Fan Simulation Resulted in Pressure Loss Instead of Gain

[hooshsim]

Ran a full model of an axial fan using SimpleFoam (FC/CfdOF, OF v2206) with the following BCs:

1. Inlet flow, outlet pressure
2. Inlet pressure, outlet flow

In both cases, the results showed low torque but reasonable value, and pressure loss across the model. At specified fan speed I was expecting a pressure gain instead of a loss. Has anyone experienced similar behavior? Suggestions welcomed.

Updated fan patch in 0/U file:

Type rotatingWallVelovity;
Origin (0 0 0);
Axis (0 0 0);
Omega 430;

[Raedchen_im_System]

Updated fan patch in 0/U file:

Type rotatingWallVelovity;
Origin (0 0 0);
Axis (0 0 0);
Omega 430;

Hi hooshsim, it would be better if you upload your FreeCad-Model for review.

The Axis in your model has the length of 0, it should have the length of 1. Example: If your model is rotating in the x-/y-plane, then the axis of rotation should be (0 0 1) or (0 0 -1). The axis of rotation must always be perpendicular to the rotation.

In your model, the origin of rotation is (0 0 0). Please check if your model is build around the zero-coordinate of the global coordinate system.

[hooshsim]

Updated fan patch in 0/U file:

Type rotatingWallVelovity;
Origin (0 0 0);
Axis (0 0 0);
Omega 430;

Hi hooshsim, it would be better if you upload your FreeCad-Model for review.

The Axis in your model has the length of 0, it should have the length of 1. Example: If your model is rotating in the x-/y-plane, then the axis of rotation should be (0 0 1) or (0 0 -1). The axis of rotation must always be perpendicular to the rotation.

In your model, the origin of rotation is (0 0 0). Please check if your model is build around the zero-coordinate of the global coordinate system.

FC model file is too large to upload (2.6 GB). How can I reduce the size? Any other options?
My bad, the AOR was a typo here. It’s correct in the model file

[Raedchen_im_System]

FC model file is too large to upload (2.6 GB). How can I reduce the size?

The FreeCad-file is "blown up" by the mesh. If you want to upload your model, you must reduce the size of the file by removing the mesh. The mesh is removed by pressing the Delete-key on the mesh:

RemoveMesh.jpg (39.39 KiB) Viewed 1425 times

[hooshsim]

Model file with the removed mesh still exceeds the FC upload limit of 1MB. . Is your personal email an option to send it? You may reply to my email cfdpro2000@yahoo.com.

Placed the model file in DropBox

https://www.dropbox.com/s/chsmxg6jnuj92 ... FCStd?dl=0

[hooshsim]

I’m very surprised nobody has replied to this thread. Does that mean nobody is using OF for fan simulation? Moderators can at least make suggestions. Thanks

[thschrader]

I’m very surprised nobody has replied to this thread. Does that mean nobody is using OF for fan simulation? Moderators can at least make suggestions. Thanks

My theme is wind on buildings.
This is an interesting thread, but at the moment I have a lot of statics stuff to do.
No offense intended because of no reply...

[Raedchen_im_System]

Sorry, I did not have the time and the motivation to let your example run. In general, I can recommend that you try out models with few elements and simple geometry as a first step in order to gain experience. I made a video about the opposite case - wind turbine. Did you watch this video?

You have a very high speed, 430 is in rad/s, which converts to a speed of 4106 rpm, is that what you intended?

In general, you should check the direction of rotation, the direction of rotation is specified by the direction of "Axis" and the value "Omega". The easiest way to check the direction of rotation is in Paraview, where you look at the direction of the speed near the wings. If you vary the speed of the rotor, you will calculate different torques. If you enter the speeds and torques in a diagram, you will get a parabola. My advice: Check if you were in an unfavorable RPM range.

In my tutorial about wind turbines, I set the speed as the boundary condition on the inlet side and the ambient pressure as the second boundary condition on the outlet.In your case, the boundary conditions are not so easy. My assumption: The task of a fan is to accelerate air, the pressure changes outside the fan area should be rather small. I would specify the pressure as 0 mPa as a boundary condition on the inlet and on the outlet. It would be interesting if you could provide feedback on whether this constraint worked in your model.

[hooshsim]

Sorry, I did not have the time and the motivation to let your example run. In general, I can recommend that you try out models with few elements and simple geometry as a first step in order to gain experience. I made a video about the opposite case - wind turbine. Did you watch this video?

You have a very high speed, 430 is in rad/s, which converts to a speed of 4106 rpm, is that what you intended?

In general, you should check the direction of rotation, the direction of rotation is specified by the direction of "Axis" and the value "Omega". The easiest way to check the direction of rotation is in Paraview, where you look at the direction of the speed near the wings. If you vary the speed of the rotor, you will calculate different torques. If you enter the speeds and torques in a diagram, you will get a parabola. My advice: Check if you were in an unfavorable RPM range.

In my tutorial about wind turbines, I set the speed as the boundary condition on the inlet side and the ambient pressure as the second boundary condition on the outlet.In your case, the boundary conditions are not so easy. My assumption: The task of a fan is to accelerate air, the pressure changes outside the fan area should be rather small. I would specify the pressure as 0 mPa as a boundary condition on the inlet and on the outlet. It would be interesting if you could provide feedback on whether this constraint worked in your model.

Thank you for the reply. Appreciate the time and the effort.
FYI, this model was simulated in a commercial CFD with identical boundary conditions. The pressure gain at 4100 rpm and 1 kg/sec mass flow, was about 253Pa. However, the FC/Cfdof results in 750pa loss.

I started building this model after watching your wind-turbine tutorial online…an excellent video.

The direction of rotational speed is correct because the predicted torque value is close to the one obtained using commercial software. My objective is to predict the fan performance curve (flow vs. P).

I just submitted a case based on your recommendation by specifying pressure at both inlet and outlet. At 300 iterations the torque and the inlet velocity values are too small and not changing much.

This model is very similar to the one used in your tutorial. The only difference is my model has a 19mm thick shroud at the fan tip. In Paraview the velocity distribution seem to be correct. At 500 iteration the torque and the inlet velocity don’t seem to change significantly. Therefore I stopped the run. Do you have any other suggestions?

I have another model of this case which represents only one blade passage. The cyclicAMI symmetry is not working at all. I’m stock here. I can upload the cyclic model if interested. Thanks