Bootstrapping a tiny but effective rolling mill

Hi Gary:

This is a serious bit of trivia, but apparently there’s some sort of
relationship between ideal roll diameter and stock thickness. For
rolling out thick silver slabs, there are industrial rolling mills
with very large diameter rolls, and for rolling out very thin stuff,
there are two different types of mills (4-high mills & cluster mills)
that use a pair of small diameter rolls backed up with either another
larger pair, or a cluster of medium sized ones to support the little
ones and prevent them from flexing under the load. I’ve never used
either of these types, so I’ve never been crystal clear on exactly
how it’s better, but apparently it is. Making a rolling mill with 16
bearings as opposed to 4 is significantly more trouble, so there must
be a reason to do it. There’s info on this in the new edition of
Grimwade’s book, I’m just too tired to go digging right now. (Taught
all day. Brains running out of ears…)

For whatever that’s worth.
Brian

Hi folks,

probably there are better technical terms than what I can offer here

  • but here goes -

The total amount of energy required to accomplish any given task -
for example reducing a particular size and gauge of sheet to a
thinner gauge (and, consequently to a larger size) remains constant.
The RATE at which the task is accomplished is dependent upon the
output available from your arm, electric motor, or whatever the
source of motive power. Measures such as horsepower include BOTH the
total amount of energy and the amount of time needed to do a
particular task. Higher HP does the same job faster. Obviously a
motor can outstrip most of our arms. However, if we are content to
accomplish the task at a slower RATE then we can do the same job by
hand - it’ll just take longer. We can make the effort expended at
any given moment easier if we are willing to take longer. So I don’t
understand what the fuss is about relative diameter of gears and
rollers etc. Any combination of gears can be rigged to turn any size
roller and the effort can be modified by making the crank handle
longer. A small set of rollers will do the job and so will a large
set. It is all just a trade off of time and effort - what is
comfortable and what is tolerable in terms of time expended.

Marty

Hi Gary,

The sharper ‘bite’ of a small diameter roller may be an illusion.
The small roller is feeding less metal per revolution compared to a
large roller.

This is the limit of my engineering as I cannot provide calculations
that include all the variables :(… but I think we are in the same
ballpark :slight_smile: !

My mill (Cavallin) has 65mm diameter rollers, a 4:1 reduction
gearbox and a 340mm handle. Without the gearbox it would either need
a handle 4x longer or it would only manage a quarter of the bite. I
love it because it can take large stock and huge bites when needed.

For the smaller tasks I made another handle half the length because
it became tedious to swing the long handle around with hardly any
resistance. The short handle is easier to spin around for light work
(about 70% of the time) and I fit the long handle when there is some
big stuff the mill down.

The home-made mill I used years ago had rollers about 25mm diameter,
a short handle, and the depth adjusting screws were separate at each
end. The shortcomings were; a) that the mill could only take tiny
bites due to the short handle deliberately made so in order to match
the frame and roller strength; b) each new bite required adjusting
both ends of the roller exactly the same to maintain a parallel gap
between the rollers; and c) the max opening was very small. I used
the sledge hammer and anvil frequently in order to get the stock
into the mill.

In fact hammers and anvil along with swages are the cheapest 'roller’
by far! They require more skill to be accurate but are extremely low
tech and versatile.

Regards, Alastair

A small set of rollers will do the job and so will a large set. It
is all just a trade off of time and effort - what is comfortable
and what is tolerable in terms of time expended I stand corrected. 

I drew a diagram (see http://tinyurl.com/l5nlbu) to illustrate what
I thought would be a convincing argument. It certainly showed that
the squeezing on big rolls took place over a longer distance, but,
what I hadn’t allowed for was that the longer distance corresponded
to a smaller rotation angle.

So, since the same amount of work was done with a smaller angle of
rotation, the torque must be greater.

Umble pie doesn’t taste nice, but its good for me.

Regards, Gary Wooding

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. 

I realize this is a somewhat old thread, but I’m getting caught up
and it’s pretty interesting in it’s way… I admire anybody who wants
to tackle big problems… A couple of observations, though.

A centerless grinder isn’t optional, it’s essential unless you want
textured metal of uneven thickness.

The fact that aluminum rollers and gears is -errrr - less than
satisfactory has been addressed, I hope.

Small rollers are used for a variety of reasons, but the main one is
surface speed, which lapidary people understand quite well. A large
wheel turns slower than a small one at the outer rim when spinning
at the same RPM - surface speed of wheels. But then the smaller
rollers need backing rollers for strength against the stresses or
they’ll deform. That’s how you get high-speed rollers with a minimum
of fuss…

You can’t make good strip on a rolling mill anyway. You can make
pretty OK strip if your mill is good and you know how to use it, but
stip and wire will always come out of the mill crooked. The mill and
the user just aren’t that precise…

Swaging has been mentioned, which is usefull. Commercial mills use
slitters to make strip. They get a big sheet and slice off strips
from the edges and put them on rolls, nice and straight…

Go ahead and make one if you like - my guess is you’ll spend twice
as much and ten times the effort to get half the mill…

Hi All:

I got to pondering just why one would go to the trouble to make a
cluster mill, as I didn’t know. Still don’t, precisely, but I found
a really interesting animation illustrating the working stresses both
in the rollers of the top half of a 12 high cluster mill, as well as
the strain within the top half of the billet.

Take a look here:
http://www.adina.com/rolling.shtml

Nothing quantitative, but interesting nonetheless. From what I
gather from the simulation’s stress levels, cluster mills seem to be
a way to get either heavy bites on softer materials, or medium bites
on harder materials. It seems to be a way to distribute the stress in
such a way that the working roller can be both very hard, and fully
supported.

FWIW,
Brian.

Been following this thread with some interest, hoping that someone
else might have ideas/drawings of a mill. I have a “someday” project
for a larger mill with rollers perhaps up to 12" long, rather then
the present 3" standard of commercial mills.

I have bought a cheap Indian mill which worked well as long as you
are gentle and take that extra step or 2. Which was a problem for me
because I don’t have that sort of patience, I broke 2 rollers and a
bearing. Fortunately being a Metalwork Teacher, I have access to
machine lathes and a milling machine, I was able to fab up new
parts.

What I’ve learnt is that:

1: You need a strong frame to handle the pressures.

2: You need the rollers to be steel that at least is case harden. The
Indian mill rollers were harden all the way through and broke.

3: The depth of the drive gear teeth determine the working opening
of the rollers.

4: The mechanism that raises and lowers the top roller must allow
some way to adjust either side so that what you you roll comes out at
an even thickness. If and when you look at a commercial mill you will
see what I mean.

5: The rollers have to be of a size to prevent deforming. Which
means if the the rollers are to small they will bend in the middle
and produce uneven thickness to your work.

What I have determined is that for a standard small mill, buy one.
The cost of making one will be pretty close to the cost of a new
Indian mill (for under $300 if you look around). And as I said if you
look after it, it will serve you well. If you are like me and can see
using a larger mill for design as well as production use then it
would be worth while to make one.

Chris Gravenor
http://www.northernlightsdesign.ca

Hi all:

In my continuing search to educate myself about cluster mills, I
found another interesting diagram of a “Z-high” mill. It seems the
point of a cluster mill is twofold.

(1) to allow large single-pass stock reductions by way of using
super hard working rollers supported by a swarm of intermediate and
backup rollers.

(2) to allow one to control the thickness and cross section of the
rolled stock very precisely, by way of reducing the flex of the
working rollers far more effectively than a traditional 2-roll
design can.

The diagram is here:
http://www.sendzimir.com/Z-HighMills-1.htm

Looking at it, the interesting thing is the taper turned onto the
outer ends of the intermediate rolls. This will cause the working
rolls to flex quite deliberately. The idea being to reduce the
‘bulge’ in the center of the rolled sheet. I’ve noticed that with my
hand rolled sheet, the edges of the billet are usually thinner than
the center, and the front and back bit are thinner as well. The cause
of this is the rollers flexing under the load. With a cluster mill
with tapered intermediate rolls, that flexing is eliminated, so the
stock is of equal thickness all the way across. This also apparently
reduces edge cracking and feathering, which is an issue when doing
the large single-pass reductions these mills seem designed to
provide.

These sorts of mills are large industrial scale things, not something
most of us are ever going to have to worry about. The designs are
interesting to know about nonetheless. (at least for a tool geek like
me.)

Thought I’d share.
Brian.

But, Gary, that was a neat diagram!

And, having drawn it, you immediately reaped the benefit of seeing
the reality and dealing with the reality, instead of stubbornly
remaining committed to the idea in your brain. Ideas can be sticky
and hard to dislodge without occasionally resorting to extreme
measures, like making a picture, for example.

picture worth a thousand words.

Marty

Nothing quantitative, but interesting nonetheless. From what I
gather from the simulation's stress levels, cluster mills seem to
be a way to get either heavy bites on softer materials, or medium
bites on harder materials. It seems to be a way to distribute the
stress in such a way that the working roller can be both very hard,
and fully supported. 

The main reason for cluster milling is that with a smaller diameter
roller contacting the ingot, a very heavy pinch can be taken because
the angle of contact (the small diameter) localizes the heat to the
smallest possible area allowing milling in a semi plastic state
preventing cracking, splitting, and stratifying. Forming the ingot
hot can improve the crystalline structure substantially making up for
any number of pouring problems. Of course a small diameter roll isn’t
strong enough to roll with that much pressure thus the reason for
backing. Cluster rolls of course are very expensive (often over
$50,000.00) and the rolling mills jewelers use are a form of
compromise. In industry to break down an ingot for rolled product,
one mill is used, and then other mills with different arrangements
are used to make finished products. We don’t have that option. Some
companies have large blocks of steel over the top roll to prevent
finger rolling and this does in some cases add enough stiffness of
the top roll to equalize with the bottom roll and help make “more
even” very thin strip. Good rolled product REQUIRES a perfect setup
but everything bends, stretches, and warps so compromise is the only
way.

Good luck
Dan
Daniel Culver

Been following this thread with some interest, hoping that someone
else might have ideas/drawings of a mill. I have a "someday"
project for a larger mill with rollers perhaps up to 12" long,
rather then the present 3" standard of commercial mills. 

Sorry, but whether you make it your self or buy it used, a proper
12’’ wide two high mill will be over $25,000.00. New ones are over
$60,000.00. I once saw one on ebay go for around $9,000.00 but it was
an old machine and, though in very good condition, it was going to be
a $15,000.00 complicated move to go less than 100 miles. I would
guess it cost about $30.00 just to turn it on.

Daniel Culver

Roll diameter is a complex thing and I’m not sure you have it
exactly right… A large roll is in contact with more metal than a
small roll and is therefore having to squash the metal over a larger
area. This requires more work and more friction than the smaller
roller - its like if you try to squash a lump of clay with your flat
hand or if you poke your finger into it…

Ian
Ian W. Wright
Sheffield UK

This also apparently reduces edge cracking and feathering, which is
an issue when doing the large single-pass reductions these mills
seem designed to provide. 

Sorry Alberic,

Reducing the flex and the accompanying center bulge doesn’t itself
reduce edge cracking or splitting. The heavy pinch and high heat
caused by the small diameter contact roll allows semi plastic rolling
and reformation of the crystalline structure producing a superior
rolled product. An experienced metal spinner will tell you that if
you are cracking the product, increase the pressure, to increase the
heat. Same principle; though easier said than done.

Dan
Daniel Culver

Hi John,

Small rollers are used for a variety of reasons, but the main one
is surface speed, which lapidary people understand quite well. A
large wheel turns slower than a small one at the outer rim when
spinning at the same RPM - surface speed of wheels. 

Sorry, I think you misspoke John.

If you do the math, you’ll find the ft/in per minute speed of a
point on the surface of a 10" diameter wheel is 10 times faster than
a 1" diameter wheel at the same rpm.

The circumference of the 10" wheel is 10 X Pi = 31.41", the
circumference of the 1" wheel is 1 X Pi = 3.14". A 10" abrasive wheel
would present 10x the abrasive particles to work piece that a 1"
wheel would assuming they’re both turning at the same speed. However
the the 10" wheel would take more power to turn it than the 1" wheel
would.

Dave

I have a "someday" project for a larger mill with rollers perhaps
up to 12" long, 

Chris, that’s an ambitious ambition - I’ve said before I admire
people’s ambitions when they are reasonably attainable. I’d suggest
you consult an engineer for your 12" rollers, though. I’m not in any
way such a thing, but I’d guess you’re looking at 500 pounds of tool
steel, at least 1000 pounds of frame and something like a 50
horsepower gearmotor. Those are some intense stresses… Not to
mention the machining. But invite me over for a look when you build
it,too… ;}

http://www. donivanandmaggiora. com

Really! I too have seen larger mills on ebay which are used
industrial forming etc., not what I had in mind. I was thinking of
following the basic design for jewellery mills but increase the size
maybe 3 or 4 times. I believe I can do it for around $500 for
materials, perhaps more for good quality round stock of induction
steel up to 4" in diameter.

Will it work? Perhaps not, that’s why I said its a someday project. I
do know that most of the pieces (using my design) can be got over the
counter and like I said I do have access to a machine lathe and
milling machine as well as welders and a plasma cutter so I can do
most of the work myself.

Yeah I could be an idiot and it won’t be the first time.

I think people look at larger mills that are on ebay or the net and
see this big machine with huge wheels large motors etc. That machine
is used to form plate steel (probably cold rolled process), where
indeed the stresses are enormous. There is a big difference in the
amount of pressure (stress) with the materials we use. Also we can
reduce the stress by how much we’re reducing per pass.

As I said in an earlier post, my concept is to just use the basic
design of present jewellery mills and increase the size 3 or 4 times.
The rolls would be 3 to 4 inches in diameter and up to 12 inches long
(plus extra for the bushings etc.). The frame would be 1/2 inch plate
and the bushings would be case iron or brass. Yep it would be a heavy
thing, maybe 200 to 300 pounds.

I haven’t taken this idea out of my hat, I have some basis to
believe this would work. Before I had my present mill I used a slip
roller (used to make sheet into cylinders and cones) with admittedly
mixed results (using the wrong tool syndrome). It did show me that
that it could be possible to make a machine for what I want it to do.

Which comes to the question on why I would want a bigger mill. The
direction that I’m going in is Mokune Gane and a process which I
call Fusion (where I fuse 2 or more metals together). The problem (in
my view) with both process is that the initial billet no matter what
the size takes the same length of time to produce no matter what the
size is. Again in my view, by increasing the size of billet you
increase production.

Experience and readings both show that with Mokune Gane (and indeed
with my fusion process) that great care must be taken NOT to stress
the billet to much or it will delaminate. Indeed we could apply this
concept to any project that we would use the mill for in jewellery
making.

So to make a long post short, I am not trying to make a mill that
can be used in industry to form plate steel, rather I’m wanting a
machine for particular processes where you would want the stresses to
be low. My idea may or may not work or it may take a few tries but
that’s fine, I’m having fun and learning things.

Now to find the time to make the damm thing.

Chris
http://www.northernlightsdesign.ca

I got to pondering just why one would go to the trouble to make a
cluster mill, as I didn't know. Still don't, precisely. 

One of the reasons you’d see a robust mill such as a Z-mill, or any
type of mill backed by very large diameter rollers is the need to
maintain thickness and flatness within a narrow range of tolerances.
While we jewelers may work with fine dimensions, we don’t attempt
thickness within thousandths, or ten thousandths, or overall
flatness within the same. It is acceptable for most of our
applications if our thickness of sheet when we roll it is close, if
for example you want 18 ga, 17 to 19 ga will often be ok. In many
industries that will not work.

The thicker the rolls supporting your anvil, the less flex it will
have, ensuring more consistent thickness and flatness.

Mike DeBurgh, GJG
Henderson, NV

I think people look at larger mills that are on ebay or the net
and see this big machine with huge wheels large motors etc. That
machine is used to form plate steel (probably cold rolled process),
where indeed the stresses are enormous. 

A mill for steel is truly huge, like building filling size. Even on
soft materials like copper the separating forces are huge. A 4"
diameter is more suitable for a 6" wide roll.

James Binnion
James Binnion Metal Arts

If you are planning to make a mill with a very wide width I would
suggest you do some engineering calculations regarding the
deflection of the rollers. The deflection will be determined by the
stiffness of the material and the dimensions of the roller (beam).
Without the proper increase in diameter you’ll end up with a greater
center thickness as you roll your ingot.

Mike DeBurgh, GJG
Henderson, NV