Ejector in Code_Saturne

Hello, all.

I am struggling a bit with analysis of a air-air ejector in Code_Saturne. Geometry and mesh have been defined in Salomé 4.1.4 and I am using Saturne 1.3 for analysis.

The model has an inlet (InletA) where the boundary condition is an inlet u-velocity, an ejector outlet (Outlet) and finally the suction port (InletP) where boundary conditions are outlet.

I would then expect at least (a) the flow through the Outlet being a sum of InletA and InletP and (b) a flow directed into the ejector through InletP and a flow directed out of the ejector through Outlet and (c) same static pressure af the interfaces InletP and Outlet. But ...

What I see is a laminar shaped distribution of flow at the InletP (suction port) where the velocity vectors point *out* of the ejector!!! and a clear *increase* in pressure between the InletP and Outlet.

I will appreciate much your comments to the analysis/model.




Just adding a remark: I assume it is either my model, the Code_Saturne analysis and/or the Salomé postprocessing that is playing me a trick as I know from real life measurements with exactly this ejector that the flow through the InletP boundary should go into the ejector and not away from it.<br /><br />Post edited by: Torben, at: 2010/01/17 17:18

(a) the flow through the Outlet being a sum of InletA and InletP

Then both inlets should be specified as inlets...InletP should not be an outlet.

Thanks for the answer. I specifed outlet for the InletP because I consider it a free passage. If specifying inlet, I have to specify a velocity/massflow through that boundary. While studying the log files from the analysis I see that the sum of ingoing and outgoing massflows are zero and Code_Saturne writes in the log that mass flows in through some outlet boundaries - but it is not mentioned as an error.

While animating the velocity vectors in InletP from the iteration history, they are seen to oscillate a bit in the beginnning going in going out, but after some 150 iterations they stabilize with the outward going distribution described above.

My initial idea was to see how much &quot;passive air&quot; will be transported by a given amount of &quot;active air&quot;. Is there a more correct way to specify a free flow boundary ?

Just a little update from my struggle with learning CFD

Checking the documentation of Code_Saturne I have confirmed that my use of the outlet as a &quot;passive&quot; inlet is essentially correct, so I decided to give it a go again.

I believe that I am now facing what makes CFD a challenge to learn by self study - the sense of balancing the mesh and solutions parameters to the task at hand.

First I cheated a bit and increased the viscosity. That naturally made the solution a bit more stable, and most recently I ran the analysis in laminar mode. Now it even converges so I believe next step will be trying to generate a mesh more optimised towards representing the boundary layers and maybe research a bit on the different turbulence models.

It ain't easy - but kind of fun

I have uploaded the velocity vector iteration history for the laminar run for you to see. It's an AVI of 10MB so it takes a bit to fetch.

tpa.orgfree.com/cfd <br /><br />Post edited by: Torben, at: 2010/01/31 19:46

Does the 'outlet' BC in the Code Saturne GUI specify a pressure? I was just thinking that if it does this might force flow out the outlets. The few times I have used Fluent I recall using pressure outlets.

If I recall correctly, in the Code Saturne documentation there were a bunch of outlet BCs but I dont remember if any specified the pressure or if that much information was even provided. In the GUI there seems to be one outlet BC. Perhaps there are settings elsewhere I missed?<br /><br />Post edited by: Matthew Bondy, at: 2010/02/06 00:55

Code_Saturne is a general purpose computational fluid dynamics free software. Developed since 1997 at EDF R&amp;D, Code_Saturne is distributed under the GNU GPL licence. It is based on a co-located Finite Volume approach that accepts meshes with any type of cell (tetrahedral, hexahedral, prismatic, pyramidal, polyhedral…) and any type of grid structure (unstructured, block structured, hybrid, conforming or with hanging nodes…).

Its basic capabilities enable the handling of either incompressible or expandable flows with or without heat transfer and turbulence (mixing length, 2-equation models, v2f, Reynolds stress models, Large eddy simulation…). Dedicated modules are available for specific physics such as radiative heat transfer, combustion (gas, coal, heavy fuel oil, …), magneto-hydro dynamics, compressible flows, two-phase flows (Euler-Lagrange approach with two-way coupling), extensions to specific applications (e.g. for atmospheric environment).

Code_Saturne can be coupled to thermal software SYRTHES for conjugate heat transfer. It can also be used jointly with structural analysis software Code_Aster, in particular in the Salomé platform. SYRTHES and Code_Aster are developed by EDF and distributed under the GNU GPL licence.