I found your explanation very much enlightening but got lost at the "nucleation sites". Could you develop it a little further?
Gabriel,
Sorry for the unnecesarily large words… Lets try and say it
all in simpler terms. In a molten, cooling metal, as it cools
below certain key temperaturs, crystals begin to form. The trick
here is “where”, “why”, and how many places these form. in
general, it takes some disturbance, or impurity, to cause a
crystal to start forming. If there are not such “triggers”,
then it’s theoritically possible to actually cool the liquid to
below it’s normal melting point without it solidifying. That’s
then called a supercooled state. Notice, I said theoritically.
In practice, that just doesn’t happen with molten metals.
Something, somewhere, always seems to set it off. Maybe a bit
of vibration, some rough spot on the container, an extra atom in
the mix that somehow creates a focus point for things to get
started… Who knows. These sites where crystalization starts
are called nucleation sites. If you can add components to an
alloy that increase these sites where crystals start to form,
then as the metal cools, a greater number of crystals will form,
leading to a smaller average crystal size. This, as I discussed
in that last post, is highly desireable in jewelry castings.
You are probably already aware of a beautifually graphic example
of crystal nucleation, and it’s requireing some site for it to
start. Ever seen one of those instant heat bags you can stick
in your gloves in winter? It’s clear, and room temp, until you
activate it by literally, clicking/bending a little metal disk
inside the container. Then it heats for an hour or so. And
later you can reactivate it by microwaving it to melt the liquid
again and letting it slowly cool… Clicking that disk creates a
little shock wave. That disturbance and little pressure
differentials, gives the liquid spots to start crystalizing which
it instantly does at a rapid speed. You know how it takes the
input of extra energy to melt stuff? You heat it to the melting
point, and then you continue to heat, and until it’s converted to
liquid, the actual temperature doesn’t rise much, and then once
liquid, it starts to rise again. This extra heat you have to
pump into the crystal (solid) to get it to convert to liquid is
called the heat of crystalization. In those little heating
pouches, allowing the super cooled liquid to crystaliza allows
that stored heat of crystalization to be released again (the
crystalized state of a material is more stable, and has less
stored energy than a liquid state), thus warming up the pouch
and your fingers as well. The key point to this illustration is
that it requires not only that the liquid be cooled to below it’s
melting point for the liquid to solidify and crystalize, it also
requires some point at which the crystalization can start.
That’s called a nucleation site, as the crystal starts at a
point called the nucleus, and grows out from it.
But I digress. (In case you’d not noticed. Are your fingers
warm, now?)
In our molten metals, being cast, as they cool, the liquid metal
becomes less and less “comfortable” being a liquid, as the energy
levels of each hot atom becomes less than really required to keep
it from bonding to it’s fellows. However, just when, exactly, the
process of crystalization starts is a bit variable, and atoms
would much rather bond to a spot where the crystalization process
is already occurring, than to start it all on their own in a new
place, since the existing site can help absorb some of the excess
energy carried by the still liquid state atoms. Once you start a
crystal growing, it often growns very quickly.
What grain refiners do is to mix into the molten metal different
atoms which will act as “starting points” from which crystals are
more likely to start growing than would otherwise be the case. I
suspect, but don’t actually have the knowledge to say for sure,
that what happens is that these are materials with a higher
melting point that thus crystalize out of the melt before the
main mass of material, leading not just to different atoms
floating around, but instead to lots of little crystals of the
other material, shich then can form a nucleus to start building
the main crystals upon (Anyone out there know if this is a
correct understanding of the process?)
And it’s interesting as well to understand the way the metals
melt. Pure metals have exact melting points, changing from solid
to liquid at exact temperatures (though that temperature changes
with pressure, but that’s another subject,) requireing only that
you add enough additional energy to the metal to overcome the
heat of crystalization. But the mixes behave differently. The
contact of one metal atom with that of another type of metal atom
can lower the temperature at which the two become liquid. For
each ratio of, say copper and silver, there is a lowest
temperature at which that mix can exist as a liquid. The
mixture of the two which has the lowest possible melting point of
any such mix is called the eutectic alloy, and the temperature at
which it melts is the eutectic temperature for those two metals.
With copper and silver, that ratio is 71.9% silver and 28.1%
copper. That mix, the eutectic allow for copper and silver,
melts at 1435 F/ 780 C. If you heat that alloy, when it reaches
that temperature, it will become all at once completely fluid,
instead of become slushy and more and more liquid the hotter it
gets. This is the same way that a pure metal melts.
When you start to melt any OTHER mix of those two metals, at
that eutectic temperture, melting starts, with enough atoms of
each type of metal becoming fluid to form that composition. The
remaining, excess metal stays solid until you raise the
temperature more. With each raise in temp, a wider range of
mixtures can exist as a liquid, and sufficient amounts of the
previously excess other metal will melt into the mix. For
example, with sterling silver, when the eutectic temperature is
reached, all the copper in the alloy immediately melts together
with enough silver to form the eutectic alloy. since there is
only 7.5 % copper in sterling silver, and the eutectic alloy
contains 28.1% copper, there’s a lot of left over silver. It
will stay solid, as silver crystals, until the temperature
raises. By the time the temperature has reached 1640 F/ 890 C,
all the silver will have melted, and the melt will have the
composition of sterling silver.
When you start to cool this mix, as it falls below the melting
point of sterling, only a mix with less silver can remain fully
fluid, so some of the silver crystalizes out as almost pure
silver crystals. This continues as the temperature falls until
the eutectic temperature is reached. At this point, you’ve got a
mass of solid pure silver crystals floating around in a fluid of
the eutectic alloy. As you cool it more, (removeing the heat of
crystalization… remember that?) that eutectic alloy now
solidifies into crystals of that formula, and cools. Solid
sterling silver is a mix of two distinct crystal formulas.
You’ll find both almost pure silver crystals mixed with those of
the eutectic alloy. And, if you heat treat the silver, you can
also cause some of the copper to literally migrate out of those
solid eutectic alloy crystals, forming new, mostly copper
crystals, at the grain boundaries of the parent crystals. This,
by the way, is what is called precipitation hardening. When you
anneal it, that copper precipitate is then no longer stable as a
seperate entity, and it redissolves back into the parent alloy
crystals.
In gold alloys, which are often far more complex, what happens
if harder to spcifically describe, but it’s similar. As the
alloy reaches a certain minimum temperature, certain proportions
of the constituents can exist as liquids. Since they can, they
do, and the alloy starts to melt, giving your a liquid of that
eutectic mix (which is probably several metals, rather than just
the two in sterling silver) with solid crystals floating in it.
You must then raise the temp until all the material can now exist
as a liquid before you’re gone through the slushy stage of
melting. Those alloys which are already fairly close to a
eutectic mix, may have fairly short “slushy” ranges, and will
melt quickly. Those with longer slushy ranges will malt in a more
gradual manner.
Goldsmiths attempting to work with gold alloys will be well
familier with the fact that some yellow golds, containing perhaps
more silver and zinc, may melt rather lower or more quickly than
others. I can think, for example, of one ring manufacturer whose
alloys were almost imposible to solder a finding to using a hard
solder. The gold would melt only a little above the solder
temp, and as it happened, this alloy was one which had a very
narrow range between starting to melt and being completely fluid
(a narro slushy range, meaning it’s basic composition was fairly
close to a eutectic composition) This alloy tended to produce
beautiful, porosity free castings, but it was very difficult to
work with, since if you overheated it just a hair, it would give
you little warning and suddenly you’d have parts of the ring
turning to puddles.
And there’s yet another interesting application of all this
which we use every day. Solders. Solders are metal alloys
designed not only to melt at lower temeperatures than the metals
we join with them, but also to flow out well on those metals
allowing the joins to be made with ease. The way the solders
flow is what’s interesting. In order for a solder to flow well
out onto the metal, it must be attracted to the metal being
joined. Yet it must not simply slump into that metal and
dissappear. If for example, a 14K yellow gold solder were made
that had a certain melting point, but who’s melting point could
be significantly lowered by the addition of the 14K alloy being
joined, then when you melted that solder it would cause the
surrounding metal in the join to simply join the molten pool.
You’re tiny piece of solder would somehow become a rather large
puddle, more molten metal than you’d intended. That wouldn’t
give you much of a clean seam.
If the addition of the joint alloy to the solder alloy doesn’t
much change it’s melting point, then the solder can flow out on
the metal without greatly dissolving much of it, beyond what gets
dissolved in by the fact that you’re using a temperature to
solder the metal that’s higher than the actual minimum melting
temperature of the solder. Something in this range is usually
the desired idea, as it allows you to “pull” the solder towards
the heat, without causing excess distruction of the joined
material. And it allows the molten solder to dissolve a little
of the parent metal, while simultaneously diffusing into it. All
good things.
If, however, the solder is made so that adding the parent metal
of the joint raises the overall melting point very quickly, then
when the solder is melted, there will be little diffusion into
the parent metal, as well as less attraction of the solder, so
that the solder will not tend to flow well. Such solders can be
made so that even the dissolution of only a little of the parent
metal from the joint quickly raises the melting point. With
these solders, there is little tendancy of the solder to flow
away from where it’s originally placed, and it will tend to just
slump into place where melted, and then, absorbing some of the
parent metal and raising it’s melting point, it will freeze
there. This is the type of alloy that was often used in dental
work, as Skip was telling us a few days ago, where multiple
layers of solder were built up on a frame to create a dental
restoration. These solders were designed to bond, but not to
flow out, allowing the technician to build up thickness and form
with repeated applications of the solder.
And that, I think, is enough. Maybe way too much… For
tonight.
Peter Rowe