![]() |
Abrupt jump at inlet in rhoCentralFoam for fixed pressure & velocity
I am trying to simulate compressible pipe flow in rhoCentralFoam. I am providing a fixed pressure and inlet velocity at the inlet but this causes an abrupt jump in properties from the inlet patch to the first cell and another jump from the first cell to the second.
I do not encounter the problem if I specify the outlet pressure (waveTransmissive with fieldInf), let the inlet pressure be waveTransmissive (without any fieldInf) and fix the inlet velocity. However, this is impractical since I would like to see how the flow develops from fixed inlet conditions. I have a feeling that the problem is due to the manner in which fluxes are calculated in rhoCentralFoam. I could be wrong though. Does anyone know of any correction to apply for the incoming flux to resolve this issue? Many thanks. |
Mathematically you can't do that. Read the book by Malalasgar about the boundary conditions.
|
Thank you for the quick reply.:)
So, am I to understand that, currently, one cannot specify fixedValue inlet conditions for pressure and velocity simultaneously in rhoCentralFoam? And, is my guess, that the central-upwind flux scheme is the reason behind it, correct? Many thanks! |
You may find a better explanation of this in Hirsch, but this should also give you the idea:
Look up an-introduction-to-computational-fluid-dynamics by versteeg, Read page 27, 2.6 Classification of physical behaviour. |
Thanks for the tip!
I have been reading up on hyperbolic conservation laws since I started to use rhoCentralFoam about a year ago. I still haven't quite gotten one aspect of it: the energy equation. I have already posted this question [Energy equation rhoCentralFoam (revisited)] before but none answered. I am repeating it here in the hope that you (or someone else) might direct me in the right direction. In the inviscid energy equation, why does sigmaDotU, clearly a viscous term, appear? Furthermore, U here is interpolated using Kurganov's flux scheme which was developed for the hyperbolic conservation laws, Euler equations. Even the wave speeds are computed based on Euler equations. My understanding is that the flux scheme is to compute a numerical flux but here they have used it to interpolate just U. I don't quite follow... Thanks once again. USV |
All times are GMT -4. The time now is 18:03. |