[yaseen wsu]

hello every one
who can suggest me an appropriate condition at inlet (for spillway model) in order to water level become constant, note that I used bulk mass flow in a chamber at inlet but the result of depth of flow (within model) lower than the experimental but when I used velocity at inlet for all height of inlet (like inlet boundary flow over bump), the water level located at the top of reservoir and then fall down to its origin and discharge (from CFX post) greater that entered.
http://www.cfd-online.com/Forums/att...1&d=1459927464
http://www.cfd-online.com/Forums/att...1&d=1460028313

[ghorrocks]

This is straight forward - use an inlet boundary with the velocity and volume fraction specified. But the problem is the system you are modelling. At a certain free surface height it will have a certain flow rate (in other words, the flow rate through the system is a function of inlet reservoir free surface height) - and that flow rate and height has to match what you define at the inlet. If they do not match you will get a strange jump in free surface heights.

Your problem is that the way you are defining the inlet makes the inlet conditions a function of the system you are modelling. This is undesirable and makes setting the boundary difficult. It means you should move the inlet boundary to somewhere the flow is purely defined by external factors.

[yaseen wsu]

Originally Posted by ghorrocks

This is straight forward - use an inlet boundary with the velocity and volume fraction specified. But the problem is the system you are modelling. At a certain free surface height it will have a certain flow rate (in other words, the flow rate through the system is a function of inlet reservoir free surface height) - and that flow rate and height has to match what you define at the inlet. If they do not match you will get a strange jump in free surface heights.

Your problem is that the way you are defining the inlet makes the inlet conditions a function of the system you are modelling. This is undesirable and makes setting the boundary difficult. It means you should move the inlet boundary to somewhere the flow is purely defined by external factors.

thanks a lot, it is very interested explain, yes creating a jump in surface height is common when I used velocity with volume fraction (and pressure in global initialization), this is happen in both cases when I specify (1- only height of water as inlet boundary condition. 2-total height (water+air) as inlet BC), but in the case when I specify total height it give me higher flow rate that I entered (using velocity), and water level from top of boundary fall down to the original water level.
but in another case I treated top of reservoir as a wall (like closed chamber) and used bulk mass flow rate, free surface level is straight ( but in this region v = 0), unfortunately this way doesnot successful.
according to your opinion moving inlet BC further upstream can treat this case, (because I put BC in that location when I know free surface level after that I dont know)?
I surprised why (in the tutorial flow over bump), the inlet with velocity has no problem but I used same condition this things occur !!!
thanks

[ghorrocks]

Yes, the tutorial example uses this approach and it works there. In my opinion it is not good practice and only works because the free surface height just happens to equal the system flow rate in that case.

I think a better approach is to say that you cannot know both the free surface height and the flow rate. So you should specify one of them in your model and the solver works out the other. And your check for simulation accuracy is that the solver gets the other parameter right.