Hello,

I'm using Cosmos FloWorks to select a fan for an enclsoure I'm designing. Thus far, I've built the pressure drop curve for my enclosure by specifying different volume flow rates at the fan inlet (5CFM, 10CFM, 15CFM, etc...). This resulted in what I consider a reasonable curve (~0 in H2O at 3CFM up to .36in H2O at 30CFM). Based on the required temperature rise, I determined that 20CFM was needed for my enclosure.

I went to a major fan manufacturer and found a fan that would supply about 20CFM at the corresponding back pressure from my enclosure pressure drop curve (~.16in H2O). Then, I built the fan curve in FloWorks and used that to validate my selection.

The results of that simulation were mixed... The temperature rise was within a degree or two of the simulation I ran just using a 20CFM volume flow. However, the pressure was a good deal higher (0.3 in H2O vs. the 0.16in H2O from the 20CFM simulation).

Does anybody have any ideas as to what might be happening? Shouldn't I be seeing about the same pressure using the fan? All of the simulations were run at a 101325 Pa environment pressure and an ambient temperature of 50 Deg. C. I took the static pressure results from the same face that I specified the volume flow/fan.

Any help is appreciated

Are the velocity profiles used at the inlet the same in both cases (i.e. volume flow rate has constant velocity, while the fan model has swirl)?

Erich,

I didn't model the fan with swirl... Just with flow normal to the inlet face (which I believe is constant velocity). The volume flow rate was modelled the same.

I don't know if FloWorks allows users to do much more with the inlet velocity profiles.

Are you familiar with FloWorks?

Thanks!

Mike,

Sorry, I am not familiar with FloWorks, we use CFX. I would guess the fan manufacturer could provide you with a velocity profile. Your support provider should be able to guide you with its use.

Good Luck,

Erich

I just started learning and using Floworks.

All or most real life axial fans have "swirl" - a tangential velocity component in addition to the expected axial component at the outlet of the fan. There are numerous experiments showing this( which I can site if want to know ), and if you do a CFD simulation using Fluent, CFX and maybe even Floworks itself on the fan, it would or should show up. Thus, if you want somewhat realistic results, try to incorporate swirl. But as far as getting actual detailed data from fan manufacturers on the outlet velocity profiles of these fans ( for even one flow rate ), good luck! They just don't have it or don't care to give it out. I've already tried many times. Well, one Japanese company MAY have the data; I'll have to call them again.