# Find alloy composition by density

I have a square piece of metal from a processor heat sink that looks a lot like copper. I'm trying to find the alloy composition but I'm not sure how.

I happen to have both a digital caliper and a milligram-precision scale. The piece has a rectangular shape so volume wasn't hard to find. The volume is 5.378 cm³ and the mass is 52.395 g, that gives a density of 9,27 g/cm³. The density of copper is 8.96, nickel is 8.91, zinc is 7.14, lead is 11.34.

If I use this formula (which may be wrong):

$$p = a \cdot p_a + b \cdot p_b$$

$a$ being the volume fraction and $p$ the density. The density is higher than that of copper, this means there has to be a metal with a density higher than copper in the alloy, maybe lead? Using the formula I get 87 % copper and 13 % lead by mass, looks a bit too much.

Is this a good way to find the alloy or an alloy doesn't have a density or volume proportional to that of its components?

• (1) Densities aren't additive like that (i.e. following ideal behavior) except in small ranges or unusual cases. (2) You have to consider the error in your measurement which was crude. (3) Think the most common compositions for electronic heatsinks would be copper or aluminum alloys. – MaxW Jun 27 '18 at 21:37
• @MaxW Alright that makes sense. Would an alloy made of denser metals (eg: with lead) always be denser than one made of a less dense one (eg: aluminum) if mixed in the same volumes? – Nicolas Jun 27 '18 at 21:40
• You'd have to check the phase diagram for the alloys in question. Superlatives like always and never make me wonder. There always seems to be an exception. – MaxW Jun 27 '18 at 21:43
• Besides, not every pair of metals forms an alloy. Cu/Pb, for one, does not. – Ivan Neretin Jun 27 '18 at 21:44
• No, it is not a good way to find the alloy. The best you can do is dissolve a piece of the metal and analyze the cations. It is also not a good way to find the density. You should use the method of Archimedes: submerge the piece in liquid and measure the volume increase. The density is probably wrong. Most alloys including SS are below 9. The closest I have found is Hastelloy with 9.245, but which is obviously not your material: does not look like copper, has a low heat conductivity (10 W/m.K against 400 of Cu and 200 of Al) and is very expensive. – Raoul Kessels Jun 28 '18 at 9:29