Correct Boundary Conditions?

[bplturner]

I am trying to simulate a vertical pipe attached to a reservoir that sits on top of it with about 100 feet of water. The height of this reservoir never changes, so the hydrostatic pressure remains the same. I want to know the major/minor losses associated with the pipe. I'm having trouble setting up boundary conditions that account for this constant pressure without specifying a boundary inlet/outlet velocity.

Can someone assist?

[tcdonley]

To find head loss by hand in this problem, I would think you'd need to calculate the Reynolds number and then dig out the Moody charts. To get the Reynolds number and ultimately calculate head loss, you must know the flow rate or velocity. Would it work to simply set (flow rate in) = (flow rate out)? I expect you must actually specify a number for the solver, but this shouldn't be too hard to do. Assuming you are pumping water into the lake against the hydrostatic pressure, then the flow rate where the pipe enters the lake should be the same as the flow rate at your pump, yes? Now given the flow rate you can find the velocity by hand and enter this in your model. If your pipe diameter doesn't chnage, then Vin = Vout anyway.

If you just want a number for head loss to size a pump or something, it's probably quicker to do this by hand then setup an FEA model unless your piping is extremely complex.

In any case - you must set conditions for velocity or flow rate to solve the problem.

Best regards,

Tim D.

[raback]

Hi

I'm having trouble setting up boundary conditions that account for this constant pressure without specifying a boundary inlet/outlet velocity.

You cannot set BCs for the pressure using the constant discretization. For more info search for 'gradp' on this forum.

Given the large size of your problem (i.e. large Re) you may run into other problems. But you could start with large viscosity to see that you at least have a solution for the BC problem...

-Peter