[Noblek330]

Hello,

I am looking for some guidance with respect to an ANSYS fluent simulation. Attached pictures demonstrate my problem. Description follows below.

I am attempting to predict the performance of a fan where there are some obstacles immediately upstream of the fan (transient sliding mesh is employed). For now, I am using a very simple geometry until I get my arms around solving these types of problems. Long term goals are to estimate total pressure drop of air and also to investigate the effect of changing the blade angle on the air flow and pressure drop.

The model geometry consists of four bodies, a fan hub and blades (solid), fan bell (solid), fan screen (solid), and air enclosure with a boolean to subtract the aforementioned solid entities. Ultimately the other solid zones will be omitted, there is no real use in including them, I keep them now only so that I can see the relative mesh motion easily. These bodies are all combined into one part in the design modeler so that the interfaces may be automatically defined by ANSYS.

The model is transient, mesh motion is enabled for the fan cell zone with a rotation of 100 RPM. The interface between the solid fan blade and the fluid air domain is evidently two parts. The wall which is adjacent to the fan cell zone will rotate as expected while the "shadow" which is adjacent to the air cell zone stays put. I have enable mesh deforming for the air cell zone and have tried playing with the shadow as either a stationary or a moving wall but have not succesfully seen the motion that I am looking for (i.e., both components of the interface should rotate together).

First, is what I am looking to do possible with only ANSYS fluent?
If so, how can I proceed from where I am?

Thanks and regards,
Ken

[ghost82]

Hi,
I have 2 questions:

1- Why do you have shadow walls?If you're modeling the fluid flow and not the solid zones you shouldn't have shadow walls; if your goal is to only study the fluid flow go back to the pre-processor and delete solid zones.

2-Why are you using dynamic mesh?Apparently your mesh doesn't need to be reconstructed (re-meshing or layering for example); you should be able to treat the problem with moving mesh (in cell zone conditions tab) and by setting absolute/relative velocities to the walls.

Then, you will have the "fan" zone that will rotate, with its interface(s) and fan walls, and the outer stationary zone, with its interface(s) (adjacent to the interfaces of the fan zone) and stationary walls that will be fixed.
The fan zone must be a cylinder to have a proper interface to be able to rotate sliding the stationary interface.


Daniele