Learning Forum

[pcaicedo]

Hi everybody,

I am using MRF to simulate a turbine into a big domain. I have created a rotor and stator zones and after applying interfaces, the software creates automatically two walls per each interface. The summary of my understanding and what I have done is the following

1) Assigning a rotational velocity in the rotor zone

2) Creating normal interfaces at the inlet and outlet of the rotor zone

3) 4 different walls are created. According this user manual (https://www.sharcnet.ca/Software/Ansys/17.0/en-us/help/flu_ug/x1-2670007.4.1.html) and some tutorials, these walls shouldn't be created or if created one should assign them as interior (I don't have that option in the BCs). I have meshed my geometry in Pointwise, but I don't think that's the problem because I can assign the BCs directly in that software and then import to Fluent. The domains have exactly the same area.

I'd appreciate any suggestion,

Thanks,

Paul

 

 

[Keyur Kanade]

Hello Paul, 

 

If your interfaces are exactly overlapping, you can ignore walls created. In this case these walls will be created but those walls will be empty. If you try to display them then you will find that nothing is displayed. 

Usually if interfaces are not exactly overlapping then there will be two zones: the overlapping part will be interface and remaining part will be wall. 

Hope this helps. 

If this is the solution please click on 'Is Solution' option so that it will help others on forum. 

Regards,

Keyur

[pcaicedo]

Thank you very much for your quick answer Keyur,

Yes, I can't select the surfaces. I was wondering if that was the problem because even at very low angular velocities (0.1 rpm) I got negative torque, and at little higher (4 rpm) reverse flow.

My case uses solar radiation to accelerate airflow. I have simulated the flow without MRF and I got around 950 kg/s with an outlet velocity of 11 m/s approx. So, I shouldn't have that physical behavior. Do I need to upload the case to get support; however, it is confidential. Or, do I need to present all the inputs here to have the concept complete?

Regards,

Paul

[Karthik R]

Hello,

Are you suggesting that the interfaces being created in Fluent are not the right ones and it is not allowing you to move forward? Did you create your geometry in ANSYS?

Thank you.

Best Regards,

Karthik

[pcaicedo]

Hello Karthik,

According the explanation of Keyur, the interfaces are not the problem and it makes sense, so it should be the physics involved. I am doing something wrong in setting up the case. What is your opinion about my second comment? Do you think is better if I create another post with all the details?

Thanks

Paul

[Karthik R]

Hello,

On your second comment, you might have to provide us with a little more details about your physical model. If you have something confidential, you might perhaps want to think about a test problem you could explain with. If it is completely different from your original question, you might want to create a new post / thread.

Hoping that helps.

Thank you.

Best Regards,

Karthik