will need to provide more details of what you are trying to do. A picture of your geometry would help too, also the CCL is good.

Glenn Horrocks

Daniel June 16, 2009 02:30

Ok some Data about my simulation,

Using. SST model, supply Gas is Argon, T is 293 K (Isothermal in the reactor), Thermal Energy Modell is used. its a very simple geometry and so is the simulation. consider it just as an 5mm long tube with 8mm cross section (D0,inlet) and after that the cross section expands abruptly to more than 10*D0. Everything else in the Sim. should not affect this inlet problem

ghorrocks June 16, 2009 08:06


We will need more details to really help. An image of the region would be good, also be more specific about the velocities you are measuring.

But some thoughts (can't be too specific as I still have no idea what you are doing):
1) If you specify the inlet as constant velocity flow the boundary layer will grow causing the flow next to the walls to slow down and the central bit speeds up.
2) You said the flow is isothermal so I assume you are not modelling compressible flow? If you are modelling compressible flow you can get some strange effects as the flow expands out.
3) Are you sure your inlet is far enough upstream to not affect the results?

Glenn Horrocks

Daniel June 16, 2009 15:43

thanks for your reply again, but now i know what is wrong. The velocity profile at the inlet cross section is plug flow and of course it speeds up then when it develops a parabolic profile. so it was point number 1 you suggested. Thanks for the help!

All times are GMT -4. The time now is 15:55.