Heat Dissipation 1

[GBor]

A bit out of my league here. I have 3 electronics boxes being installed on a rack. They exhaust heat into a plenum which has a pull-through cooling system. I have the fan characteristics with CFM to static pressure. Plenum is a little deeper than 3", and we are looking to attach a 3" diameter duct. Required airflow for the three boxes is 27 CFM at 40C, 27 CFM at 40C, and 20 CFM at 55C. Static pressure drops max are 0.07 inH2O, 0.06 inH2O, and 0.13 inH2O, respectively. The boxes are each sitting on a 4MCU short-short Airinc tray and completely cover the top of the plenum. Friction due to bends will be minimal (probably negligible).

When I start to think about the restrictions caused by knocking out the plugs in the bottom of the Airinc trays, I get turned around in the analysis. Fluids book isn't helping me here...been too long and cobwebs are really strong!

Any guidance on how to balance the airflow?

 

[HANDCALC]

GBor,

Can you measure the cooling parameters you are interested in, rather than trying to calculate a theoretical answer? The "knock-out plugs" are intended to adjust the airflow to obtain the desired cooling results.

 

[GBor]

I may be able to, but we're on a pretty tight program and I was trying to get close to the right answer. We're modifying an existing installation that used to need a significantly greater amount of cooling. I'm concerned that we may collapse the flexible ducting if we don't have sufficient open surface area through which to draw air. On the "flip side", we don't have a lot of room to "hard pipe" it. The fan installation isn't lined up with our foundation and moving it so that it is doesn't appear to be an option. I think I'm making some progress understanding the fluid calcs that I have attempted to purge over the last 20 years, but if you have some guidance, I'm interested...

Thanks,
GBor

 

[HANDCALC]

Unfortunately I've only been involved with Avionics Tray and Shelf manufacture at the OEM level, where the "metering plates" at the bottom of the trays, regulate the airflow and were adjusted to match the drawing requirements.

Maybe someone else can point Gbor in the right direction?

 

[kellnerp]

Gbor,

A spec. for the equipment to be cooled might look like this: (73 lbs/hr airflow: 18 CFM, 0.20 in. pressure drop). If you have three boxes you might assume:
[ul i]
[li]The flows are minimum required to cool the equipment.[/li]
[li]The pressure in the plenum is equal at each restrictor plate.[/li]
[li]That when all the plugs are out of the restrictor plate the box sees the plenum pressure.[/li]
[li]That the equipment with the lowest pressure drop will see the greatest flow.[/li]
[li]That downstream from the fan is at the same absolute pressure as the upstream of the equipment. If not you have to add this into the pressure drop across the fan.[/li]
[/ul]
Then a first order approximation would be:
[ul a]
[li]Add up the flows required for all the equipment boxes.[/li]
[li]At the pressure drop for the box with the highest drop, without any restriction, will the fan capacity be enough?[/li]
[li]Add restriction to the other bays so that the overall drop through the remaining boxes and the plenum equals that of the box with the highest drop.[/li]
[/ul]
What is not accounted for here is that the air warms as it passes through a box.

Qout/Qin ~ (Tout + 273)/(Tin +273)

If it is 20c in the cabin or wherever the air enters the box and 40c when the air leaves the box then there is a 7% increase in flow exiting the box. Probably not significant in this situation.


TOP
CSWP, BSSE

http://www.engtran.com

http://www.niswug.org

"Node news is good news."

 

[GBor]

Thanks, kellnerp,

Fortunately, my flow requirements through each box is pretty close to the others and the restriction may just balance it out pretty well. My requirements doesn't say anything about pounds/hour.

We're building a test platform, but I will probably try to figure out what we're going to end up with just to exercise my "geekiness"

Thanks, Again!
GBor

 

[kellnerp]

There is implied in the requirements a mass flow rate because you threw in temperature. PV =nRT, remember? You probably also have to consider altitude because you don't necessarily maintain standard conditions in a cabin.

Would be fun to run this through a simple CFD analysis. Second order stuff and unexpected stagnation points.

TOP
CSWP, BSSE

http://www.engtran.com

http://www.niswug.org

"Node news is good news."

 

[GBor]

I may try that. I'm not a big CFD kinda' guy, but this could be a good opportunity to practice.

Thanks, Again, for the thoughts and assistance.

GBor