changes required at outlets to avoid inflow ?
 

Hi,

I have been facing inflow at one of the two outlets of my fluid domain. I extended the outlet pipes to a significant length but solver still blocks 100% of its area. I can not use 'opening' dur to certain reasons.

I have been specifying 'Average static pressure' (relative pressure=0)at this outlet BUT the function calculator in CFX-Post shows some different value (-2 kPa). As 'Average static pressure' allows the pressure to float locally, would it be causing problem?

Please help me with following questions:

1) why does CFX-Post not give correct specified pressure at outlets? (is it cell averaging?)

2) Should I specify 'static pressure' instead of 'average static pressure'? Would it help?

3) I got another idea to change this particular outlet to be 'mass flow' specified outlet and use the solution obtained from the previous settings as initial condition. How about it? Would this initial guess work?

Thank you so much, KK

Re: changes required at outlets to avoid inflow ?
 

I would change the outlet to opening for some time and then switch back to outlet.

1. CFX should give correct specified pressure at outlet. If it doesnt, i would say something is wrong with the set up of problem.

2. That will not make too much difference.

3. Yes, this sounds good. You should try that.

Khawaja

Re: changes required at outlets to avoid inflow ?
 

Do you have convergence and what is your convergence cretiria?

Re: changes required at outlets to avoid inflow ?
 

Thanks Khawaja and Omer,

Yes I have convergence (RMS 10^-4) and I monitor velocities at several points and they reach a stable condition where they oscillate around one average value .(not steady state as I am doing LES)

Because I am doing LES, opening type BC will cause solver to fail. (CFX manual also says that opening may make solver fail because it allows eddy entering across BC). Hence I can not go with opening.