Rolling mills square wire grooves

Greetings all.

Speaking of cheapo rolling mills, I have a comment and a question.

Does anybody know why the square wire rolls in every mill I’ve ever
seen (dozens) have flat bottoms to the square wire grooves? (Thus
giving ‘square’ wire with the corners beveled off.) This holds on
cheapo Indian mills, Durstons, or those green German ones. Clearly
there must be a reason, but what is it? Normally I don’t care, but
I’m working on some square wire based Mokume things, and it’d be a
whole lot easier to get to my final form if I could get sharp edges
on the wire. (I tried rolling it through the flat rolls to give me
sharp edges, and that works, but it tends to go trapezoid, and then I
end up beveling a corner as I knock it back to square.) (Remember:
This is mokume. Twisted mokume. It really doesn’t want to do
anything evenly, especially rolling.)

Which leads me to a comment about the Indian mills. I have several
real rolling mills, including one that I converted from a dead
powered mill, and machined up new rollers for. (Just for reference,
the rolls on that were case hardened.050" deep, to RC 55-60. ) I
ended up picking up one of the cheapo indian mills at a garage sale
for…not much. The handle had snapped off, probably before the
original owner got a chance to use it. So it really does not
impress me for quality. But for cheap, for something that I don’t
feel guilty re-machining the rolls on, yes. (This one is a ‘Moon’,
for whatever that’s worth.)

Since it has square wire rolls, and I already have a real square wire
mill, the thought floating through my head would be to re- machine
the grooves in the square wire rolls so they had full 90 degree

Has anybody ever used a mill with full square rolls? Anybody know why
they’re usually not?

Brian Meek

Has anybody ever used a mill with full square rolls? Anybody know
why they're usually not? 

Stress cracking. Rolls are under tremendous strain in use, sharp
corners will focus the strain and the rolls will crack at the tips
of the “v” groove.


James Binnion
James Binnion Metal Arts



Those “flattened” corners on the rolling mills wire grooves were
originally designed, I believe, to relieve the stress in the wire
formation. Forming square wire out of a round ingot, the normal way,
would put a lot of stress on a perfectly square corner, and you’d
probably see many cracks form as you made the square wire.

However, those flattened square grooves do serve a wonderful
purpose. Next time you are making square wire stock, say you are
flattening a square wire to make a rectangular-shaped wire, and your
wire ends up a little wider than you wanted, you will have a way to
make it narrower. After annealing your too-wide stock, put that
rectangular stock vertically in the largest groove that it will fit
in, and gently roll it through the mill. That flattened bottom of the
groove will put a flat edge on your stock, and narrow it as well.
When you have the right width rolled down, then gently roll the wire
flat again on the flat roller, and you have a perfectly flat,
straight stock with perfectly flat edges. When you practice it, you
will do this operation in minutes. Don’t forget to anneal!

If you want a perfectly square wire with sharp corners, just make
your square wire with the grooves, normally, but make it a shade too
thick. Then, take your wire and gently run it through the flat
rollers, alternating flat sides, until you have squared up the stock
to the right size and squareness. The trick is to put very l;ittle
pressure on the rollers for each pass. Works well.

Good luck!
Jay Whaley

1 Like

Hi Brian, I’ve often asked myself that question but have assumed
that this is engineered in. A sharp square corner seems like it would
be much more prone to generating the small cracks that grow if
unattended into larger, deeper cracks and then eventually splits…
It’s like the rounded inside corners on cast iron castings. The wider
surface spreads the energy out rather than funnel it in to one point.

This could, of course, be totally wrong. We’ve all pushed the wire
mill a bit too much-- right before “finning” happens-- and then
repeated this in a 90 degree turn, yielding truly square stock…

Take care,
Andy Cooperman

Hi Brian,

if you have a groove in your mill that has a sharp corner it makes
it more likely that you will roll over a corner to produce roaks.
(which is what I was told those thin strips that appear later out of
the edge of badly rolled square wire are called).

I have always assumed that you are supposed to finish with a square

I also understood that to make wire you started with a square strip
off thick sheet, not a cast round rod, owing to the way the metal
was distorted, ie more distortion gave you a better grain structure,
so perhaps the mill is designed to turn sheet into round wire and it
makes octagonal wire as an intermediate step.

Anyone worked at a bullion dealers care to comment?

Tim Blades.

Does anybody know why the square wire rolls in every mill I've
ever seen (dozens) have flat bottoms to the square wire grooves? 

I would guess that it’s because most people are going to make round
wire and by “cutting off” the corners, 75% of the work is already



Two reason for flats in the bottom of the square wire forming

  1. If the forming groove were fully vee formed the wire wouldn’t
    roll out with square corners but rather a wavy edge that was
    fractured continuously. The procedure is to roll down to a size just
    over your final size and proceed with alternating rolling to get a
    square wire. I believe details are in the archive. The reason a full
    groove would roll wavy damaged corners to your wire is that rolls
    never actually line perfectly and don’t roll at exactly the same
    speed, if they did then this would indeed be possible. A specialized
    mill would have to be built.

  2. Cost. Any machined vee groove will have to have a flat or radius
    in the bottom of the groove. Pick your size and then your price. I
    would guess that with a .010’’ (about 1/4 mm) radius properly ground
    and hardened in a decent set of rolls (not the whole mill) would cost
    you about $4,500.00 and maybe much more. A .005’’ radius could go
    over $7,000.00.


the thought floating through my head would be to re- machine
the grooves in the square wire rolls so they had full 90 degree

How about drawing the not-exactly-squared wire through a
square-holed drawplate after you’ve milled it?



These are cut on Turning Lathes. The Tool bit is not narrow and
pointed you cannot be changing the tool bits so often either.

Even if one makes a 90’ on a roller and say with a stroke of luck
matches the top and bottom what happens to the other two corners of
the wire that are now between the rollers.

Rolling mills are not for perfect square or round wires. You have to
use draw plates.

Kenneth Singh
46 Jewelry Supply
Karat Rolling Mills


For what I have seen, granted it was high speed, but the metal sticks
and cracks. If you can machine your own rollers, give it a whirl,
might work fine slow. But the engineer working on that project had to
redo the rolls to add a radius to the design. He was not a mechanical
engineer, but electrical, maybe a mechanical guy will chime in and
tell us both why you put the radius in.


Greetings all.

Thanks for all the replies, much appreciated.

Once I got my brain on straight, Jim (et al) are correct: stress
cracking, both of the rolls and of the wire seems to be the best
answer for why that’s a bad idea.

This was mostly a case of getting annoyed while arguing with a
billet, and having time to scheme and ponder as I was rolling it

The billets in question are on the order of .625" thick to start,
and the end goal is something closer to .25" so a drawplate was just
right out. (at least for the valentine’s day present I was working
on.) (Now done.)

Again, many thanks,
Brian Meek.

Does anybody know why the square wire rolls in every mill I've
ever seen (dozens) have flat bottoms to the square wire grooves? 

Lots of other answers have suggested various reasons, but I suspect
the best answer has not yet been given, or if it has, I missed it.

When you roll wire, you’re going from one size, and then to the next
size down in the next groove. The rolls only compress the wire
vertically, reducing the vertical or roll-to-roll dimension. they do
not make the wire any narrower in the horizontal dimension. In order
to progress from one groove to the next smaller size groove, the wire
has to be no wider than the new smaller groove, or it will pinch in a
nasty flange along the sides. The flats at the bottom of each groove
mean that after you roll the wire through a groove, the measurment
top to bottom, across those two small flats, is smaller than the
width of the wire measured side to side. Those flats are sized so
that this difference equals the reduction in groove size between one
groove and the next. So you turn the wire 90 degrees before going to
the next smaller size groove, and the flats now mean that the width
you’re putting into the groove matches the width of the new groove,
but the new smaller groove rolls the old wider dimension thinner,
whilst putting new small flats on, ready to be turned again 90
degrees for the next smaller pass. If you turn it 90 degrees and put
it through the same groove, you now end up with a cut corner square,
which is much better for subsequent drawing than a fully square
cornered wire would be. It doesn’t need too many passes through the
appropriate plate to become either fully square in a square draw
plate, or round in a round plate, etc.

But the main thing is simply producing a rolled wire with dimensions
that physically can be fitted entirely in the next smaller groove.
Fully square cornered rolls would simply not work in a two rolled
rolling mill. The so-called “turks head” drawing fixture can be made
with rolls instead of adjustable friction plates, and I saw one once
that was an actually motorized rolling mill, with four small rolls
arranged to produce an adjustable dimensioned square or rectangle
sharp cornered wire. Because all dimensions were being reduced
simultaneously, no flanges were being produced. But the design, at
least the one I saw, was difficult to adjust and operate, and quite
slow, and could not handle a large reduction in size per pass… It
was intended more for producing a single precise dimension of wire
rather than reducing a larger wire progressively smaller…


1 Like
But the main thing is simply producing a rolled wire with
dimensions that physically can be fitted entirely in the next
smaller groove. 

I didn’t know the reason for this either, though I guessed most of
it once the question was asked. I suspect that the real reason for
the flats is “all of the above” that has been posted - Peter’s
thought, above, and the cracking and wavy edges and all the rest.
Just the best way to design it.