[Showseph]

Hello all,

I wanted to know the best ways to melt down metal alloys while best maintaining the composition of the alloy. We have two methods that we have attempted this with varying results including: placing samples in an enclosed furnace and melting samples in a simplified "CianFlone"-esque setup in which we use a TIG welder to melt them down. We want to melt our samples down for analysis on our optical emission spectrometer, and the alloys that we test are aluminum, titanium, cobalt, nickel, iron, and copper. I have previously used a Cianflone to melt titanium down consistently, but not for any of the other alloys that we test.

[Borek]

Perhaps I am missing something, but I don't see how you can change the composition of the mixture by melting it unless it contains volatiles or some of the components get easily oxidized. Inert atmosphere could help deal with the latter.

[Enthalpy]

Evaporation is known to change the composition of alloys. For instance if welding Al-Mg by TIG, we use a seam alloy richer in Mg to compensate its evaporation. Al in molten Ti has a similar vapour pressure, in the order of 1kPa.

Vacuum remelting is used to make expensive steel more resilient by evaporating embrittling elements.

Air might shift the amount of N and O in metals. Traces of N₂ serve to harden duplex steel while traces of O spoil magnetically soft iron. Traces of N or O embrittle Ti alloys, so welding by TIG apparatus demands a much stronger protection against air than just blowing Ar as for aluminium alloys.

The melting pot uses to add Si to steel. That's so common that we don't even mention Si amounts in steel below an important threshold.

I haven't read about W contamination by TIG welding, but since the electrode is slowly consumed, it must necessarily happen. Traces of W would be undesired in magnetically soft iron.

----------

One relatively clean means to heat metals is induction. Contamination by the melting pot but not by the ultra-hot electrode.

I expect - this is only a guess - that melting pots made of metals with a small vapour pressure contaminate less the melt. That would mean W. Available from Plansee for instance. Cheaper would be ceramics, where oxides of Al and Mg shouldn't hopefully react with Fe, Ni, Co.

My preferred way would melt only locally the analysed alloy. This is done for deposition of Al by evaporation for instance. In microelectronics we used an electron beam for that, but a laser should do it too.

----------

Optical emission spectrometer has the capability to measure alloy composition locally, pointing to precipitates, segregation and do on. Do you want to melt the analysed alloy to get a mean composition? This may demand to exceed widely the temperature at which the alloy begins to melt.

[Showseph]

----------

One relatively clean means to heat metals is induction. Contamination by the melting pot but not by the ultra-hot electrode.

I expect - this is only a guess - that melting pots made of metals with a small vapour pressure contaminate less the melt. That would mean W. Available from Plansee for instance. Cheaper would be ceramics, where oxides of Al and Mg shouldn't hopefully react with Fe, Ni, Co.

My preferred way would melt only locally the analysed alloy. This is done for deposition of Al by evaporation for instance. In microelectronics we used an electron beam for that, but a laser should do it too.

----------
Optical emission spectrometer has the capability to measure alloy composition locally, pointing to precipitates, segregation and do on. Do you want to melt the analysed alloy to get a mean composition? This may demand to exceed widely the temperature at which the alloy begins to melt.

Thank you for the reply,

We've actually got an induction heater collecting dust in our workshop, I will give that a shot tomorrow to see if it is suitable for our needs. We have ceramic crucibles available, so I think those should suffice.

W has been an issue that we have been running into with the TIG welder, but it is difficult to avoid it with the setup that our machinist built.

We're looking to melt down samples that are either too small or irregularly shaped into a slug/button so that we can determine a mean composition. Do you think that using a slower heating process would be more beneficial over a faster one for what we are trying to accomplish?

[Enthalpy]

Do you think that using a slower heating process would be more beneficial over a faster one for what we are trying to accomplish?

Huh, I dunno.

Re-thinking about a W crucible: ceramic should be better, as some metal from one sample might wet the surface of a metal crucible and contaminate the next sample.

Though, the melting points of Fe, Ni, Co, Ti exceed the operating temperatures given by most suppliers for alumina and zirconia (which must be stabilized). Silicon carbide exceeds 1500°C but it will dissolve in molten metal. Magnesia looks like the best choice. It's also less likely to react with metals, molten titanium being bad for that, magnesium even worse and it's present in aluminium alloys.

[Ithakaz]

Have you tried acid digestion?