Back to Ganoksin | FAQ | Contact

[Correction] Sterling annealing

hey all, in rereading what I wrote in annealing the other night,
I realize that I’ve mixed up some info here. I was talking about
heat treatment to age harden sterling silver, as well as
annealing it, and mixed things up a bit.

I wrote:

      In normal shop practice, we anneal sterling silver by
torch heating just till we see some "glow", and then waiting
just till the glow is gone, and quenching in water.  For
maximum softness, use a kiln set at 1375-1400 F (745-760C). 
Hold at that temp for 15 minutes, and then quench in cold
water.  (allow to air cool in cool air till the bright glow is
almost gone, to avoid cracking the metal)  This results in a
hardness of about 56 (vickers).  cold worked, such as rolling
to a 60 percent reduction, will produce a hardness of 140-180
V.    If you take the annealed metal, and hold it at a temp of
600 F (316 C), and holding it there from 30-50 minutes, then
allowing it to air cool (no quench), you'll raise the hardness
back up to about 110-120V, the same as about a 50 percent
reduction in the mill, (half hard, which is defined as reducing
the thickness by two B&S guages). 

You’ll note that I’ve here suggested that a normal annealing
temp requires 1375-1400 degrees.

I got that pretty much from paraphrasing the section in the
Handy and harmon “handy book” where it’s talking about age
hardening, figures that corrolate as well with those given by
Alan Revere for the age hardening process, in his fine book (I
surmise that Alan probably got the figures from handy and harmon
too, though I don’t know).

Typing with the book in front of me, I just copied figures
without quite thinking about them and what those temps actually
were. Late nite messages… Later, thinking about it, I realize
that this was an error on my part, and that this was not, in
fact, the right way to anneal. Checking, I found that it was
correct for heat treating, however.

Anyway, in the above paragraph, I describe that initial
annealing temp as the normal best anneal… THAT’S NOT THE CASE.
The above sequence is way too hot for a normal anneal, and in
fact may be too hot for many fabricated items as well, since it’s
well above the melt point for most solders. It’s the optimal
preheat temp before age hardening, since what’s desired here is
to both maximize grain growth as well as initially redissolving
the copper uniformly prior to precipitation hardening. the above
sequence does NOT produce the best anneal. metal heated this hot,
though now soft, will produce an orange peel surface when bent,
due to the resulting large crystal size. Note that for items
that cannot take this pretreatment, the later low temp heat soak
will also increase hardness almost as much, though without the
large grain size, which increases hardness even more. Also note,
that for this sequence to work, it’s almost essential to have an
atmosphere controlled furnace. 15 minutes at that high temp
will defeat pripps, and just about any other flux used to control
fire scale. Without an atmosphere controlled furnace, you’ll
have a mess on your hand.

So then. The CORRECT procedure for normal annealing of sterling
silver is this:

SILVER IS BETWEEN 1100 F, AND 1200 F. (593C-649C) Temperatures
above 1200 tend to dissolve the copper rich phase in sterling,
and maximum softness will not be reached. In practice, this
means you heat the silver until just a trace of glow/cherry red
is seen. Sterling anneals very quickly, so a “soak” at this
temperature is NOT needed at all. The silver may be air cooled
or quenched. The handy book suggests there is little difference
between air cooling and quenching, but in practice, most
silversmiths will tell you that quenched silver is slightly
softer. too slow an air cool will start to age harden the metal
a little. However, if quenching, be sure to let that cherry red
glow dissappear before quenching. You can crack the silver if
you quench it from too hot a temp. I’ve actually used this on
purpose a few times, intentionally overheating the silver to a
fairly bright red, then quenching from that temp, to produce a
series of shatter fractures and cracks that were quite
spectacular. this was in sheet metal that I’d abused in
misaligned rolls and then bead blasted until it was quite
distarted and warped, and then it got shattered like that. The
cracks are things you could never do with a sawblade, and the
result was a piece of highly distressed looking metal that was
perfect for the piece I was making… But that’s another story.

Hope this helps. And I hope I didn’t mess anyone’s metal up too

Peter Rowe.

Thanks Peter I had wondered about quenching of annealed metals
can you tell me does the same rule(s) apply to gold and platinum
as well?

 Thanks Peter I had wondered about quenching of annealed
metals can you tell me does the same rule(s) apply to gold and
platinum as well? 

Yes, but not completely the same. yellow golds will be softer
with a quench, and can be treated like sterling. White golds
have to be quenched after all trace of red glow is gone, or you
can crack them easily, but you should also note the they will age
harden quickly if you cool them too slowly, so quenching is
important. With platinum, the key is annealing it long enough,
as it is NOT instant. It may require as much as 20-30 seconds at
a bright orange heat to completely anneal thicker pieces.
Quenching doesn’t seem to make too much difference once it’s
completely annealed, though it speeds up your work.


In my experience with platinum the following would apply. Prior
to annealing the platinum the item should be placed in a bath of
pickle to remove any contamination of steel on the surface from
the mill or any steel tools. If you suspect that there is a base
metal contamination a warm nitric acid bath for several minutes.

In my experience with platinum you have to heat the platinum to
a bright orange (temperature at this color that would be 1025
degrees Celsius or 1877 degrees Fahrenheit) and sustain the heat
for 1 minute per 1 millimeter thickness of metal. So if you are
annealing a 4 mm thick square wire you must have a sustained
heating of the metal at bright orange for 4 minutes to completely
anneal the square wire. Again the metal can be quench or not.
Platinum can be reduced by about 70% but one should be aware
that it becomes very dense and should be reduced slowly to avoid
denting stakes or rolling mills.

You may consider visiting the
site for more complete info on this.

Happy Holidays to One and All
Edward J. Friedman

Regarding the heat treatment of sterling to increase its
hardness, Peter mentioned the table in my first book,
Professional Goldsmithing, called Physical and mechanical
properties of silver alloys. This table presents many figures
including a comparision of hardness after annealing, work
hardening and age hardening (heat treating). The
comes from Mark Grimwade’s book, Introduction to Precious Metals.
It states that to gain the maximum hardness, hold sterling at
750B0C for 30 minutes followed by quenching and then heat it to
300B0C for 60 minutes. This will yield a Vickers hardness of
110/120, which is significantly harder than sterling’s annealed
hardness which of 56 Vickers. Work hardening can reach a hardness
of 140-180 Vickers.

What I find even more interesting is that in certain gold
alloys, you can reach a greater hardness by heat treating than
work hardening. For instance 18k gold alloy of 750 Au, 45 Ag, 205
Cu that is worked to the max has a Vickers of 240, but when age
hardened (in this case by heating at 280B0C for 60 mintues after
annealing), the hardness jumps to 325 Vickers! The jump is
greatest for 18k red, with a composition of 750 Au and 250 Cu.
Here working leaves the metal at a maximum of 240 Vickers.
However after age hardening the number jumps to 340 Vickers. This
comes from the Alloy Data Sheets provided by the
World Gold Council in 1990.

   .... The jump is greatest for 18k red, with a composition of
750 Au and 250 Cu. Here working leaves the metal at a maximum
of 240 Vickers. However after age hardening the number jumps to
340 Vickers. This comes from the Alloy Data Sheets
provided by the World Gold Council in 1990. 


That alloy is one of the scarier ones to work with, due to it’s
formation of an ordered array structure that is very brittle and
cracky, when cooled slowly through around 700 degrees F. or so.
(don’t have the exact temp right here.) I find myself wondering
whether this age hardening process for this particular alloy may
actually be the formation of that unfortunate ordered array

Just wondering…

Peter Rowe