Back to Ganoksin | FAQ | Contact

Ripples in metal

Hi all: I have a problem and I know someone out there has the answer.
Recently I rolled a sheet of 10 guage sterling to 16 I annealed it
twice during the rolling allowing it to air cool before picklling to
remove the fire coat .When I went to use the sheet I found ripples
running through the sheet in a vertical direction. These have been
very hard to remove,I checked the rollers and they are not out of line
or to my eyes have any dents in them. Any suggestions Thanks

You may want to try, placing the silver between two flat peices of
wood and pounding on it with a mallet… you may have to do this a
few times… but it usually works for me.

Good luck AW

AW Thanks I had tried between the bench block and anvil to no anvail
( sorry couldn’t resist ) but will try the wood Thanks again

i have been able to avoid ripples by continually rotating my piece
when rolling it out. Also I turn it upside down every other pass.
Think of your piece as being North, south, east and west. I mark it
with a permanent marker at one side so that I know where I am. I
put the north side through first, face up. then rotate it so that
the east side is put through face down, then the south face up, then
the west, face down. then back to the north, and the process is
repeated, with annealings in between. It seems to work for me as I
am getting less distortion following this method.-

G’day; for small pieces of sterling, when I get ripples in it I make
rude noises and remarks to the rolling mill. This doesn’t improve
the process or flatten out the ripples a bit but lowers my blood
pressure. My next trick is to carefully anneal and pickle the work,
then adjust the rolls on the mill to put a very little pressure on
the work. I then enter the work between the rolls at 90 degrees to
the previous direction used - sideways in fact. If the ripples aren’t
gone, I repeat the bad words incantation and repeat the rolling with
a tiny bit more pressure. This I admit doesn’t always take all the
ripples out without altering the thickness of the work, but it does
help. Anyone got a better idea? Or a plausible explanation of WHY
we got ripples in the first place? Cheers - and I am looking forward
to some ripp(l)ing good yarns. –
John Burgess; @John_Burgess2 of Mapua Nelson NZ

   Or a plausible explanation of WHY  we got ripples in the first

John, In answer to your question The reason the rippling or wave action
occurs is due to the gear ratio of the mill. For example a 3 to1 gear
reduction will give you 3 ripples for every revolution of the
cylinder. One would be high whilst two would be low or vice versa
depending which way you hold the piece :-). A 1 to1 mill will give
you a long curve verses the up and down wave action as seen with a
3-1. Best Regards Neil George

Hi John, I’ve not to much experience working in silver (don’t like it)
but generally find getting the piece red hot and rolling out whilst
still hot seems to work o.k. Regards Andrew

At one time, I took 2 plates of 1 in. wide X 1/4 in. thick X 12 in.
long pieces of polished stainless flat stock and sandwiched my silver
between the 2, then gave the metal a light roll through the machine
and out it came on the other side … perfectly flat and quite
Daniel Grandi

John, In answer to your question The reason the rippling or wave action
occurs is due to the gear ratio of the mill. For example a 3 to1 gear
reduction will give you  3 ripples for every revolution of the

Yeah; g’day; Oh dear, this is a bit awkward! You see, the rolling
mill I use regularly is direct drive; that is the 18 inch long handle
drives the bottom roller directly without any gears at all!
Furthermore, I get ripples in a piece of sterling only 4 - 5 cms long
2.5 cms wide and 2mm thick. Which means the rolls only turn around
half to three quarters of a turn at the most to roll the whole length
of the strip. So I find the above suggestion not very plausible. Tell
you what I’ll do: next time I do a roll job I’ll measure the
wavelength of the ripples; perhaps we can get at it that way? Doing
a final anneal and putting the piece in sideways with only a little
pressure a couple of times helps, but isn’t the whole solution.

I once worked in the Mechanical Working Division of the British Iron
and Steel Research Association, and although rolling wasn’t part of my
own job, I did notice they got wavy steel sheet quite often with cold
rolling. I have an idea they were puzzled too. They would talk
glibly about ‘stick-slip’ but couldn’t really explain it.
Occasionally the wire drawing folk would get 'string-of-sausages’
wire - which they too called ‘stick-slip’! However, my days of
meticulous research are long over! Like I said, I dunno.

John, I beg to differ, but what you are telling me is actually a
variation to the problem in your case.

   You see, the rolling mill I use regularly is direct drive; that
is the 18 inch long handle drives the bottom roller directly without
any gears at all! Furthermore, I get ripples in a piece of sterling 
only 4 - 5 cms long 2.5 cms wide and 2mm thick.  Which means  the
rolls only turn around half to three quarters of a turn at the most
to roll the whole length of the strip.  So I find the above
suggestion not very plausible.

With you having a direct drive, it is in actuality a one to one ratio
which should give a long curve, but in saying that, you have another
issue with your mill. Without seeing the mill you have, I am assuming
that the top roller is technically freewheeling and is driven only by
the rotation of the bottom roller which utilizes the metal contact as
the motion provider. The key to your problem, is that because the top
roller relies on the bottom roller for motion, you are getting an
inconsistent simultaneous revolutions of both rollers, which offer
extremely small variations in rotation which in fact form ripples. If
you want to call it a slip then we will call it a slip. There are
only 2 ways to eliminate the ripples and to have a long curve in a 1
to 1 ratio. First is to have both rollers driving and taking up the
work load simultaneously and synchronized, or secondly and preferred,
is to have a flat bottom bed that does not move, and therefore, only
the top roller is moving and taking up the work load whilst providing
a consistent down force onto the metal without variation. If you think
about a single roller moving back and forth along the surface of a
strip of metal on a flat surface, the work hardening of the metal is
going to move that metal in a certain direction to start forming a
curve. It will curve, but will not cause a wave action. On a 1 to 1
twin drive where both rollers are driving, the roller that applies
the force is the one that will determine which way the metal curves.
This is why a mill/foundry use a single roller and flat bottom with a
direct power drive to produce their stock. On some pieces if
necessary, they will flip a long piece of stock to induce a curve in
the opposite side to maintain flatness. Now with a 3 to 1 drive ratio,
it is actually translating that ratio to the metal. A round surface
in contact with metal does not create a flat, it creates a curve. Does
this mean that two rollers form a wave?, On a 1 to 1 with twin
driving rollers the answer is no, but it will produce a curve. On a 3
to 1 yes. Pay attention as you rotate the handle on a 3 to 1 and you
will see that the handle at 12 o clock will form a specific ripple in
the same place every time. Try measuring the length from the crest of
one ripple to another, and work out your data from the roller
circumference and see what you get. You will see it corresponds.

When I buy steel for tooling and molds, it is of extreme importance
that I match the width of the rolled material to the longest part of
the mold. For example, if a mold is 6" x 4", then I will buy 6" wide
stock and whatever length I need for the job. Why is that important
you ask, well the reason being, is that the grain of the material is
going in the direction that it was rolled in, therefore, when a
cavity is cut and stress is induced into the metal during machining,
it will curve no matter what in the direction of the grain. By having
the shortest side of the mold going with the grain it will make it
harder for the material to bend in on itself. Therefore this makes the
longest part of the mold more stable.

If nothing else it makes for an interesting thread. Let me know your
thoughts, as I value and respect your knowledge base appertaining to
many facets of discussion.

Best Regards.
Neil George

Hello Ron

Just as a note,make sure that your sheet is totally dry before you
start rolling it down!! Depending on what kind of silver sheet I need
(soft,medium or hard),I don’t anneal silver to often.If you do so,you
would create a big crystal structure of your silver giving you the
same kind of failure. Don’t apply toomuch pressure at ones and it may
help a lot to change the milling direction AFTER your annealing
procedure.I think this is your biggest problem anyway if I may say

Regards Pedro

Haven’t we had a lot of good comments on this subject, and though, I
would like to throw in my two are/(Danish ‘cents’) worth here.

All of us who regularly use rolling mills surely have experienced
these d… ripples and also have wondered how/why they ‘happen’.

To my humble opinion they are due to the fact that you are applying
your force to one of the booms and that the applied force then is
transferred to the other boom by some gear wheels. If you look at the
action of a set of gear wheels yoiu will see that at a certain time
the inner part of the tooth of the driving wheel will engage the
outher part of the driven wheel and vice versa.

I don’t think that it matters whether you have a 1:1 or 1:3 drive
ratio. As John from NZ so wisely said, the distance between the
ripples does not really fit with the drive ratio figures. Anybody got
a better explanation?

Kind regards
Betty & Niels Lovschal, Jyllinge, Denmark
phone (+45) 46 78 89 94

I’ve had the problem of ripples in the metal while rolling. A few
things I remember about it was, I was in a hurry and didn’t anneal
well enough, and I didn’t rotate the metal after each pass. I know
the rippling appears after I use too much force from either not enough
annealing or adjusting the roller too far downward between passes.
Haste always ends up costing me more time than if I didn’t rush. Steve

A comment and observation on this question of rolling-mill-ripples and
where they come from. Actually, if you take a piece of sheet and pound
it with a hammer on an anvil, worked without annealing, it will
eventually develop a tendency to concave slightly when struck and
become springy. I assume most all of the metal artists have
experienced this effect to some degree, intentionally or otherwise .
It seems to me from my own applications that when the tendency towards
ripples in sheet develops from the rolling process it is a very
similar symptom, and I believe that it possibly occurs from not
annealing the metal early enough in the course of rolling.

If the metal is reduced with sufficient pressure at each pass it will
require annealing at the proper point, usually at a reduction of 50%
in thickness, or an increase of 100% in length. This is a general rule
of thumb although it may vary considerably depending on the specific
metal/alloy being rolled.

In my opinion, the most likely cause of these ripples comes from not
reducing the mill enough on each pass, and subsequently developing a
work hardening tension from passing it through the mill too many
times either before or in-between annealing. This is evidenced through
the attempt to anneal the metal and pound it flat again, if it is
difficult to flatten after a typical application of annealing, it
might help to anneal it a second time before getting too frustrated
with the resistance to flattening.

Try to insure that you are bringing the metal to the correct
annealing temperature and color and that you holding it there long
enough to be effective given the size and thickness of the item you
are working. It may be helpful to darken your studio by closing the
blinds and shutting off some of the lights which will assist you in
making this visual determination.

Once you have succeeded in rendering the metal annealed and pounded
flat which quite possibly could take several attempts depending on
just how hardened the metal has become, as you again start milling try
to cultivate a sensitivity to how much the metal reduces with each
pass through the rolling mill and how much it stiffens. Over the
course of time you will notice that determining the proper time to
anneal is intuitive more from a tactile sensation of the metal in
your hand for the brief moment in-between each pass in and out of the
mill than it is from any visual observation.

I have learned that efficient milling of both sheet and wire requires
timely annealing, I often anneal wire after pulling only 2 or three
holes if it is large diameter, and when fabricating my tubing if it is
thick walled I sometimes have to anneal after pulling each hole. (This
is working with 18kt yellow gold):>)

I hope this helps,
Prosperity to all from your earnest efforts.

Dear Lovschal, Closer than you think :slight_smile: I did not think it would get
this deep, but here goes. You are 100% correct in that you get ripples
with a 1-1 ratio that has a gear, in fact any mill with gears has the
problem, I assumed that John had a freewheeling roller on top. A 3-1
mill has more teeth per say on the gears which therefore allows for
more of that misalignment to translate to the metal. If this was a
belt or clutch driven roller you would get a long curve. The metal
does pick up the transmission of power from each tooth on the gear and
really does translate the pattern to the metal only more so on a 3-1
and I will prove it to you. The jewellers type of rolling mill is
notorious for this because of the lack of precision in aligning the
teeth especially when depending on how wide apart the rollers are.
This distance of separation will either decrease or increase the
problem because the teeth do not come into full contact with each
other until the rollers are fully closed (unless you happen to have a
spring loaded one which does help a little). There are mills out there
that use hydraulic power, yet the pulsing of the fluid going through
the pipes to the rollers also translate that pattern to the rolled
metal. Gears are not noted for being used in precision application
and for this reason, it is why a clutch type power mill is better
because of the smoothness in power transmission. If anyone has used a
hydraulic winch to pull anything like a boat out of water for example,
you will note that by holding the rope tightly in your hands you will
feel the pulsation in action.

This is a basic engineering problem and is best clarified in showing
reasons why in other applications.

  1. All precision finishing lathes that require a high surface finish
    are all done with a clutch or belt driven power systems. You will not
    see a geared head finishing lathe.

  2. High speed CNC Mills have a Direct Vector drive spindle which has
    no gears or even better an air driven spindle for high RPM which again
    has no gears. Only large roughing mills have a geared head to enable
    Hogging or in other words, the removal of a huge amount of metal in
    applications that require a lot of torque, but again, never for

  3. Printing machines which believe it or not, requires very high
    tolerances of precision, so even though there are gears in the system,
    the sections that need precision have cone clutches in place and not
    gears. Paper mills have the same issue of translating transmission
    power onto the paper, this is again why they use a cone clutch because
    of the smoother ride.

  4. Sailors in racing yachts had problems for years in the necessity
    of high speed delivery of power to trim the sails etc. They did
    succeed however, in building a geared device that does the job
    extremely well, however take a look at the price of these precision
    winches. The life of these are minimal in that as soon as there is any
    wear they start failing and therefore need replacing. This is I guess
    is a slight contradiction, but really says that there is a small
    solution for a high price.

For sure someone will say what about watches, don’t they have gears?

Best Regards.
Neil George

This is a common problem with most flat stock. The thicker the metal
the more common the problem. When metal passes through the mills it is
only in contact with a very small surface area of the top and bottom
rollers and is being squeezed to reduce thickness. i.e., the metal
leaving the mill is harder than the metal entering the mill (work
hardening). The metal reacts to the change in hardness and reduction
in thickness and “moves” due to the stresses of cold working.

If you continue rolling in the same direction the problem may get
worse. Depending on the length of metal and the width of the flat
rollers, ripples can usually be removed by passing the metal at an
angle, say up to 45 degrees to the previous pass. Alternate and adjust
this angle on subsequent passes until the metal flattens out. Care
should be taken not to reduce thickness too quickly. The adjustment
applied to a 1:1 mill should also be applied to 4:1 and powered mills.
This avoids over stressing the metal.

Points to be considered to preserve metal during rolling: * Cold work
metal prior to annealing. The common work range is 30% to 70%
reduction between annealings. This is a generous work range to cater
for soft and hard alloys. Rose golds and nickel white golds work
harden quickly and you may have difficulty exceeding 40% reduction
before rolling becomes difficult, whereas platinum, white gold setting
alloys and 18ct YG could roll closer to 70%. I generally work to 50%
between annealings; it only requires a simple mental calculation at
the mills.

  • Exceeding 70% reduction without annealing can lead to distortion
    in the metal i.e., buckling or cracking. * Annealing before 30%
    reduction may cause grain growth and brittleness. The same can occur
    if you work the metal correctly but then over anneal. * Alloys
    anneal (metal grains recrystalise from distorted to undistorted
    grains) at different temperatures: Sterling silver 500 degrees
    Celsius, first red (during heating metal will go frosty white then
    first trace of red appears). Yellow gold alloys 650 degrees Celsius,
    dull red. White gold alloys 750 degrees Celsius, cherry red.

  • Use a reducing or neutral flame and evenly distribute across
    metal. * Allow metal to cool to black heat (just below red heat)
    and quench. The safest practice is to quench in water and then remove
    oxides in the pickle.

Happy rolling,
Graham Farr, from Sydney, Australia.