[ehsan120]

Hello everyone,

I am a new fluent user and I am recently working on a problem to solve. You can see it in the attached picture. This is a multi-species problem.

The container is full of gas1 (O2) initially with e5 (pa) pressure, and I want to fill it (fully) with gas2 (N2). Let's assume that I know the pressure outside of the inlet (which is not that much higher than the pressure inside). The container as you see does not have an outlet and I need to calculate the time required to see 100% of the gas2 inside. It means, in the end, we should just have N2 in the container. For this, we must have an outlet in the container, as far as I understand.
My question is, can we define and inlet-outlet BC, i.e., flow come insider and go outside of that region at the same time? If not, then how I can simulate this in the fluent?

Thanks in advance for your responses.

[LuckyTran]

Stuff can flow out of a pressure inlet. Stuff can also diffuse back up an inlet.

But stuff tends to not flow from low to high pressure. Just make sure you have the right BC so that you simulate the right thing.

[vinerm]

Pressure Inlet as well as Pressure Outlet boundary conditions allow flows in both directions. However, your scenario is unphysical in reality; even in real life you cannot use one boundary to fill in one gas while allowing other gas to move out. You have to separate the inlet and outlet, such as, by using a nozzle to fill. Therefore, even in Fluent or for that matter in any other software, even if you write it yourself, until and unless you separate the boundary, you will not be able to do what you want to do. If you apply pressure inlet, then the same pressure will prevail throughout the boundary. So, if the gas inside has to come out, either it will come out of all faces or none. Separation of boundary does not mean literal separation, it could be separated due to pressure difference, say, few faces will lesser pressure than others, or it could be due to some sink term near the inlet for the gas inside.

[ehsan120]

Thanks, vinerm and LuckyTran for the response.

vinerm:

Yes, I understand that just introducing one inlet and no outlet can raise some problems physically. Let's assume that we don't have the symmetry lines and the interior walls inside the container. If we have the container in the middle of a chamber, like a picture I have attached here, then that's a physical problem that mimicks what I want to model. Please note that this is a simplified case and there are so many technical issues and because of these complications, I have to just simulate the container and cannot go for the entire chamber.
Now, if this explanation is good enough, then how you can model the container to reach a result roughly close to the reality (when we have the entire chamber). We know the flow rate at the chamber inlet FYI.

Thanks for the help in advance.

[vinerm]

Yes, this is good and doable. You have to use species transport with two components, the gas already filled and the gas you want to fill. Chamber inlet can be pressure or mass flow inlet. Outlet has to be pressure outlet. Container inlet will just be an interior. Initialize with oxygen mass fraction of 1 in the container. Mass fraction in the chamber has to be what you expect it to be. If the objective is to predict the time required for purge, then the initial mass fraction of oxygen in the chamber will make a difference. The walls between container and chamber can be of 0 thickness or finite thickness; won't affect the solution much until the thickness is of the order of container size itself.

[ehsan120]

Thanks for the prompt and clear response.
How if I couldn't include the chamber in my simulations. I mean, I have just the container with one inlet. What boundary condition I can apply for the container inlet to get a result more or less close to the case when we include the chamber?

Thanks

[vinerm]

When you include the chamber, the inlet of the container will have flow leaving on some regions while entering on some. However, this ratio of outgoing to incoming flow may not be fixed. Depending upon the pressure inside the container and the outside, as well as the mass fractions of the species, this fraction of out going mass to incoming will keep on changing. Therefore, you may not be able to work with just the container. You can run one case and observe if the fraction of the faces at the inlet to the container that have out going flow remains more or less same.

If that happens, then you can identify the locations where flow is outgoing. Then create a model of the container alone and create two separate boundaries, one as inlet and the other as outlet; outlet will be composed of the faces where the outgoing flow was observed for the first case.