[BasG]

Hi CFD-friends!

I am trying to model a rotor with forced rotation in combination with a static tube around the rotor (see attached image 1 (static tube = red, dynamic part = gray) & 2). At the moment I am using the PimpleDyMFoam solver in combination with the AMI approach using dynamicMotionSolverFvMesh - solidBody. I managed to setup a working simulation with the rotor only. Therefore I selected a rotating cylinder around the rotor (see also attached image 3 & 4).

The next step is to combine the dynamic rotor with the stationary tube. The design of the rotor is such that it almost 'touches' the static surrounding. This means that there is hardly (if no) space between the rotor and the surrounding tube. It is therefore that selecting a rotating cylinder around the rotor (but within the tube) with an interface at which the mesh quality is high enough to not result in errors in the solver seems impossible.

Do you have a suggestion how to tackle this challenge? Is it possible to include both the static and rotating geometry inside the rotating cylinder and somehow freeze the static geometry within this cylinder (e.g. by means of its boundary conditions)?

Is the use of AMI indeed a good choice or should I rather look at different approaches such as the overPimpleDyMFoam-solver?

I hope to hear from you!

All the best,

Bas

[BasG]

After trying many attempts, I was able to place the AMI-interface inside the wall of the stationary cylinder. By setting the U-boundary condition of the statinonary cylinder to fixedValue 0, the flow velocity along the cylinder is set at 0. However, the inner side of the cylinder still moves with rotating mesh zone. Since the cylinder is (almost) perfectly round, this approach is considered acceptable for my case.