Code_Saturne: Fluid properties

Hi.

I have searched the user interface as well as manuals available for Code_Saturne, but I cannot find where to specify the nature of the fluid in terms of compressibility. Essentially I would believe that for a liquid the calculation should assume incompressibility whereas for a gaseous fluid the variation of density with pressure should be taken into consideration. According to the specification sheet CS should be able to handle both ...

Can someone please help me with this?

Thank You very much for comments and advise.<br /><br />Post edited by: Torben, at: 2010/02/15 00:47

In the GUI when the fluid properties are specified there is a drop-down-list where the density can be set as constant or variable (I am unsure of the term used). This would be in the version included in CAELinux 2009 (Ubuntu 8.04LTS). I dont know if anything comes up when that option is changed, I have never run such a simulation.

Good luck,
- Matt Bondy<br /><br />Post edited by: Matthew Bondy, at: 2010/02/12 02:16

Hi, and thanks for the input. When density is specified variable, a routine should be present in usphyv.f calculating the density as function of other parameters. In the file a programming example is present. It is however not totally clear to me if this step is actually required or if a flag can just be set somewhere, telling that the fluid is incompressible or will be treated like an ideal gas.

In the example below I have an ejector with an annular suction port with specified 10m/s, and an inlet with active air at 1000m/s or roughly mach 3. A dynamic analysis gives the following distribution after 0.5ms of local velocity and total pressure.



To see if the flow is compressible or not I enclose a view of the U velocities at 50, 100, 150, 200, 250 and 300micro seconds after start. From these views I estimate that the disturbance travels with a speed of 120-190m/s. This would indicate a compressible flow.



One additional detail, however speaking against this is that the lowest absolute pressure is actually negative, which should in my expectation not be possible for compressible flows.

... or do I by any chance overinterprete the results?<br /><br />Post edited by: Torben, at: 2010/02/14 20:56

Post edited by: Torben, at: 2010/02/14 20:48<br /><br />Post edited by: Torben, at: 2010/02/14 20:48

Having just tried in a simpler model (straight pipe constant section, even velocity distribution) I am now convinced that the model straight off the SaturneGUI is incompressible, as the travelling speed of a plane disturbance is infinite. So I was overinterpreting the curves before. Sorry for this confusion.

Consequently it seems a custom usphyv.f is necessary to calculate compressible. I just have the impression that this user fort is intended for temperature dependency. If anyone here has tried this modelling rho as a function of pressure I will be pleased to hear about it. It could be implemented by the Ideal gas law.<br /><br />Post edited by: Torben, at: 2010/02/15 02:11

So, I found a lead: The compressible module is activated in usppmo.f. This module is described in the users manual section 4.38 (p.88) so that a C_S user with some experience in user fort application can understand ... though I may have to read it a few times.

I will get back once I get the module to do my analysis. In the mean time - if anyone has experience with the compressible physics module, feel free to share.<br /><br />Post edited by: Torben, at: 2010/02/16 12:49