# Need advice on how realistic these CFD results are. Intuitively they seem way off.

Tags:
1. Apr 1, 2012

### TJ_Laser

I am using Solidworks to model the pressure distribution of air around a large fan. The fan is 10' in diameter and moves 94,000 CFM. I need to know the effect adding a large duct around the inlet of the fan would have on the pressure. If the duct causes a huge pressure drop then we can't use it.

The duct would completely enclose the inlet and extend 18'. After modeling the air flow Solidworks tells me the lowest-pressure point in the entire duct would be only 0.4 psi below atmospheric pressure. It seems to me that a fan of this size pulling air through 18' of duct would cause much lower pressure than that.

I have added some pictures of the model and results. Air enters through the rectangular inlet shown in the two pictures of the model. It exists through the opening, which I have modeled as a red disk. As you can see the pressure inside the duct is barely below atmospheric pressure. Intuitively it seems like it should be much lower. As a new engineer with little experience my intuitive frame of reference may not be realistic. Any thoughts?

The length units in the attached pictures are in inches.

File size:
17.8 KB
Views:
115
File size:
12.2 KB
Views:
102
File size:
38.8 KB
Views:
143
2. Apr 2, 2012

### AlephZero

Re: Need advice on how realistic these CFD results are. Intuitively they seem way off

Using Bernouilli's equation (assuming the air is incompressible) will give you a sanity check on what pressure changes to expect.

3. Apr 2, 2012

### TJ_Laser

Re: Need advice on how realistic these CFD results are. Intuitively they seem way off

We did that and found a much higher pressure difference than that predicted by Solidworks. I'll just have to go back to the model and see if I made any mistakes.

4. Apr 3, 2012

### bill nye scienceguy!

Re: Need advice on how realistic these CFD results are. Intuitively they seem way off

Your first mistake was to use Solidworks.

In my experience (admittedly a few years ago now) Solidworks is workable if you want to compare flow paths on, for example, different iterations of a model but if you're after an absolute value don't bother.