mixing gases with different water content

[drick]

Hy everyone,

I am new to Code Saturne, and I need to simulate how effective is dry nitrogen purging wrt humidity control. The geometry would be a simple cabinet (just a box), where nitrogen with 5% relative humidity (RH) is injected with a predefined volumetric flow (through a short pipe). Initially the cabinet is full of air with a RH of about 60%. The cabinet is not hermetically sealed and in communication with the external environment through an aperture of a predefined area.The environment external to the cabinet is again air with a RH of about 60%. I would need to simulate the RH humidity content in the cabinet in function of the volumetric flow of injected dry nitrogen and also in function of the aperture size between cabinet and external environment. Having read the documentation and some of the forum posts I believe this should be possible through 'species transport' but I actually do not know how to do it. I guess I have to create a partial pressure scalar for the water content in the air and in the flushing nitrogen but I do not know where to start and how to set the problem. I hope someone can advise me on how to implement such a simulation. Even advices on simpler case based on the pipe tutorial of the CAELinux would be very valuable help. For instance by reverting the flow and having two mixing inlets and one outlet. One of the inlets with 60% RH and the other with 5% RH.

Thank you
Riccardo

[Yvan Fournier]

Hello,

In the previous posts you may have looked at, the closest one to your needs is the following:

http://code-saturne.org/forum/viewtopic ... bb1d#p6415

You are right about using species transport: simply define a "species" matching the humidity content, and assign the needed content for initial and boundary conditions.

For fluid properties (such as density, viscosity, ...) which are significantly dependent on the humidity content, you may need to define user formulas (in the GUI) or program user properties (using user subroutines) to define the those properties as a function of humidity (which, being defined as a species/scalar, will be an available quantity).

You do not need (or want) to define a separate partial pressure. A species represnenting humidity will be enough, and simpler.

Regards,

Yvan

[drick]

Thank you very much Yvan.
The post you refer in your reply is indeed the one that gave me hope...

I started to study it but I miss to which geometry it applies. When I open the xml and go to mesh selection, I get a warning message stating that echo.unv.gz and echo_extension.unv.gz mesh files are not present in the meshes directory search path. I also tried to look for these files in the CAELinux directories but I am always getting an error referring to search abortion. Am I missing something?

Thanks again
Riccardo

[Yvan Fournier]

Hello,

The example is from a validation test case, which is not published, and not a tutorial.

So you have to assign your own mesh (in the example, there are 2 joined meshes, but you may have something different, such as a single mesh).

You will also need to adapt the physical property definitions (the example is for a He/CO2 mix), and the boundary condition selection criteria.

So the case is only intended as an example.

If you want to start from scratch instead, the 3 import things do do are those I indicated in my first post:

- define a "species" (user scalar) representating the humidity
- define initial and boundary conditions with the humidity you want to model
- define physical properties based on your own equations if necessary (that is, for properties which are significantly dependant on the humidity in the chosen humidity range)

Regards,

Yvan