Back to Ganoksin | FAQ | Contact

Gravity casting into plaster


I have been experimenting with gravity casing into carved gypsum
plaster molds. This is part of an ongoing research project into how
certain medieval fine metalwork may have been done. I can make my
molds work in a number of ways using modern materials and techniques
that would not have been available in the 7th century, so for the
sake of demonstration I am trying to learn how to make this method
work using just the most basic direct methods.

This is not lost wax I am talking about, but directly carved plaster.
Investment, although it is gypsum plaster based, doesn’t work very
well for carving. There is plenty of archaeological evidence for clay
molds during this period, but clay does not carve well enough to
accomplish certain techniques. Plaster will make exactly the mold of
the correct shape and level of detail. The problem is getting it to
fill with metal.

It takes a very long time to drive off the water from the molds in
order to get them dry enough that they do not release steam when
silver or bronze is poured. Baked hotter than 500 degrees F (260
C)and the material cracks. A 3/4 inch thick mold takes more than 10
hours in the oven to dry. Cast in a centrifuge it casts just fine,
but gravity cast it seems that the material just is not that
permeable to let the gases out and give a good fill.

It appears that there is an entire group of modern plasters
available for making cope and drag sorts of molds, so somebody must
be actually doing this. Additives like talc or pumice increase
permeability but ruin the carvability of the material. I am thinking
I should try thinner molds and maybe more permeable materials for the
backup plate, but I would appreciate hearing advise from anyone who
has experience with casting into plaster molds.

Stephen Walker


Hi Stephen,

I’ve never tried serious medieval casting, but is it worth pondering
what tricks they may have used? I’m thinking some variation on
steam casting. Sort of an “open face” steam cast? Just my $0.0000002
worth… Either that, or sling casting.



Hi Stephen,

Soapstone moulds were used, as was tempered clay.

Clay doesn’t have to carve well, you make a model, and take a clay

There was an article on using iron cement moulds for metal casting,
I could dig it up and sent it to you if you like.

I can also put you in contact with people who actively experiment
with bronze casting using Viking techniques.

Another avenue you can try is using hydrocal and unscented baby oil
(the scented stuff produces an evil odour), more of a sand casting
approach, but it does eventually dry out, and go hard.

Kindest regards Charles A.


Uh, have you looked at the possibility they used soapstone? There
were a lot of soapstone molds made in the Early Medieval period (the
National Museum of Ireland has a collection) and it will hold pretty
fine detail.



This is not what you asked about, but I want to comment as a
professional potter for 20-some years before taking up
metalsmithing. I used to do extremely detailed carving with really no
functional limit on the amount of detail. I used porcelain, but any
really smooth clay body should work. The trick was to keep it leather
hard while carving was being done. Once bone dry, it was too brittle
for some types of fine carving. Obviously this was a bit if a PITA
but I did it all the time.



Hi I am a member of a historical recreationist group of medieval
eras. Many molds were carved from different stones depending on the
region. here is a link to our forums database on metallurgy.



Interesting subject…

What do you mean by Gravity Casting? You said plaster molds, so you
are just pouring into the mold? Why stop there. Would they have
thought to use some wet and soaked paper or skins to do a steam
cast? This causes a layer of steam as you cap the mold and FORCES
the metal into the mold. Maybe they had an accident where something
caught fire and they put it out with wet rags so they wouldn’t have
the metal explode by pouring water onto it… They noticed that it
was better cast, and eventually they tried it again and again to
refine it.

You said Centrifugal cast works for you, but why could they not of
poured the mold and then spun it with some type of sling? Now you
have Centrifugal casting. :slight_smile: I know it sounds funny, but I think we
sometimes just assume ancestors didn’t know what to do but I think
they were pretty smart. Look at all the bronze age castings of axe
heads, etc… I think you basic smithy that did this magic on a
daily basis would sometimes have a happy accident and remember how
to reproduce it later. Also they would of worked as a team not an
individual craftsman. So while the smith may prep the mold and the
metal, the apprentice got the dubious task of spinning it a sling.
No workmans’ comp back then, so if something happened, you just got
a new apprentice.

Also why not sand casting? You get a nasty chunky result, but after
refinement and fileing,etc you end up with the finished piece. Not
right out of the mold, but with refinement it can be done. It really
depends what you are trying to replicate that they would of made.
The sand casting would be easy if they knew to pack sand around the
item… Cope and drag specifically refers to sand casting by the
way… That is how the early bronze casting would of been done. So
you have to think of you timeline… What is plaster? ground up
limestone/cement. Would they have it available back then? Don’t
know, that’s why I’d look at the sand casting.

If not poured plaster, maybe from Chalk. You have a HUGE coast of
chalk in England used for many things. It would be fun to carve some
of those blocks up and cast. In a similar manner, the Natives of
North America used a rhyolite tuff for casting that they would
carve. But they were shown how by the Spaniards from what I recall,
so I don’t know if they naturally developed it or the spaniards just
helped them refine an existing technique.

And my last thought is that if you want to know how they did it in
the medieval times look to some undeveloped countries that exist
today… Many of those processes were documented and still in use to
some extent today. While “WE” in a modern society have moved on, the
old tried and true methods are still in use and handed down from
generation to generation.

Just some thoughts for ya. Fun topic.


What about carving volcanic tuffa or cuttle bone. You get
interesting texture for free!

Charles Friedman DDS
Ventura by the Sea


Thanks for all the suggestions, both online and off. The specific
technique I am trying to reproduce relies on being able to modify a
mold that begins as an impression of a model, but then has detail
added by carving directly into the surface. Stone molds are not
going to do it. Cuttlefish bone might work but many of the medieval
molds made in this technique are too big to have been made that way.
Sand molds can give excellent detail in the impression but not when
carved. Clay almost works and there is plenty of evidence of clay
molds unearthed in archaeological digs, but so far none of the clays
I have tried, including porcelain, will hold the kind of detail
without chipping out small spaces between the carved lines. When I
have tired it with clay it can be done, but frequently fails in the
way mentioned above. Having examined many medieval pieces I have
never seen the kind of chip-out failures that a constant struggle to
overcome in clays in finished pieces, even some that were very
shoddy workmanship, so I must expect a more cooperative material.

Gypsum plaster is just that material. As I mentioned before, there
is no problem making it work for sculpting the kind of detail I am
trying to reproduce. But it cracks up when heated to over 500F and
doesn’t fill very well at that temperature without adding some
pressure in the centrifuge.

The best advise so far comes from my son, who is a 1st year
engineering student. He thinks that the grain size of my modern
plaster is so small and that the plaster would be more porous if it
were of a lesser quality. Not so pure and not so finely milled. Sand
is regularly added to foundry investment that is commonly used for
burnouts at temperatures over 1000F. Talc is also an ingredient in
metal casting plasters. These additives mess with the carving
properties but I am running a kiln test to see how various
combinations will hold up, including one that is a sand and plaster
mix with a thin straight plaster surface that is carvable. So far
even the plaster sample with just 5% talc is taking much more heat
than the pure gypsum plaster. At 750F, where the kiln is now before I
turn in for the night, none of my test samples are showing any
cracks, so I am hopeful that tempering the plaster slightly might be
all it takes so that a much hotter mold can be cast.

The type of detail I am trying to achieve is like that of the stem
of the Ardagh Chalice or the body of the Tara Brooch. The Pictish
hoard of brooches known as the St. Ninian’s Isle Treasure includes
three brooches that are variations on the same model, done in the
same technique, but not of nearly the same quality the examples
mentioned above. My current experiment is trying to reproduce the
molding technique for these three penannular brooches. The main
problem that remains is getting the molds to fill, which seems to be
a matter of mold temperature and material permeability.

Stephen Walker


A couple of suggestions.

The type of mold making you are describing, in which a smooth mold
has decoration added by carving directly into the negative surface,
is the technique used for ancient Chinese bronzes and still used for
the casting of large Buddhist temple bells. These techniques have
been quite well documented, so you might look at books on Chinese and
Japanese bronze casting of ritual objects.

Are you certain that the details you want to reproduce were cast? A
cursory look at a large photo of the Armagh Chalice suggests to me
that the knotwork and animal interlace plaques that decorate the
circumference of the bowl, between the silver & niello bosses, were
pierced and chased. The plaques in the recesses of the Tara Brooch
may well have been done that way. I’m not familiar with any technical
examination of the Tara Brooch; medieval gold work has had less
technical research done on it than has Classical gold work. However,
since you’re talking about techniques prevalent in the 7th century
you might look at Merovingian Garnet Jewellery by Birgit Arrhenius.
It has an extensive technical section, I believe.

Just my 2,


about working with clay. clay passes through several stages, where
working with it, carving/burnishing are quite possible without any
cracking/chipping. your experience says you found it chipping easy
while carving. with some finesse bone dry clay can be carved with
incredible detail. i don’t know what tools you are using, maybe if
you could offer some exact details, this could be further explored.
i’ve scratched around on stoneware with crude tools and with a light
hand, didn’t get any chipping out on the edges a bowl shape. i also
don’t know at what drying stage your attempts were applied.

there are fine carvers of clays for sculpture, the details would
knock your socks. julietta wehr is a facebook acquaintance of mine,
her work could shed some light as to the “carvability” of ceramic

My current experiment is trying to reproduce the molding technique
for these three penannular brooches. The main problem that remains
is getting the molds to fill, which seems to be a matter of mold
temperature and material permeability. 

Those examples are mostly hand worked pieces, moulding may not cut it
(excuse the pun). CIA


I recently had great success in pour casting into carved casting
investment. I make 2 bricks (of whatever size I think necessary) of
casting investment, one a bit thinner than the other. Carve the
thicker, put the two together and cast. Have had a couple of great
results and a couple not so great. As in any casting, preparation is
important. I do not bake or heat them before casting…just let them
dry naturally. I do soot the cavity though.

from Don in SOFL


I made two changes that have pretty much solved the problem. I added
risers, suggested by an Orchid reader offline, (thanks Dave!)and I
used a different material for the back plate. One sample used
Satincast 20 for the back and the other a plaster sand mix. Both were
cast with the mold at 500F. Both worked just fine. The risers alone
may have been enough to fix it.

Are you certain that the details you want to reproduce were cast?
A cursory look at a large photo of the Armagh Chalice suggests to
me that the knotwork and animal interlace plaques that decorate the
circumference of the bowl, between the silver & niello bosses,
were pierced and chased. 

The Ardagh Chalice and Tara brooch combine lots of techniques. The
stem of the chalice was cast and a few other parts. The band around
the circumference was done with gold filigree and granulation on gold
foil. None of it is chased, but, yes, it is pierce on those panels.

Thanks everyone for your suggestions. When the project is done I
will make the complete report available on Orchid.

Stephen Walker