Copper vs Aluminum vs Ect.

[ryansoh3]

Diamond heatsinks anyone?

However, not so sure about its specific heat capacity, which is more important in this case.

Cheers! :beer:

 

[BShanahan14rulz]

Why no one is trying to make a sink out of aluminum oxide.

What do you mean? Plenty of people use alumina on the surface of their heatmass because of the better heat properties of alumina compared to "bare" aluminum... All of my flashlights and lasers are alumina coated.

 

[Encap]

What do you mean? Plenty of people use alumina on the surface of their heatmass because of the better heat properties of alumina compared to "bare" aluminum... All of my flashlights and lasers are alumina coated.

Have a look at a thermal conductivity chart of materials.

Alumimum Oxide has poor thermal conductivity and is not a good as a heat sink material

 

[BShanahan14rulz]

I'm not talking about heat conductivity or specific heat capacity. I agree, a heatsink made completely out of alumina would not perform as well as aluminum.

 

[TheDukeAnumber1]

I'm not talking about heat conductivity or specific heat capacity. I agree, a heatsink made completely out of alumina would not perform as well as aluminum.

Yeah we get it your talking about emissivity which depends more on the color and the finish surface than most anything else. If you have some useful data then please present it.

 

[BShanahan14rulz]

All I'm saying is that if you don't think about radiating that heat, or removing it from the system some other way, then you will still see temperature increases at the junction. Not that you wouldn't otherwise, but as most of these threads conclude, it depends on your application. I would rather have an aluminum head coated in either natural oxidation or purposely oxidized (anodized), rather than a tarnished copper head. If this part were simply to hold the heat, or spread it out and pass it on to the radiative part of the heat system, and I didn't mind cleaning the heatsink from time to time, or actually took the care to not let its surface deteriorate, I'd go for copper too.

Emissivity Coefficients of some common Materials < interesting data.

 

[TheDukeAnumber1]

All I'm saying is that if you don't think about radiating that heat, or removing it from the system some other way, then you will still see temperature increases at the junction. Not that you wouldn't otherwise, but as most of these threads conclude, it depends on your application. I would rather have an aluminum head coated in either natural oxidation or purposely oxidized (anodized), rather than a tarnished copper head. If this part were simply to hold the heat, or spread it out and pass it on to the radiative part of the heat system, and I didn't mind cleaning the heatsink from time to time, or actually took the care to not let its surface deteriorate, I'd go for copper too.

Emissivity Coefficients of some common Materials < interesting data.

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

This is from the reference you just used.
Radiation Heat Transfer
q = σ T4 A (1)

where

q = heat transfer per unit time (W)

σ = 5.6703 10-8 (W/m2K4) - The Stefan-Boltzmann Constant

T = absolute temperature Kelvin (K)

A = area of the emitting body (m2)



q = ε σ T4 A (2)

where

ε = emissivity of the object (one for a black body)

XXXXXXXXXXXXXXXXXXXXXXX

 

[Blutquell]

 

[TheDukeAnumber1]

That pic literally made me lol.

 

[Cyparagon]

You screwed something up. Two things, actually. You're not accounting for the energy the object receives from the surroundings, and you forgot that fourth power.

(5.6703*10^-8)(300-20)^4 = 350 watts of heat

A more typical value of 20C above ambient would be 9mW of heat, so your point is still valid.

 

[grainde]

You screwed something up. Two things, actually. You're not accounting for the energy the object receives from the surroundings, and you forgot that fourth power.

(5.6703*10^-8)(300-20)^4 = 350 watts of heat

A more typical value of 20C above ambient would be 9mW of heat, so your point is still valid.

Hang on if you use the first equation at 30 deg C or 303 K with ε = 1 and Ac = 1 m^2

q = ε σ T^4 A you get: (5.6703*10^-8)(303)^4 = 477.9 W

Which is basically what Cyparagon said, although he calced it incorrectly. If however, you you use the second formula he found on that page to account for the surroundings and also using ε = 1 and Ac = 1 m^2:

q = ε σ (Th^4 - Tc^4) Ac you get: (5.6703*10^-8)(303^4 - 293^4) = 60W ! Which is relevant!

You must use Kelvin in the calculation and the formula was K2^4 -k1^4 not (K2-K1)^4

:beer:

Edit: All the same it seems a little high, guess you cant really use an emissivity of a black body for the calc...

Using 0.2 - 0.31; which was given in a table for the emissivity co-eff of some common materials ie Al heavily oxidised

We get 0.31 x 60 = 18.6 W or 12 W for 0.2; which would seem much more reasonable and very relevant!

BTW polished Al has a co-eff of only 0.039 - 0.057 suggesting the radiative emission would be considerably worse for a highly polished host, which has already been stated.

 

[TheDukeAnumber1]

Thank both of you for cleaning that up, one more correction though. Area isn't even close to 1m^2.

(5.6703*10^-8)(303^4 - 293^4)(.07) = 4.2W

I was being quite generous at .07m^2
Still higher than what seems like it should be, I still wonder if there is something we are missing.

 

[TheDukeAnumber1]

Here are two sources of experiments done with varying surface finishes. One source from an individual and one on behalf of a company.

http://www.dfrsolutions.com/uploads/white-papers/WP_Heatsinks.pdf

http://www.molalla.net/members/leeper/coatbar.htm

From our own calculations and these sources it looks like the surface finish does make a notable difference in temperature even at the handheld level of application.

 

[Blutquell]

atleast five characters

That pic literally made me lol.

Oh and very interesting calculations. op:

 

[Multimode]

.......will someone buy this software and give us the answer please :friend:
ThermoAnalytics : RadTherm: Professional CAE Thermal Modeling and Heat Transfer Software

 

[TheDukeAnumber1]

I think we cleared most of the issues up. What's your question?