Fan Study

[Anonymous]

Hi

I am attempting to optimize a fan by changing the geometry of the blades. I found a tutorial on centrifugal pumps and figures that I would set up something similar, just focusing on a fan. When I try to run I find that I am lacking a license for the motion solver. Is there another way to model this problem? Or perhaps it is time to upgrade the license.

Thank you

Dylan

[Jon.Wilde]

Hi Dylan,

You would need motion for this, yes.

Check out the guide here too, might be useful 🙂

Thanks!

Jon

[Anonymous]

Thank you for this.

Now I have a second question that is about modeling an entire airflow system that includes a fan within it. I have an internal fan with prescribed fan curve, then I also have the inlet and outlet to the air handler. Is it appropriate to model the inflow and outflow both as pressure BCs with P = 0? I have also included a resistance region within the system that would allow me to modify the amount of resistance (and therefore the system curve) until I get the CFM that I am trying to model. Does this sound reasonable?

Thank you

Dylan

[Jon.Wilde]

Hi,

This sounds reasonable - although could you not apply a cfm at the inlet with a P=0 at the outlet to give you the desired cfm?

I am not entirely clear of the goals here, just a thought.

Thanks,

Jon

[Anonymous]

Ok, so this is how I thought about modelling this and other problems that include a fan for a sample problem such as the one I have attached here.

1) P = 0 at both inlet and outlet. Model the fan with a fan curve. In this case the fan is actually inducing the flow rate through the air handler.

2) P = 0 at outlet. Flow rate at inlet. Model the fan with some resistance. This would be because I am assuming the fan is inducing the flow rate seen at the inlet.

3) Something else.

What is the recommended method on solving a problem such as this?

Thanks

Dylan

[Jon.Wilde · Accepted solution]

Hi Dylan,

Typically we would go with option 1 here. Be sure to apply a good uniform mesh on the fan - at least 4-5 elements from inlet face to outlet.

Kind regards,

Jon

[Anonymous]

So, I have set up this problem and am trying to run it with both inlet and outlet to be at P=0. I assumed that the internal fan will set the fan curve while the system curve could be adjusted by some internal resistance between the inlet and the fan. The idea is that I could then adjust the resistance value in somewhat of an iterative manner so as to arrive at a flow rate that I am looking for (around 50,000 CFM).

Unfotunately I am getting flow rates that are high (2-400,000 CFM) and inelastic to resistance value adjustments.

Is there any advice that you can offer to me in the way I am modeling this problem?

I have attached the .cfdst file

Thank you

Dylan

[Anonymous]

Thank you for all of your help but it looks like I have actually solved this problem. It turns out I was using a fan curve for all 9 fans as opposed to just a single fan. After making the adjustment, I am getting results that are reasonable.

Thanks again!

Dylan

[Jon.Wilde]

Great 🙂

________________________________

Autodesk Limited
Registered Office: One Discovery Place, Columbus Drive, Farnborough, Hampshire GU14 0NZ
Registered in England and Wales, No. 1839239