[amintalezade]

Hello every one,
I'm trying to simulate the fluid flow around DTMB 4119 marine propeller. the steady simulation has been performed very nice and the complete-unsteady simulation using pimpleDyMFoam solver has also been simulated. the problems happen in periodic simulation.

I used cyclicAMI B.C for both side and inside faces. my question is that

1)does the mesh should exactly be the same in these both coupled faces?
2)how can I make this kind of mesh? (pointSync did not have any effect)
3)and finally how can i fix the attached error?



that would be really appreciated...

[sathya123]

Hi!

I am not in a position to answer questions 1 & 2, but looking at the solver messages for turbulent kinetic energy, I suspect there could an error with the numerics. Moreover, the solver does not seem to complain about the AMI, right? Could you share some specific details of your case? (Especially b,c,'s for turbulence, etc.)

[amintalezade]

Thank you Sathya for your fast reply
My case is a 3 blade marine propeller as depicted here with it's cyclic AMI boundaries. as I want to use pimpleDyMFoam solver i have separated these two zones. the innere cylinder rotates while the other is stationary. the P,U and k files are attached.

thanks you indeed.

[amintalezade]

I'm sorry, here is the B.C files

[sathya123]

Hi Amin,

We can start getting into the details of your case, by looking at the following:

1. Your inlet boundary conditions for k and epsilon. I just took a look at your boundary conditions for 'k'.
What turbulence intensity have you based it on? Also, what formulations have you used to specify 'epsilon'?

2. What discretization schemes have you used for the 2-equations (k-epsilon)?

3. Have you tried playing around with the relaxation factor ?

[amintalezade]

Hi Sathya

Finally I could solve that problem!!. The error caused by bad geometry definition for the inner Cylinder since I had imported the periodic geometry as .STL file. I chhange the file to the complete version and it was change to periodic as the blockMesh command was used. The previous and the new innerCylinder geometry has been attached for the others. it could be useful !!
But now I have another problem the process starts for a few itearations but it exits with this errorr.

Thanks a lot for your attention.