CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   CFX (https://www.cfd-online.com/Forums/cfx/)
-   -   How to simulate valve using pressure BC?? (https://www.cfd-online.com/Forums/cfx/79200-how-simulate-valve-using-pressure-bc.html)

 Luk_Fiz August 16, 2010 11:46

How to simulate valve using pressure BC??

Hi,
I am trying to simulate the following situation during 1 second transient run:
The domain is a pipe with inlet and outlet:
a) between 0[s] and 0.5 [s] there is given mass flow in the inlet. In the same time the outlet is closed by the valve (no mass flow). In this time the domain is "pumped up", with growing gas pressure.
b) between 0.5[s] and 0.8[s] both inlet and outlet are closed.
c) between 0.8[s] and 1[s] the outlet valve is opened giving free outlet (zero pressure).
I want to use compressible air.

And now: inlet is simple: MassFlow with step function. But I have no idea what to give to the outlet. I was trying to set up pressure equal to inlet but anyway there is mass flow through the outlet (because inlet mass flow sets up velocity).

Any ideas?
Big thanks,
Luk

 serezhkin August 16, 2010 14:01

Imho, you can split it into 2 runs:
1) 0-0.8s, wall BC on the outlet
2) 0.8-1s, pressure BC, using ouput file from run 1 as IC for run 2.

 Luk_Fiz August 16, 2010 14:29

Well, I agree but You know...:)))

 serezhkin August 16, 2010 14:46

I guess another way of doing it is to set pressure BC side as outlet and specify a very high pressure for 0-0.8s. When ANSYS detects a backflow from the outlet, it puts a virtual wall over it to prevent it, and gives you a warning message. This way, you will have a virtual wall over the outlet for 0-0.8s.

Pretty sure this is not what ANSYS programmers had in mind when they implemented this, though :)

 ghorrocks August 16, 2010 18:27

Your high pressure outlet idea might work, but it also might be dreadfully numerically unstable. As you say, it was never intended to do this so who knows what might happen. Splitting the run into two and changing the outlet BC seems like the best approach to me. Using the new multi-definition run thing (V12) I think you can do this in the same def file so it is easy to implement.

I assume these flow regimes come from opening and closing valves. Are you interested in the flow near the valves? In this case you should model the valve opening and closing and that will give you your boundaries and there is no need to restart the run.

 All times are GMT -4. The time now is 05:46.