Heat disipation, rectangular enclosure - natural convection

[Anthonycrf450r]

Hi All,
I have been working on how to calculate the maximum heat load allowed within a small rectangular alloy enclosure subject to natural convection only. I have calculated the heat transfer coefficient's for each of the internal and external surfaces (these are just plain flat surfaces at this point in time).

Can someone clarify how to combine each of the internal and external surfaces into an average transfer coefficient for each (internal / external)????? My two thoughts are;
1. Obtain the average h by multiplying each coefficient by the percentage of surface area the correspond to and then sum all values to obtain the total.
2. or can the sides / top / bottom of the enclosure be turned into a parallel resistance network and then the total R obtained.

Essentially what i am trying to do is turn an enclosure with 6 internal heat transfer surfaces, and 6 external heat transfer surfaces into a standard heat transfer calculation through a wall (1 internal surface, 1 external surface) based on the max allowable internal and external air temperatures.

Thanks

 

[boyze]

depending on the dims of the box and heat sources the internal flow patterns may not behave like free air flat surfaces. If open ended and wide and deep enough then, probably. But if narrow, then flows will combine. If capped off then internal recirculating flows are possible. There's also the possible complexity of constant heat flux vs constant T influence on the hc's. Then there's the heat source location issue and so on.

Tell us more.

Have Fun!

James A. Pike

http://www.xl4sim.com

http://www.erieztechnologies.com

 

[Compositepro]

For example, there will be relatively little heat transfer out of the box at points that are below the level of the heat source in the box, unless there is a fan.

 

[Anthonycrf450r]

Yeah I understand the true affects of what happens in the box is extremely complicated. It is a completely sealed enclosure and the size relative to the internal components is such that you could assume that heat flux is distributed evenly inside the enclosure.

The driving parameter for design would be the maximum internal air temperature for the worst case component. eg: max temp = 80 deg C, I am trying to work out the max allowable wattage of internal components based on this internal max temp, and of course a max expected external ambient temp. So essentially you could ignore the internal components and just assume that the internal air temperature is uniform 80degrees C, then I am after the final wattage which the external surfaces will dissipate.

For the calculated internal h values, I am assuming these can be calculated the same as external... noting that this may not be true depending on temperature hotspots (and obviously done by using the correct Ra and Nu values for the internal air).

The original plan was to place a fan inside the enclosure to circulate are and reduce hot spots. However I have read that for small enclosures, the gain versus extra wattage is negligable...so at the moment this is still TBD.

Sorry if this is confusing, or if my head is screwed on backwards just let me know.

 

[boyze]

My first pass would be to assume no internal convection. Just conduction through the box wall coupled with external natural convection.

Have Fun!

James A. Pike

http://www.xl4sim.com

http://www.erieztechnologies.com

 

[Anthonycrf450r]

^^^ Boyze,
So your saying just assume the internal surface temperature of the enclosure is equal to the internal ambient air? Is that too conservative though, as this says there is no resistance between the internal air and enclosure?

Re- the original question on how to average the transfer coeficients to obtain a single wall problem...I will ignore this for now and treat it as individual walls and just sum the heat losses.