Shrinking models


I have a museum artifact that I have been asked to shrink (a winged
phallos from Pompeji actually). It is currently about 6 cm which is a
bit large to wear around your neck and I would like to make a 3 cm
version and one 1 cm version (for earrings). Any sugestion on the
best method to do this?


The most accurate way is to have the original scanned either with a
touch probe or laser scanner, import that into a suitable CAD
program, scale it down the desired dimensions and either mill it or
grow it on an RP machine.

You can also hand carve or fabricate copies, but depending on the
skill of the person involved, there can be some discrepancies.

I have both a probe and laser scanner, and do that sort of thing all
the time, if you’re interested in exploring that option, I’d be
happy to help you out.



Make a mold of it and use one of that shrinks when it cures.
Smoothon has such (I think it is called “Reduce It”, Douglas and
Sturgess in San Francisco, CA can get it for you if you want), as do
most other mold making suppliers/mfgrs. These products are usually a
water mixed material that once formed and solidified, the piece is
removed from the mold and heated to drive off the moisture and piece

John Dach

currently about 6 cm which is a bit large to wear around your neck
and I would like to make a 3 cm version and one 1 cm version (for
earrings). Any sugestion on the best method to do this? 

To my knowlege, which in this case is probably accurate, there is no
practical way to shrink anything that much by any sort of molding
process. 10%, yes. A factor of 6, no. I saw a demo of molding a mint
coin, and then molding the product of each mold over and over, and
the coin basically disappeared pretty quickly. When you say “shrink”,
I think of something like that, and that sort of shrinkage is
unpredictable, anyway. Pantograph, 3d scanner, or the old fashioned
way - carve them in wax.

Yes use “Redu-cit” rio grande sells it. it’s kind of crazy to work
with but if you experiment a few times you can get aprox. a 50%
smaller object you won’t be able to burn it out though, you will
have to RTV it and shoot a wax copy, you will also probably have to
fill in som air hole and touch up the shrunk version.

Basically it will take alot of trial and error and “playing around
with” ; I guess what I’m saying is I hope there is room in the
project to pay for some of your time!..if
you can’t find it at Rio I know other sources also.


Any sugestion on the best method to do this? 

Try gelatine:

  • Make a mold, silicone will do
  • Fill with semifluid gelatine
  • Put in the fridge to cool down
  • Take out the gelatine model and let shrink/dry at room temperature
  • Cast the model as if it were wax

I regulary use this method to size down rings where I only have one
mold. It is getting more and more difficult, though, to find
gelatine, but I guess that ultimately gummie bears will do as well.


dear Bjorn. one of my favorite jewelers in the northeast is Anthony
lent. he is at FIT in New york. he does alot of sculptural work.
then he offers the pieces in many different sizes. of course you can
do this in CAD if the origional design is in CAD. or if you do A
three-d scan into a CAD format.

however mr Lent used a product called…are your ready for
this…DIK reducer… (May also be called Dick reducer) you make a
mold, inject this stuff and then it shrinks to produce a smaller
version of the master piece, you can do this over and over again. if
the piece is the same thickness it will reduce evenly, if the design
has many thick and thin parts it may not reduce evenly. mr Lent did
a lecture on this at the MJSA trad show a few years ago. their may
be something in the archives of ganoksin, or MjSA may have a paper
somewhere on it.

wayne werner
baltimore md

Hi Donivans,

As to the shrinkability via mold and Reducit, a friend of ours molded
his hand with a finger pointing out and reduced it twice or maybe 3
times and had a very small sized charm like item, and you could still
see his fingerprints with enough magnification. It was rather

John and Cynthia Thomas-Dach

Alot of people don’t look to PMC for mold making. The original type
shrinks 28% so if you make your modeal and than put it in the kiln
you have a smaller version, than if you mold it and put it in a kiln
again it will shrink even further. I use this often when I want
different sizes of the same design. Than I either make them in PMC
or send them to a caster to mold and cast in the approiate metal. Its
so much better than wax as you can file and sand the piece easily
when its in the leather hard state so that after you fire it you will
have to do very little work to it. Plus, its much easier and quicker
to file it in the clay form than in metal.


If you are into CAD/CAM:

Scan the item (laser scan, flat scan if one sided) ==> save scan as
STL file (or whatever format you can use) ==> open file in CAD
program (Rhino, or software with that comes with the scanner) ==>
reduce size of item by 16.6% (1/6) or whatever you want ==> export
the file ==> load file into CNC software ==> mill the item in wax ==>
invest/cast as usual.

I do this stuff all the time. Shrink or enlarge, same principle.

I regulary use this method to size down rings where I only have
one mold. It is getting more and more difficult, though, to find
gelatine, but I guess that ultimately gummie bears will do as

Knotts gelatin should be available in every decent sizes grocery

John Dach

There have been quite a few suggestions for this, and it’s true that
some mold compounds might actually work. The problem, to me, is what
was asked for. The job is to make a duplicate of an artifact to 1/2
size, and also to 1/6 size. The key word here is “duplicate”, which
doesn’t mean “pretty close” or “near enough” or “kinda sorta like”.
If there is a molding system that will do that down to 1/6 scale or
even 1/2, it’s news to me, but maybe there is. I personally would try
to avoid chasing down some shortcut that ends up not working anyway
and just carve them, but I can do that, which not everyone can. I
don’t say this to slam what has been suggested - if they work, they
work, and I’d be glad to hear that. Just that molding like that is
unpredictable. Depending on shrinkage, which is almost always uneven
depending on the original, is unpredictable. A 2mm detail is going to
shrink away, also, and a 3mm section will become paper thin - issues
like that. Carving is predictable. Or 3d scanning, which is the
high-tech version.

I’ve been following this thread and thought I’d chime in: I
successfully used Reduc-it to shrink a detailed (though one-sided)
model a while ago. It did work well for preserving the details of the
piece, but I found that the resulting model had a slightly grainy
surface texture, as if it were molded from fine sand. Also, I had to
mix the dickens out of it (an electric mixer was helpful) in order to
not have lumps show up partway through the drying process. How does
this compare with your experiences?

I’d be keen to hear more about the DIK reducer (how appropriate!),
and I’d love to see a photo of the original model, if you have one
handy. After all, who doesn’t love a good winged phallus?

Best to all,

Jessee Smith
Cincinnati, Ohio

Hi again

Thanks for all the replies. As for now I will be going with the
sugestion of 3D probe scanning since I have a contact nearby with
such a machine. I that doesn’t work out I will be going on to DIK (I
just have to try it. Just think about the accompanying infoletter:
Pompeijan phallus, shrinked with DIK. :wink: ) If I get the time I will
also try with gummibears just for the fun of it.

Here Jessee, if You want to have a look at the said pice You can
have a look in my newly opened webshop. Beware of serious spelling
errrrorssss though. I haven’t really gone public yet, and feel free
to comment on that or anything else in private or here, out in the
open. Have you over there heard of “Mickey” on the front page by the
way? Discovery Chanel apparently had a th=ing about him some time

Here’s the adress:"79&grp=629=56 Or here: To
the shopfront: (click on Kilian
Stobeaus kuriosakabinett)

Bjorn Gedda