Boundaries definition
 

Hi all:):

Is there any way in a free-surface model to don`t specify the normal speed in the inlet boundary? Well, of course I need to specify initial values but i don`t want to specify that value (the normal speed) in the boundary. I just want to give an initial height of the water (in the upstream and downstream sections) and I hope that the model converges to some discharge. Is this possible?

How can I create periodic boundary conditions?

Many thanks

Look in the available options in CFX-Pre. You can also set a pressure.

No. You can specify the inlet water height but not the outlet.

This is described in the documentation. It is also mush easier in V12. Are you using V12?

Salam = Hi
you can use static pressure or total one
I think I'm working with a project with the same definition for inlet but not periodic
· I used static pressure and defined a function in expression. see tutorial 7 , I used the same function for UpPres and put it for static pressure. in this tutorial you can learn how specify the Hd hydraulic head. see expressions and use the step function. learn about it in guide.
I have version 11
good luck

salam = hi


in tutorial 7 it used outlet water height if i'm right.

• could you please tell how much it's different in version 12 from 11?

regards
zandi

From memory improvements to periodic boundary conditions were made, allowing you to set an inlet and outlet to be linked with specified flow rate or pressure drop or a few other options.

Hi All.

Thanks a lot for your suggestions.

Yes ghorrocks I am using V12. I have already saw in the documentation how to set the periodic boundaries (it`s called domain interface isn`t it?). However I still have one question. In the option Interface model I have 3 options : mass flow rate, none and pressure change. At the present moment I have chosed none...Is this a problem?

Setting periodic boundary conditions there is no need to define inlet boundaries with velocities isn`t it?

Zandi, I am going to see tutorial 7.

Regards

A periodic boundary is a type of domain interface. There are other types as well.

The three options relate to how the flow goes out one end and comes in the other. Mass flow rate means the mass flow will be set to your specified amount. Pressure means the pressure rise/drop over the boundary will be set. None means all flow variable map over.

If your model is one chamber in a cascade of many identical chambers then periodic boundaries can be a good approach. You the use a pressure or mass flow periodic boundary to drive a flow through the domain and you will get the representative flow in the chamber.

In your case I assume you have a mass flow rate, so in this case you will not need to set velocities or fluid heights. However an initial guess which is close may help convergence.