CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLOW-3D (http://www.cfd-online.com/Forums/flow-3d/)
-   -   set the outflow curve at downstream (http://www.cfd-online.com/Forums/flow-3d/102029-set-outflow-curve-downstream.html)

Andy Chen May 20, 2012 23:34

set the outflow curve at downstream
 
Dear experts:

When I set the outflow curve(B.C. use volume flow rate) at downstream. Then water flow toward upstream. That is not right. It Should toward downstream by gravity. How did I set Boundary condition if I want to set the outflow curve at downstream.

Thanks your help!!

JBurnham May 21, 2012 10:34

Set the velocity component to a positive or negative vector that points out of the domain. By default, the volumetric-flow boundary vector points into the domain, normal to the boundary, and you have to specify it if you want a different direction. Note that a downstream Vfr boundary will only remove the fluid which is in contact with the boundary. If dry spots appear, the flow rate will oscillate. Why not use a Vfr upstream instead?

Andy Chen May 21, 2012 21:09

First of all, thanks your reply, JBurnham.
I simulated the reservoir case. It should be controlled by gate at downstream. But I don't know how did use gate in FLOW-3D. So, our team consider that will set outflow curve(volume flow rate of B.C.) at downstream to solve the problem. Do you have any suggest can provide our team?

Thanks!!

MuxaB May 24, 2012 22:05

Quote:

Originally Posted by Andy Chen (Post 362315)
First of all, thanks your reply, JBurnham.
I simulated the reservoir case. It should be controlled by gate at downstream. But I don't know how did use gate in FLOW-3D. So, our team consider that will set outflow curve(volume flow rate of B.C.) at downstream to solve the problem. Do you have any suggest can provide our team?

Thanks!!

Andy, just like Jeff said, set the flow rate vs. time curve and define the flow direction vector pointing OUTWARDS. By default it points inwards.

Andy Chen May 25, 2012 10:19

Quote:

Originally Posted by MuxaB (Post 363010)
Andy, just like Jeff said, set the flow rate vs. time curve and define the flow direction vector pointing OUTWARDS. By default it points inwards.

Yes, I already solved about flow direction vector problem. I just reply that Jeff said "the flow rate will oscillate. Why not use a Vfr upstream instead?". I simulated the resvoir and it was controlled by gate at downstream(I set Vfr at upstream). So, our team consider that could set Vfr at downstream. Maybe it can solve about "the resvoir was controlled by gate" problem. That was I do why I set Vfr at downstream

JBurnham May 25, 2012 11:21

I assume that 'the reservoir is controlled by a downstream gate' means that there is gravity-fed overflow through (over or under) a gate, and that the flow is critical (Fr = 1) somewhere on the gate. Is the gate inside of the domain? If it's not, then you might as well use Vfr at the downstream boundary. If the gate is in the domain, however, you will want to let the flow go over the gate under the control of gravity, and be solved by the Navier-Stokes equations. In that case, it would be over-specifying the problem to define a flowrate at the downstream boundary. If I'm correct in my assumption (that Fr = 1 at the gate) then the downstream boundary will not have an effect on the reservoir level itself (no information propagates upstream from the critical transition point, and you could use an Outflow-type boundary downstream instead). You can use the Vfr boundary downstream - it will remove fluid at the rate you specify. I was just suggesting that you think carefully about whether that's a physical representation of the problem or not.


All times are GMT -4. The time now is 23:24.