# Where can I learn to calculate the need for cooling?

1. Mar 10, 2009

### TSN79

I'm on the lookout for a web site dealing with calculation of cooling effects in deep freezers (for food storage and such). The main goal for me is to learn the basics of how to find the needed cooling effect from input values like amount of food, required temperature, room size, insulation, etc. If anyone know any good places I'll be happy :)

2. Mar 10, 2009

### Naty1

You could also get empirical information off the back of any freezer...or the owners manual for one....

One thing I can tell you in general: heat loss/gain is directly proportional to the difference in temperature between the ambient and refrigerated area....so keeping something at zero degress in 80 degree ambient takes twice as much cooling (power) as in 40 degree ambient.

3. Mar 10, 2009

### alxm

Saywhatnow?
Stefan-Boltzmann's law:
$$j^{\star} = \sigma T^{4}$$

Does that look linear in T to you?

4. Mar 10, 2009

### Staff: Mentor

That's only for radiation and a refrigerator is almost entirely conduction and convection and is linear.

5. Mar 12, 2009

### alxm

I know it's for radiation, and at least where I learned to do these kinds of calculations, radiation losses were never neglected.

Conduction is generally treated as if it was linear, but it is not. It's an approximation. The same goes for convection, only it's a worse approximation there.

Using a linear approximation is one thing, stating that reality is linear is something else.

6. Mar 12, 2009

### Staff: Mentor

That's surprising since radiation really is insignificantly small for this case. A tiny fraction of a percent of the heat transfer.
I've never heard that before - could you provide a source? Here's the typical description: http://hyperphysics.phy-astr.gsu.edu/Hbase/thermo/heatra.html
Convection is only approximately linear, but in a situation where you have a lot of insulation, the conduction through the insulation is by far the dominant mode of heat transfer. So again, other nonlinear modes are insignificantly small.
I think you are being overly pedantic. When someone is looking for 95% accuracy and an effect is .0001% (estimates), it can safely be ignored. This is an engineering question and one of the more important skills in doing an engineering problem is figuring out what you need to consider and what you can safely ignore. If an engineer tries to consider all effects - no matter how small - in all cases, they'd never finish solving a problem.