[hani]

I am trying to evaluate OpenFOAM for a workshop in hydraulic water turbine drafttube computations (www.turbine-99.org). One of the cases in the workshop is to generate a steady solution of the flow in the drafttube, so I use simpleFoam. Niklas Nordin has helped me generate the appropriate inlet boundary conditions for velocity and turbulent quantities. The grid is highly non-orthogonal (specified at the workshop homepage), so I understand that I must have at least one non-orthogonal correctors in the SIMPLE algorithm. Is that correct?

Of course I have convergence problems, and I see at this discussion page that another user was recommended to use the inletOutlet boundary condition when there is a risk for recirculation at the outlet. In the drafttube case I am sure that I will get recirculation at the outlet. Right now I use zeroGradient for U, k and epsilon, and fixedValue, uniform 0 for pressure. Then I get the same kind of problem with divergence of the continuity as the other user had. What boundary condition is recommended to use, and how do I specify it for U, p, k, epsilon?

In our in-house code a Neumann condition on all variables works fine even if there is recirculation. Of course the inflow values are not correct, but it does not seem to influence the results much. We don't specify a constant pressure at the outlet, but we set a reference pressure anywhere in the domain.

Is there any other reason why the continuity diverges although the prime variables seem to converge?

Håkan.

[rcpoudel]

Dear Håkan,

I an new user. Just wondering how did you people (including Niklas Nordin) generate the appropriate inlet boundary conditions for velocity and turbulent quantities.
Appriciate your detail answer/hints.

Thanks in advance,
Ram

[hani]

Ram, I have previously written a pretty detailed instruction of how to generate inlet profiles in the forum. However, I now also post an example (warning, it is an ugly piece of code and should be improved):

setCombBC.tgz

This example is for a circular inlet facing the y-direction with axisymmetric flow in the negative y-direction. The inlet boundary condition is linearly interpolated from profiles that are determined from the boundaryProperties and
inletProperties.dat, which you should place in your constant directory. The example is a pre-processing code, i.e. it should be run as a separate application just to set the boundary condition.

Read through it and adapt it to your application, maybe you will learn something.

Håkan.