Maximum temperature inside of an enclosure?

[al124]

I'm currently doing a research project that involves a bunch of electrical components that generate a certain amount of heat in a cast-iron explosion proof box. There is a sensor inside of the box that has a maximum operating temperature of 113 °F, and we are planning to deploy this device out in 100 °F weather. The box will be covered with a reflective tarp, so I want to neglect heating from the sun in my estimate. We may also be able to use a large industrial fan to cool the outside of the box. We are assuming that we are operating the device for 8 hours. How do I go about calculating a rough estimate of the maximum temperature inside the enclosure? I know the dimensions of the box and the maximum power loss of each of the components inside of it. Any help, even a start is appreciated! Thanks!

 

[IRstuff]

Well, you've already bounded the problem. 13°F of temperature difference between the air and your component. Since you know the total power, you just need to look at the conducted, radiated, and convected heat.

Which isn't very much to play with btw.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[itsmoked]

I predict abject failure.

All you need is a silver dollar sized beam of sunlight to land on the enclosure and it's over 113F.

If you can run a fan on it is the fan explosion proof?

Do you have compressed air available?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[KEVINZY]

Agree with IRstuff. You should calculate whether do you need a cooling fan or a refrigeration. Depend on the power loss.

 

[itsmoked]

Hello?? Anybody there?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[LittleInch]

Well let's start assuming you're talking about steady state. Your CI box will be at the ambient air temp 100F.

To loose any heat from the inside it needs to get to a temperature higher than 100. You need to consider the energy required to heat up your CI box by say 10.

If your internal heat load is low then you might get lucky.

If not figure out the heat loss from the surface area of your box with a 10F difference. If that doesn't work then you're into fans and fins.

How much hear power are we talking about?

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

 

[hendersdc]

A while ago I made a spreadsheet for simple first order approximations like this. If you know the total power in the box, the outer dimensions, and the ambient temperature I can give you the outer surface temperature of the box assuming only natural convection. With this approach I avoid all the necessary detail about materials, placements/dimensions and individual components. If, in this case, the outer surface temperature of your box is greater than 113F then it doesn't matter what the temperature of the sensor is because it would be greater than that.

For example, a 20cm cube in 100F ambient temp that is dissipating 20W will have a surface temperature of about 134F (assuming no hot spots and uniform surface temp) and neglecting incident radiation and any heat conducted through the enclosure mounting.

 

[itsmoked]

hendersdc; Please read my dissertation on the subject in thread248-340558. You should correct any spreadsheet to the only heat dissipation being from the TOP horizontal surface that's 3 inches bigger on all four sides than actual top surface of the enclosure. There is NO dissipation from the rest of an enclosure's surfaces.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[hendersdc]

Keith,

I agree that the spreadsheet doesn't accurately represent reality. It is however very useful for first order approximation to understand if more detailed, and accurate, analysis is justified.

In the case of the OP, with such a small delta T between what the sensor can operate in and the ambient temp, this approach will likely very quickly show that a new sensor needs to be selected or some method of active cooling is needed.

 

[itsmoked]

I'm just saying your spreadsheet will be considerably more accurate by at least a factor of three if you consider your enclosure to be only the horizontal top and only the top two inches of the four vertical sides or the top three inches mashed flat along with the top as one horizontal plane.

As for the OP where'd he go??

Keith Cress
kcress -

http://www.flaminsystems.com

 

[IRstuff]

Where did that come from? This simple has a limiting value for the transfer coefficient, not a dead stop at 3 inches

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[itsmoked]

Where did that come from?

From my experience building enclosures stuffed with electronics for power plants, oil fields, dams, water filtration systems, and SCADA systems.

That is a rough estimate as I'm talking to hendersdc's quickie spreadsheet. If you want more accurate, which he didn't need, then you have to take a bunch of other things into account. As an example the back does no cooling.

Some boxes have hot/warm sides down an inch, some down 4 or five inches, most about 2 or 3. The rest of the box surface does almost no cooling what-so-ever.

Keith Cress
kcress -

http://www.flaminsystems.com