axial fan issues

[Anonymous]

Hi all.  I'm running in to a snag, and its probably due to not using this tool often enough.  

I have a electronics chassis and an internal axial fan volume modeled (cylinder for the fan blades, a solid plug, and a cylinder for the outter housing.  I have a fan curve, rpm (3400), and slip factor of 0.3, and the flow direction is axial to the fan.  While the solution is solving, it does not seem to push much into the box (axial), but instead it seems that the air is spilling tangentially.   I acutally setup my models with a volume inlet and outlet to the box openings to help with solving since the fan is at the inner surface of the box wall.  I can see good velocity through my inlet air volume into the fan, then everthing "stops".  I've added a jpg of the simulation to show what I am talking about.  

Any ideas?  

Thank you.

[Jon.Wilde]

Hi,

Do you have the fan material applied to just the one cylindrical volume yes? Is the direction of flow set to be 'y'?

Make sure that the fan has a good uniform mesh with 5 elements from the inlet to outlet face. Use Advection Scheme 5 too 😉

Thanks,

Jon

[Anonymous]

Hi John,

Yes, fan material is just the the one cylinder, and I do have ADV 5 choosen.  There is 15-ish elements before, across, and after the fan.  The flow direction is set in z dir.  You mentioned Y, but a typo, yes?

Tom.

[Jon.Wilde]

OK, so the mesh is good 🙂

This might be down to something else in the model, are you assigning flow conditions at the main air inlet and outlet as well as having the fan drive the flow?

If so, this would be overconstrained. Does just the fan drive the flow in reality?

[Anonymous]

Hi John,

I have 0 pressure conditions on the boundaries.  I have added the model for your review.  Its a dumb model, so there is no company info in it.  

Please take a look and let me know what you find.  I would like to resolve this quickly, as Monday, I have to complete the design, and the fan is the last check needed.

Best Regards,

Tom.

[Jon.Wilde]

Hi,

I wonder if this is down to the rpm. I have run a host of different models:

• Without the fan (just an inlet flow rate, just to be sure there wasn't just a strange effect happening within the geometry) - this looked OK
• With the fan, with different slip values
• With and without a fan curve (vs constant)
• Lowering the rpm

The latter seems to make the most difference. I actually have it at 0 now as a test and the flow looks more like it did without a fan (as expected).

With an rpm,  you have a lip above the fan though, it does seem like a lot of flow is exiting upwards and this is causing a strange flow pattern globally, that is encouraging this effect to take place.

I have not seen any problems with using fans before so I wonder if with an rpm, this is really what happens. Have you any real world tests to compare with?

Thanks,

Jon

[Anonymous]

Hi Jon,

The fan will spill some air tangenetally out of the opening, but it will have most of the air going forward, as a good fan should.  What recommendations do you have to set up an internal fan?  If we used 0 rpm, this describes no rotational motion of the air. Not real, but is it adquate?  Is there a issue with the rotational speed vs. the amount of mass flow?   I would expect this kind of arrangement to have about a 0.2" H2O pressure drop.  When I look at hte pressure results on the planer surface, there shows almost a zero pressure differential, through the box, which I think would suggest the highest flow rate from the fan possible.  But the fan stops pushing air forward.  Ideas?

Thank you.

Tom.

[Jon.Wilde · Accepted solution]

Hi,

It could be the low flow vs rotational velocity that is causing this, yes. I guess a workaround might be to really drop the rpm to near zero, to try to avoid it.

I can give this a go here also?

What happens if you have no fan, flow at the inlet and a P=0 at the outlet? That should give you the dP of the system, without the fan. I guess this should match the 0.2 in H20?

Thanks,

Jon

[Jon.Wilde]

Hi,

Did this work out OK in the end or is the solution still unsatisfactory?

Thanks,

Jon

[Anonymous]

Hi Jon,

I had to give up on working with the internal fan and resort to running a set of constant flow rates and ploting the system resistance curve against the fan curve.  It works, but I would like to know why the fan produced such a pattern.  

There is not any indepth information in the help menus for internal fan setups.  Only the one generic setup.  I would have thought that with all of the electronics being developed with this tool (going back to CFDesign) that there would be more of a detailed trouble shooting and detailed fan setup based on type of fans.  

Best Regards,

Tom.

[systemfanglobal_com]

Hello, we specialize in manufacturing industrial axial fans that ensure the specifications required by buyers.
Hope you can visit our website: https://systemfanglobal.com/ Thank you.