Confused - Fan pulling air through a heat sink

[CnoJ]

Hi,

I have a cooling application that needs fans downstream of a heat source. When I try to account for the air density change (hot air at fan) I run into this problem:

-A fan pulls air through a heat sink.
-Constant heat is applied to the sink. Its exhaust temp rises and air density drops.
-As the fan has a constant volumetric flow rate, it's mass flow rate is reduced.
-As mass flow rate through the sink is reduced, exhaust temp rises, air density at the fan drops and so mass flow rate drops...

This seems like a runaway condition with nothing to stop the heat sink temp from continuously rising. Is there something I'm missing?

Thanks.
Jon C

 

[LittleInch]

I think so. The density and temperature of the air entering your heat sink stays the same.

So long as you calculate what air flow you need into the heat sink and then allow for the change in density then you will be ok. Initially thet will be too much air but then over time you will achieve steady state providing your fan is sized properly.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[hendersdc]

Assuming the air temperature rises enough to significantly change the air density then the fan will not have a constant volumetric flow rate. The pressure drop over the heat sink will reduce as the air density drops, causing the fan operating point to move along its operating curve and increase the volumetric flow rate.

 

[CnoJ]

OK. So for fan choice my system curve should use CFM exiting the sink?

 

[hendersdc]

Your fan operating point (how much CFM it is moving through itself) will be determined by how much pressure loss it needs to overcome. The pressure loss will be determined by both upstream and downstream conditions. In this case, upstream of the fan you will have a heat sink that restricts the flow. You have not provided any detail for what is downstream of your fan but things there have the potential to affect the fan as well.

Do not select a fan based on peak rated CFM, you will not get that much flow out of it because your system will have some amount of pressure loss to overcome.

 

[LittleInch]

Can you get the air flow to go vertically up?

That would help a lot. ..

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[chicopee]

Have you been able to develop a heat transfer calculation between the heat sink and the air flowing around it? The coefficient of convective heat transfer will be the hardest value to determine.

 

[CnoJ]

Well I already know the max and min kg/s I need to pull thru the sink, the kW I'm sinking, and I have system curves for the sink and for the enclosure. So now I have have to find a fan that will pull the required kg/s.
I'm thinking I should increase the sink system curve CFM by the volume increase from heating, before combining it with the enclosure curve and matching fan curves to it. Is that the correct way to go?

 

[chicopee]

The intersection of the system curve and the fan performance curve will only indicate the fan air removal in CFM for a particular temperature at a particular pressure drop. That CFM for the incoming air can the be converted to a mass value. The convective heat transfer between heat sink item and air should equal the increase in internal energy of the air flow. An increase in the internal energy of the air flow means a reduction in the air density. There should be no increase in the heat sink temperature unless you have underestimated the coefficient of convective heat transfer or the air flow pattern around the heat sink. Your air flow rate appears inadequate and you can only determine its adequacy by doing the heat transfer and internal energy calculations.

 

[hydtools]

Would it help to change perspective to that considering atm pushing air across the heatsink?

Ted