I've noticed that the area there seems slightly indented right
after the pour as though it perhaps shrunk in that location.
That's a fairly important clue. The metal did in fact shrink there.
The cause is that this is apparently the last part of the ingot to
solidify, so thermal shrinkage is trapped there, with surrounding
areas already solid, the metal must shrink in the middle. The problem
with that can be that in shrinking, it can literally be tearing
itself apart. Silver, when hot, has very little strength, so as it
cools, it cannot easily actually stretch the way it can do when
Although you cannot see them in the raw ingot, it will already have
micro cracks and porosity in that area it was shrinking in. Some at
the surface, more inside the ingot. This is the reason why forging
before rolling doesn't help much. The metal already has the cracks
that will expand to visible size when you roll.
What you need to do here is a concept familiar to those who do lost
wax casting in order to get complete fills with minimum porosity
(which can be caused by the same shrinkage situations). Progressive
solidification. As you're doing now, the mold is preheated. But
you've got the ingot mold currently heated evenly.
What you need to do is heat the top upper section of the mold hotter
than the bottom, so that when you pour metal in, that bottom of the
mold solidifies first, and solidification proceeds upwards, only
solidifying at the top of the mold/ingot after the lower parts of the
ingot have solidified, rather than solidifying last in the center of
the ingot. That way, as the metal cools, the shrinkage proceeds at
the boundary between lower solid metal and still molten upper parts
of the ingto, rather than being trapped in the middle with nowhere
to pull still molten metal from. You'll know you're getting this in
some cases when after you pour, the top edge of the ingot has a
sometimes fairly deep depression in the top edge. Sometimes it's
quite distinct, with an actual little hollow cavity in the top edge.
This can be annoying, since it ruins the very top of the ingot. But
that's OK. It means the rest of the ingot was able to solidify in a
progressive manner from bottom to top, so it did not trap
solidifying and shrinking metal in the middle. You can always (and
should) lop off that top edge anyway, since it will also be where any
impurities, flux inclusions, oxides, etc, will concentrate. When your
ingots do not have that shrinkage depression in the side, you won't
have those micro cracks to expand, which should help. So when you're
preheating the ingot mold prior to pouring, concentrate the flame on
the top of the mold more than the bottom, especially at the end of
the preheating operation. That will control when the metal solidifies
in each portion of the mold.
R. E. Rourkes suggestion of a refining flux like sal ammoniac
(ammonium chloride) is a good one for removing traces of other
metals, especially iron or lead or tin that may have accidentally
gotten into the melt. Be sure to use very good ventillation when
using the stuff though, as it releases chlorine gas during use.
The chlorine is what makes it work, combining with the baser metals
to form chlorides of those metals, which are then not soluable in the
molten silver, so rise as slag to the top, where they can be skimmed
Another comment, which is probably not your problem, but worth
mentioning, would be your source of copper. Some commercially
produced copper items (some plumbing fittings) include traces of
lead. That can be disastrous for your silver. Be sure to use a copper
source that is free of any lead. Electrical wire is one such source
of high purity copper, if you're not already using such a source.
Your melting flame should be reducing, but not highly so. Only
slightly to the reducing side of neutral. You need the heat, and
relatively shorter melting time is more useful to preventing
impurities and oxides etc, than using a more reducing softer but
cooler flame which then can take a lot longer. Also, too reducing can
end up increasing gas absorbtion by the molten metal, which is the
opposite of what you want.