[renyun0511]

Hi, all,

Recently, I try to simulate the transient inner flow of centrifugal pump by PimpleDyMFoam on OpenFOAM-2.1.0. I am so puzzled about the setup of interfaces between different cellZones. As we known, pump contains both rotational part and stationary part, and I add an outlet pipe with about 4 times the diameter of volute outlet. The attached pictures show the 3 parts of my pump model are impeller, volute and outlet pipe. The type of the interface between rotor and stator is defined as cyclicAMI. But, I don’t know what type is defined as the interface between stator and stator?

Here, the impeller is rotating around the positive direction of Y axis. Both the volute and the outlet pipe are stationary, if I define the type of interfaces between volute and outlet pipe as cyclicAMI, and I will see the torque is turned from negative to positive, which is wrong. If I delete the outlet pipe from my model, and there is only two interfaces left, which are the interfaces between impeller and volute. I defined the type of the interfaces between impeller and volute as cyclicAMI, and everything is OK, torque here is negative all the time.

So, my question is: what type of the interfaces between volute and the outlet pipe is defined?

Thank you very much!

Yun Ren

[fredo490]

Maybe you can have a look to this presentation: http://www.tfd.chalmers.se/~hani/kur...inery-OFW7.pdf

[renyun0511]

Hi, Heckmann,

Thank you for your reply! I don't understand it still. The .pdf shows more information about OF-1.6-ext. But, here, I use OF-2.1.0. I'm afraid that there is a little difference about them with the setup of interfaces. While it use GGI or cyclicGGI to define the type of interfaces by OF-1.6-ext, and I just know cyclicAMI on OF-2.1.0. I also try to use cyclicSlip or cyclic to define the interfaces between stator and stator, but it get error!

Regards,

Yun Ren

[fredo490]

If you want to run your simulation in a steady state, you can try to look for "MRF Openfoam" on Google. You will find lots of info. You can also try "RBF Openfoam"

I know that OpenFoam 2.2 has implemented the MRF in another way than before. It is introduced here : http://www.openfoam.org/version2.2.0/fvOptions.php

Previously, we had a dedicated solver for the MRF. But now with Openfoam 2.2 we have a general solver able to handle all "options" such as MRF.

You can try to look at the previous Workshop (7 maybe) if there is any example/practice with a rotor. You might be able to download the full case and look at the settings.