Rotating Radial Channel with Through Flow

I am trying to model a simple radial channel with rectangular cross-section that rotates about an axis with through flow using Caedium Professional. I obtained a solution with good convergence but the results are not what I expected from potential flow theory which gives a linear relative velocity distribution across the channel in the circumferential direction.

I am quite new to Caedium so it is likely me doing something wrong in the set-up process.

The channel is modeled in the XZ and YZ planes. I assigned water as a substance and specified the face at the inner radius as the inlet and the face at the outer radius as the outlet. The boundary conditions are the inlet volumetric flow and the angular velocity about the x-axis which I assigned to all of the walls. The flow rate should have been set high enough to prevent separation or recirculation.

Is this the correct approach? Any help would be appreciated.

No votes yet

Use a Moving Reference Frame (MRF)

If I understand your simulation correctly then the approach you've taken is incorrect. The angular velocity boundary condition applies only to axi-symmetric walls, such as a cylinder or sphere topology - as in the tutorial "Flow Over a Rotating Wheel with Moving Ground".

I believe you need to assign a moving reference frame (MRF) to your flow volume - for a tutorial see "Rotating Paddle using Moving Reference Frame".

If I haven't interpreted your simulation correctly please share a picture.

Use a Moving Reference Frame (MRF)

Perhaps I was trying to over simplify the problem by first only modeling a single channel. The actual geometry of the final result is as shown in

All of the walls in this model rotate about the Y axis. The face perpendicular to the Y-axis on the cylinder is the inlet and the exit faces on each of the ten radial rectangular channels are the outlets.

I will again review the MRF tutorials and will attempt a solution with the complete model.

Thank you for your quick response and valuable suggestions.

Validate with simple model using MRF

Given the image you shared - MRF is the only way in Caedium to model the channels.

A simplified single channel model is a good place to start. I would encourage you to validate MRF with your original simplified model. If successful then step up to the full model.

Forced Vortex Model

I have been able to model the radial channels with through flow and obtained reasonable converged results for the stationary case. I would like to now model a rotating case similar to the mixer tutorial.

Before that, however, I have been trying to model a simple forced vortex, with and without through flow, as a way of stepwise learning but have not achieved good results. The first approach is to have a simple cylinder that rotates about the Z-axis. If I understand the moving reference frame correctly, two volumes are needed, one stationary and one rotating. How do you suggest I configure the stationary volume in order to demonstrate a simple forced vortex within the cylinder?

MRF can be assigned to a single volume

You can assign a MRF to a single volume. The rotation is felt by the walls. You will also have to assign a Condition->Angular Velocity to the inlet and set the BC on the inlet to Physics->Inlet->Type:Default->->U = Fixed Value, U = [0 0 your-z-velocity].

If you purchase a subscription to Caedium Professional I can provide one-on-one support and help you setup project files as needed.

MRF can be assigned to a single volume

Thank you for your prompt response.

I was able to achieve good results for several angular velocities and various inlet velocities after your valuable suggestions.

Trending Now