Bootstrapping a tiny but effective rolling mill


I tried asking this on the forums but evidently my question never
made it there, I will try again.

Since I am still in bootstrap mode for getting tools together for
creating jewelry, Ive determined that one of the things that I need
is a rolling mill.

I really dont need anything general purpose. I simply would like to
start out with a sliver of fine silver about 1 inch long by 1/8th
square, and deform it either into thick foil or wire.

I have a tabletop CNC milling machine, a lot of aluminum, some 1016
Iron, and a few bits of brass.

Does anyone have suggestions for a basic, low-end design for a
rolling mill which can get very small jobs done until I can earn my
way to something better?

One thing I wonder about is whether I can make do with aluminum
rollers to deform the silver or whether I should use the iron for
rollers and the aluminum for the gearing?

Andrew Jonathan Fine

Hi Andrew:

As an exercise in "no money, but lots of time and tooling"
skylarking in grad-school, I took a dead powered rolling mill with
square wire rolls, and rebuilt it into a manual mill with flat rolls.
Machined up new rolls & everything. MAJOR pain in the tail. It’s
do-able, but probably not cost effective, especially with cheapo
mills in the $250- $400 range. It also cost nearly that much to do,
and that was 15 years ago.

Perhaps a better idea would be to figure out what sorts of things
you could make reasonably easily with your CNC rig that would garner
you enough money to buy a small mill. A tabletop CNC rig probably
doesn’t have the rigidity (& capacity) to deal with the steels you’d
want to use to make a decent rolling mill. They’re not especially
complex beasts, but they do have to be both tough and accurate to do
any sort of meaningful work.

An alternate idea for small work would be to make what one of my
English tutors called a ‘pull swedge’, or a rolling drawblock. It’s
essentially a very small free-wheeling rolling mill. Instead of
having gears to crank the material through, you file it to a taper
onto the material stock, feed the tapered bit through the rolls, and
then use a drawbench to yank it through. Typically, these are
produced with profiled rollers (that you machine up yourself to
whatever shape you need), as a way to get wires with oddball cross
sections. Mine only has one roller (that spins freely on a hardened
dowel pin), and a flat polished steel pressure plate that presses
down onto the profiled roller, to hold the wire into the machined
groove. I’ve seen them with two rollers, but making the bearings for
a two- roller version is a whole lot more complex. The pressure plate
version can handle wire, but won’t do much with sheet: the
friction’s too high. For sheet of any width, you really do need two

I can post pictures of mine if anybody’s interested, so long as you
make allowances for the fact that I was ?19? when I made it. Pull
swedges are pretty much all hand-made, and rare as hen’s teeth.
Other than mine, I think I’ve seen maybe 3 others over the years.

Brian Meek.

Aluminium is totally useless for a rolling mill; iron, or better
still, hardened steel is what you need. The stresses involved in
squeezing even annealed pure silver are far more than aluminium
could withstand. With a decent workshop you could, perhaps, modify an
existing small bending rolls unit. They are used for giving smooth
curved bends in sheet metal and small rods. You could possibly get
second-hand from eBay or such.

Regards, Gary Wooding

One thing I wonder about is whether I can make do with aluminum
rollers to deform the silver or whether I should use the iron for
rollers and the aluminum for the gearing? 

You’ve got to be kidding, right? aluminum will deform as quickly,
maybe even quicker, than the silver. You’re rolls need to be STEEL.
Not soft iron or cast iron, or brass or aluminum or cardboard. Steel.
Preferably capable of being heat treated. You did say you want to
make foil, right? That takes nicely flat rolls that won’t deform
under pressure. And for the rest of the rolling mill, well, steel or
heavier cast iron for the frame, and steel for the gearing. Aluminum
gears will never hold up. Probably wouldn’t hold up past the first
day of use. Bearings for the rollers can be brass or bronze if you
like. Even with your proposed tiny pieces of metal, you’d be
surprised at the loads your little mill will need to handle. Thus,
steel is the required metal for most of it.

Frankly, Andrew, making your own rolling mill isn’t going to be as
easy as you think. To work at all well, even with small pieces of
metal, it has to be fairly well engineered and made of decent metal.
Rolls need to be very round and concentric with the bearing surfaces.
That needs a high quality lathe, not a CNC benchtop mill. Even better
would be centerless grinding after the lathe work is done. You might
be able to have a local machine shop help you with this.

But frankly, I wouldn’t bother. High quality mills are indeed costly
tools that need some saving up for and earn pride of ownership once
you’ve got them. But you can get reasonably cheap mills. Frei and
Borel has a cheap one for less than 300. Harbor Freight sells an
even cheaper (in all senses of the word) mill for a bit over 250, if
I recall right.) All of these will get most of what you wish done.
Yes, this is still some money to spend. But it’s less than the value
of the time you’ll spend trying to build one yourself, not to mention
the cost of hair replacement treatments you’ll also need if you try
to build your own judging from the level of equipment you’ve got
available, and your apparent limited understanding of the
engineering requirements for a rolling mill indicated in your post.

Now, if even these prices are too much, and you really only need to
mill metal an 1/8 inch wide, then you’ve got another option. Not
great, but it might work. You can get a tool made to enlarge stone
set rings that basically is a micro rolling mill intended for
rolling the metal in a ring shank. These usually come with a set of
various profile shapes for different shank profiles, but can also
handle flat shanks. Generally limited to about 1/4 inch wide or so.
Some resemble a sort of micrometer handle mounted to a base, with the
handle adjusting the rolling pressure. Another I’ve seen actually
looks a bit like a small rolling mill, with rolls on one side and
hand crank on the other side of the base. This sort of mimics the
costly Kagen ring roller, but for vastly less money. Check ebay. The
cheapest of these little tools can be found for less than 40 dollars
or so. And they actually are sometimes useful for more than
stretching ring shanks, such as for producing various profiles of

Hope that helps.
Peter Rowe


I would be very interested in seeing pictures of you pull swedge.
Please post pictures, from several angles if you can. Thanks, David

David Luck

A few observations:

You need to think hard about characteristics of the metals you want
to use in the construction of a rolling mill. Rollers and gears
should be of a hard material to avoid deformation in the case of the
rollers and wear in the case of gearing. Aluminum is far too soft for
either of these applications. Iron might hold up for a while. Do you
intend to cast your rollers, then grind and polish them? Do you have
access to a complete machine shop so you can cut gears? I would think
all of the effort would be better applied to proper steel alloys
rather than iron so hard surfaces may be achieved.

Converting a 1/8 inch fine silver ingot to wire could be
accomplished by drawing it, but a piece 1 inch long is probably too
short to be able to grab it with draw tongs.

Dick Davies

Hello Andrew Jonathan Fine,

I am very sympathetic to your approach. I too prefer to use my own
ingenuity and resources in place of cash whenever possible - not
because i am cheap, but because it is innately and deeply satisfying
to my idea of who I am. That said, I can’t imagine anything you
build, given the time it would take just to design it, much less to
actually execute the design, would be anywhere near worth the time
and trouble when ready-made rolling mills are available for only a
few hundred dollars.

Aluminum for gears? Not a chance. Nor for rollers. Where have you
ever seen that? Not even an ordinary can-opener would survive long
with aluminum gearing. Iron? What you need is steel - especially for
the rollers, and hardened steel at that. Even case-hardened will do,
but iron or mild steel? No. I know you have described your desired
machine as “tiny” but regardless, silver, especially as it
work-hardens, is harder than aluminum. The pressure per square inch
needed to deform silver is still large, regardless of the size of
the workpiece. As you want to make foil, your rollers will have to be
wide enough to span the finished piece. Even as “thick foil” your
1/8" square might yield you a foil of 1/64" (fairly thick for foil)
so your rollers will need to be somewhat wider than 1full inch. That
is not a whole lot smaller than one of the small commercially
available mills so your “tiny” mill is not really that tiny, nor much
less demanding in its strength requirements than a cheap
off-the-shelf model.

It could be an entertaining project if entertainment is what you
have time for - but the idea of doing this so you can earn your way
to something better is an oxymoron. If, after going to all the
trouble of building the contraption, the first thing you do is to
replace it with “something better” then the work was truly wasted.
You’re only talking a very few hundred dollars. Think this through.
I’m interested to see what other advice you get.


Hi Andrew,

Steel rollers are worth the extra effort because after all the
bootstrapping you want something that lasts a few years. When you
see imprints in silver from rolling lace or hairs into the silver,
some work hardened silver will certainly leave imprints on soft iron

The gears need to work at the maximum opening where the tips of the
teeth are only just meshing. Iron gears may do. Case-hardening of
iron rollers and gears will almost equal hardened steel.

The roller diameter determines the gear diameter which in turn
determines the maximum opening of the rollers and also the torque on
the handle. The biggest gear teeth give the widest opening and the
bigger the gear diameter the more torque or the longer the handle
needs to be.

Brass is best for the bearing shells for the rollers and aluminium
will do.

The frame can be fabricated from plenty of mild steel.

In my early days I found a hand wringer from an old washing machine
and had plans for modifying it…never happened. I once used a home
made mill (not mine) and it had severe limitations.

I think you will find that the cheapest commercial rolling mill will
be far better and more cost-effective.

All the best, Alastair

I would be very interested in seeing pictures of you pull swedge. 

The concept is also similarly described in some detail, with how to
make them, in Herbert Maryon’s book.



a note of encouragement - you can make many wondrous things with
your benchtop CNC mill. they are absolutely great - with no limit on
machine shop expense, you can now make whatever you can dream up.

if you want to have some real fun, try your hand and making stamping
dies - the kind you can use in a hydraulic press. proof a pattern in
aluminum, just to see that you have the machining and toolpath
geometries correctly, then go to a ‘heat treatable’ tool steel.
using a piece of steel, say 1/2" think and 1" in diameter, this is
well within the capabilities of a benchtop cnc mill, and also a small
hydraulic press. if you do not have a press, and will be hammering
instead, then use a 1/4, or at most a 1/2" round.

you can have great fun with a benchtop cnc - the rolling mill is
perhaps not the best choice for a project, but there are many
others. I have built a 6’ drag saw (for cutting large jade boulders),
countless fixtures, dops, strange faceting machine angle
attachments, and 3 carving machines using mine. oh, wax dipping forms
(for lost wax casting), a 4th axis cnc table, a slab holder for the
24" diamond saw, a profile grinder to fit on my diamond pacific, a
face cutting indexable grinding fixture, and on and on.

oh, how could I forget this one? i used a small CNC benchtop mill to
build a large cnc benchtop mill.

while perhaps now I am knowledgeable about CNC, when I first got the
mill, I knew nothing about CNC.

So, go for it.
warm regards
Mark Zirinsky

Hello Andrew,

Your situation is a lot clearer now. Thanks for explaining all that.

I’m no expert !!! (that’s an honest disclaimer) But if it is wires
you want out of those small ingots you described - could you draw
them through a drawplate? Are they long enough so that if one end
was tapered (by forging so there is no loss of material) they could
be pulled through a drawplate? Drawplates can be bought ( expensive
but nothing like a rolling mill.) Or, as you have machine tools, I
bet you could DIY a drawplate. You might machine a small drawplate to
suit your needs using annealed tool steel - then harden it and temper
if necessary. That involves boring a few careful holes in a single
piece of steel - no moving parts - no connections. The hardening/
tempering process requires only a torch and a pail of water or oil.
Surprisingly easy to do too. After a near-lifetime of thinking it was
out of my range of abilities I finally did some tempering and it
worked well. I assume you have a torch. Whether you end up either
rolling or drawing the ingots into wire, you need one to do the
frequent annealing of the silver .

I don’t think turning ingots into wire on a lathe is practical.
(Never mind making the lathe itself!) I’m not meaning to be a wet
blanket. Lots of material would be lost and technically it’s a
challenge anyway. A work-hardened 1/8" billet might be stiff enough
to just start machining, but as you reduce the diameter I believe it
would very quickly become unmanageably thin, flexible, twisted -
impossible to make a controlled cut.

Tough choices. There’s probably lots more advice to be had from
folks who know more than i do.


Hi all:

The concept is also similarly described in some detail, with how to
make them, in Herbert Maryon’s book.

Peter, now that you mention it, I think they’re also described in
Bernard Cuzner’s book as well. Still very rare.

Meanwhile, here are a pair of pictures of it. They should be on
the Orchid server at the following two addresses:

Not the prettiest thing I’ve ever made, but it works. The pressure
screw is at an angle because (A) I was 19, and (B) didn’t yet know
to be suspicious of all machine tools and their settings. The table
on the drillpress wasn’t tight, and it slipped off angle as I started
to drill… Thus an angled screw. Since you don’t tighten it under
load, it doesn’t really matter.

It’s rather taller than it really needs to be. The height was
largely a function of the scraps of toolsteel I had laying around.
(Now that I think about it, the size may well have related to the
width of the support pins on the drawbench I had access to back
then.) The body is all O1 toolsteel, and the rollers are W1, but un-
hardened. The pin is O1, hardened to about RC 40. The big block at
the bottom is to give a vise something to grab onto. The two grooves
in the roller are designed to produce rectangle wire with a step in
one side, and triangle wire of a particular size that I needed once
upon a time. The nice thing about it is that you can just grab a
hunk of steel, turn a groove into it, and go to town. Once the
frame’s built, it doesn’t take long to lathe up rollers. (Assuming
you have a lathe…)

Brian Meek.

The roller diameter determines the gear diameter which in turn
determines the maximum opening of the rollers and also the torque
on the handle. The biggest gear teeth give the widest opening and
the bigger the gear diameter the more torque or the longer the
handle needs to be. 

Not quite true.

The roller diameter certainly must match the pitch circle of the
gears, but as the diameter increases the torque required decreases.
Its like wheels - its easier to push (or pull) something with big
wheels than something with small ones.

Regards, Gary Wooding

Firstly in reply to the original enquiry about producing very thin
strip. If you cant afford good rollers or get a friend to roll the
strip order some from your silver supplier, at a price.

I intend to put a series of photographs and text on my blog post
about SWAGE or swedge presses and their uses.I have used this process
to produce numerous sections for rings, bangles, necklets, decorative
edges for bowls etc. from tiny wires to thick wide ridged strips.

The beauty of it is you can use the same swage block for a thick
bangle or ring to a thin strip depending on the number of pulls. A
strong well geared drawbench is essential. Give me a couple of days.

David Cruickshank

Aluminum can’t be used in a rolling mill and it has a blackening
effect on fine silver. You need case hardened steel rollers- you can
buy them from Rosenthal supply, you need a turn bolt to put
euqivalent force on the upper and lower rollers at the same time,
and a gearing system to raise and lower the heads- you’d do better to
invest the cost of the parts in a reasonable rolling mill from FDJ
tools or Rosentahls- while it is tempting to want to make your own
the cost will come out to around 200 bucks before the milling you’d
have to do with iron to make the frame- You could always buy a ring
stretcher with dies that doubles as a mini-rolling mill if all you
want to do is make bezel strip or foil- they run about 80 bucks or
less from Ikohe, FDJ Tools, and many other vendors. MAKE magazine’s
website does have plans and an instructable on how to make one
yourself, but all in all for the expense, time and precision - not
to mention lasting qualities, you would come out far ahead to buy a
mini- mill or a standard size mill appropriate for the job- they are
highly resealable should you decide you are done with the hobby, and
if it’s a matter of cost, just buy the strip and foil. It’s quite
inexpensive relative to even one roller!.. rer


I am not sure I understand how I can replicate your device based on
your drawings, or how to operate it.

I’m more than 350 miles away from a university public library, if
someone pleae can photocopy the pages from relevant chapters and
email them to me, I would most appreciate it.

Andrew Jonathan Fine

I have just completed short notes and photographs of a swage press
and its products. Had a bit of trouble here and there with computer

please see David Cruickshank’s blog on Ganoksin

Hi Gary,

Your analogy is correct for the rolling resistance of wheels, but
when driving the wheels the torque in the axle increases with the
wheel diameter.

In a rolling mill the ratio of the roller radius and handle length
must remain the same in order to apply a similar force to the handle
during use.


Hi Alastair,

In a rolling mill the ratio of the roller radius and handle length
must remain the same in order to apply a similar force to the
handle during use. 

Consider two rolling mills, one with large diameter rolls and one
with small ones. Set them up to roll the same sized piece of metal to
the same thickness. Clearly, the horizontal distance from the point
where the metal first touches the rolls to the point where the metal
just starts to exit is greater for the large rolls than the small
ones. In other words, the squeezing process is more gradual with the
large rolls. Since it is more gradual it requires less force, and
therefore less torque.

Regards, Gary Wooding

In other words, the squeezing process is more gradual with the
large rolls. Since it is more gradual it requires less force, and
therefore less torque. 

The word “gradual” is a bit of a fuzzy word to use in this context,
and if using words to do with time, I would have said that the mill
with larger wheels carries its work out more quickly than the mill
with smaller wheels - not more gradually or slowly.

I would think it has more to do with the principle of moments/levers
than it does time. The further away from the fulcrum (pivot point,
ie centre of the wheel), the less force is required to do the work,
such that the larger wheels are requiring less force and therefore
torque to do the same work.

A rolling mill seems like quite a complex piece of equipment to make
at home, but it’s an interesting thread nonetheless.