CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Siemens (https://www.cfd-online.com/Forums/siemens/)
-   -   How can I solve this problem? (https://www.cfd-online.com/Forums/siemens/53474-how-can-i-solve-problem.html)

 Lee September 5, 2003 10:04

How can I solve this problem?

I want to simulate a model which involves the following physical phenomenom: In a straight round pipe which has a closed end, there are two parts separated by a sliding valve in some place of the pipe,the two parts contain AIR and CO2 espectively.At the beginning of the computation, an exhaust pulse was triged into the air in the fist segment at the mean time the sliding valve was remolved,after some time,the open end of the pipe also would be closed.

Because of the two different materials,so I need create a wall or baffle boundaries to represent the sliding valve,but I should remove the baffle or wall boundary after the beginning of the computation.

I suppose maybe I should use some "events",is it right?So how can I solve this problem?

Thank you very much for your help!

Lee

 cjtune September 8, 2003 04:18

Re: How can I solve this problem?

If your valve is to be removed very rapidly and the effects of the moving wall friction due to the valve sliding is negligable, you can just create two scalar quantities AIR and CO2 (for just one material type) and specify their initial concentrations in both parts of the (initially) separated pipe according to cell type number via the INITFI.F user subroutine. No need for any baffles.

If your the valve sliding motion cannot be ignored, then you have to consider a moving mesh sim. Again, specify the initiall concentrations via INITFI.F but this time you have an intermediate layer of (moving mesh) cells separating the two parts, pretty much like the 2-D piston-movement example in tutorial #13. But I can envision that it won't be that simple due to the circular cross-sections (of both your pipe and your gate valve?). You need to use arbitrary sliding interfaces.

 Lee September 8, 2003 05:17

Re: How can I solve this problem?

Thank you very much Cjtune!

 All times are GMT -4. The time now is 07:19.