Twisting triangle wire

Jewelry Discussion
1

I need a better idea for twisting triangle wire that has a tapering
end.

I am using 6ga wire (14k in this case) and tapering one end about
1.5 inches to almost a point Picture: http://tinyurl.com/2ebp9nh

The idea is to twist the wire only on the tapered portion and get it
even Picture: http://tinyurl.com/28fboho

You can see, not quite even. The problem is that you can’t just grab
the end and start twisting. Because of the taper, it won’t work. I
made this tool… Picture: http://tinyurl.com/25w9678

With it I can grab the wire, twist a turn, move it down, twist more
and so on to the end. It works but is inconsistent and some rework is
required to remove dents.

I also tried using a draw plate as a “wrench”. Big mistake.

I have to make quite a number of these so it’s important that the
twists be even and the method something which is sure and does not
mar the wire requiring rework.

Any ideas are greatly appreciated.
RC2

2 
With it I can grab the wire, twist a turn, move it down, twist
more and so on to the end. It works but is inconsistent and some
rework is required to remove dents.

To achieve consistent twist on tapered wire, one must insure that
cross-section though tapered area is constant.

Let’s say we have 5mm wire tapering to 1mm point. Cross-section of
1mm point is 3.14mm^2, which means that 5mm section is not round but
rectangle 5mm X 0.63mm; 4mm section would be 4mm X 0.78mm, and so on.
Some adjustments are needed to compensate for geometry, since it is
easier to twist rectangular profiles that round ones, so instead of
0.63mm, 0.7mm would be better. Some experimentation requires.
Annealing becomes a factor as well.

To generalize - while wire is tapering from 5mm to 1mm in X-plane,
the wire must taper from 1mm to 0.63mm in Y-plane and direction of
tapering must be reversed by 180 degrees.

If you need more than one, make a model and cast it.

Leonid Surpin

1 Like
3

Hi Bob,

I’ve done similar things in steel, using differential heating, but
that’s not going to work in gold. (Gold’s hot-short, meaning that
it’ll break if you try to move it much while it’s hot.)

The problem you’re having is that you’re trying to get an even twist
on a structure of differing strength. It’ll twist most quickly where
it’s thinnest. About the only thing I can think of is that you forge
it down to a taper, and hope that the extra hardening from the taper
forging is enough to exactly mirror the reduction in strength due to
reduced cross section. If you manage to balance them out exactly,
you might just get an even twist.

Oh, wait, it’s a triangle. Forging a triangular taper’s a right pain.
Probably best to use a tapered 2 sided die to forge that. Much fun. I
might try it with a 4 sided tapered form in silver, just to see how
close you can get it. It may turn out that the answer is to taper it
to a certain point, anneal, and then forge more from there, in order
to get the stresses set up to allow you to get an even twist. It may
also be that you have to forge the whole thing down from a larger
stock size, to get the main body half hard to begin with, and then
forge the point from there without annealing at all. There will be a
lot of experimenting to try to figure out what to do to generate the
right degree of work hardening in the right places to pull this off.

This may just be possible, but it certainly won’t be easy, or
anything I’d rely on for a production piece at $1300/ozt.

FWIW,
Brian.

4 
The problem is that you can't just grab the end and start twisting.

Well, Bob, that’s a tough one… As you say above, and I’m sure you
understand, you can’t just twist it because the fine point will
over-twist befor the heavier wire even starts. I would suggest
something (I just thought of…) which may or may not work. Make a
mounting for the wire in a vise, if you haven’t already - that would
be a flat side with a notch on the other side, to hold a 3-cornered
piece. Get a piece of metal - brass might work but steel would be a
lot better - and cut a series of triangular holes in it that are
small, medium large, etc. However many you feel you need. I would
mark the wire in 2-3mm increments and cut the holes till they hit the
marks. Then put the large hole on the wire, twist, then the next
smaller and twist and etc. Seems like it would work fairly well with
some tinkering and a bit of learning curve. How much to turn on each
increment, etc. It could be that you might have to sacrifice some
perfection in the name of expediency. “It’s not a perfect twist but
it IS a pleasing twist”. Otherwise known as making lemonade… Tough
problem though…

5

Bob,

I need a better idea for twisting triangle wire that has a
tapering end.

You’ve set yourself a difficult puzzle.

No doubt you can make one or two, or many, so long as you don’t mind
taking the time and effort to tweak and adjust, and to scrap some
that don’t work.

But when you say you need many,

There’s only one really easy answer to this.

Industrially, someone would make a proper die, a die striking or
perhaps a swaging die, that could product this form. Lots of money to
do that, but once you have the die, and the press or swaging machine
to use it in, you could bang em out.

However, that’s unlikely to be reasonable for you.

So.

Make one perfect sample. Take however long it takes to get it there

Note that a twisted triangle wire, once twisted, is round. The
triangle wire means the grooves in the twist are deeper than a square
wire would give, and the cross section tells you how many ridges or
threads are going around the wire.

But you could, if you needed, produce the form literally by cutting
it into a round taper with a file. Slow, but possible. Using the
threading fixtures on appropriate machine lathes could also do this
in the unlikely event you could figure out how to hold the gold twist
wire centered and rigidly enough in a lathe. If it was larger, you
could do both the twist and the taper with the same cut. But again,
in a smaller gold wire, this might be a very difficult cut with even
the best machine lathes. A grinder, rather than a lathe tool, such
as might be used industrially to make small milling cutters or drills
might also be an inspiration as to how to cut these.

And on that thought, As a model, for some “speeds” of twist, you
might even be able to find an existing cutting tool to use as the
sample. I’m thinking of tapered spiral milling cutters like those
used in CAD/CAM milling machines to cut wax models. These are also
sometimes a spiral triangle form, available in different degrees of
taper. Most that I’ve seen though, don’t have as fast a spiral twist
as your photo.

Anyway, however you make the single perfect sample, make it perhaps
five or ten percent oversize. (whatever is needed to adjust for
casting shrinkage)

Then make a good rubber mold, or better, have a metal mold made. The
idea is to be able to get really high quality wax models with a
minimum mold mark.

Then cast them. This will be more efficient in terms of gold use,
much more efficient in terms of time, and will solve the whole puzzle
of how to actually fabricate these repeatably and efficiently. With
14K gold, choose the right alloy, and with centrifugal casting
(allows lower flask temps, which gives a finer grain structure), you
then can heat treat to get a degree of hardness/springyness if you
need it, and you should be able to get pretty close to the
metalurgical properties of what you’d get if you fabricate these.
Not exact, but perhaps close enough…

Now that I’ve written all that, (which I won’t bother to delete
because it may still be the best answer if you need many) I’ve got
another idea that just occured to me as I type…

Take a pair of decent round nose pliers (yes, round nose. That
allows the spiral to be gripped without creating a flat in the long
dimension. On one jaw, file one or more triangle grooves, following
the round jaw profile. The triangle grooves should be not only
following around the jaw of the plier, but also angled to roughly the
desired speed of the spiral. If you get this right, you’ve got a
plier that can grip the spiral at any point along it’s length before
or after twisting, without leaving an ugly scar or plier mark. You’ll
want to polish those holding portions of the plier jaws for the same
reason. You might have to play with that angle a bit to get it to
work, since it also has to grip the wire before it’s twisted. And you
might need a couple grooves of varying depth to grip the taper well
as it gets smaller but also at it’s larger dimentions.

I’m kind of imagining that the groove you first use to grip the wire
could be straight. But you’d need the angled grooves to then go back,
if needed, and adjust and tweak the twist after the initial twisting.
Perhaps a “hybrid” groove in the plier jaw could do both… (forgive
me if this seems muddled. I’m kinda trying to think it through as I
type. Doesn’t make for a great description)

You also need a second such fitted plier, but this one cut in
flat/needle nose pliers with the groove straight and flat and not
angled, to hold the untwisted part of the wire, while the other plier
is doing the twisting. Also polished…

To use this, you’d firmly grip with the modified flat/needle nose
plier, while the other started at the large end and gradually slid
down the taper while twisting. This would take a bit of practice, but
should work. The twisting would be taking place mostly right next to
the round nose plier, since the metal there is thinner, while as you
slide that plier towards the end (while twisting), the thicker
tapered portion, being both thicker and now already a bit work
hardened, shouldn’t twist too much more. So how much you twist is a
balance between how fast you twist and how fast you slide the round
nose plier down the wire…

Hope that gives you some ideas…

Peter Rowe

6

Hello Bob, How long is the finished piece to be? I’m sorry if I
missed it in a previous post. If it is not too long and you don’t
need many pieces, Taper it after it is twisted. Twist a non-
tapering piece of triangular wire and then file one of the flat
sides On a three sided piece, you only need to file one of them to
make the whole thing taper.If it needs to be 3 incheslong, I’ll go
back to the drawing board.

Have fun. Tom Arnold

7

Using the threading fixtures on appropriate machine lathes could
also do this in the unlikely event you could figure out how to hold
the gold twist wire centered and rigidly enough in a lathe.

Just fantabulating at this point… It is a deceptively difficult
problem posed here. Peter is correct (of course) about using a lathe,
theoretically. How to hold it is the only real problem, and that’s
not insurmountable. Set your tailstock to the taper, cut that taper
and then use some threading speed to cut the spiral. The real
problem with it is that your taper is going to be cut round, not
traingular, and your threads will be threads cut into that - they
won’t come off the faces and edges of the original triangular wire,
IOW. You’ll have a triangular wire that stops, and then a round
taper with graduated threads cut into it - threads being a generic
term here. The shape of the cut is determined by the tool used.

Molding is the real solution, as always, but maybe that’s not an
option to the maker, I dunno. I wouldn’t twist it at all, in that
case. I’d just do the whole piece with files and filing. It’s much
easier and greatly more precise… I made a golf tee once, and it
being essentially cylindrical I didn’t want distortion or parting
lines. Using “Plast-o-wax” I could cut the mold above and below the
model, but I didn’t actually cut out the model. I could just pull
the wax out directly, and got a perfect wax every time. It’s tough
stuff… Cutting above and below, which makes a sort of "finger"
gives much more flexibility than leaving it imbedded in the bulk of
the mold…

8 
hold a 3-cornered piece. Get a piece of metal - brass might work
but steel would be a lot better - and cut a series of triangular
holes in it that are small, medium large

Having done something remotely similar(you know, the same thing but
different?) I’ll say that the thickness and hardness of your brass
or steel die will be critical. A thicker gauge die will tend to leave
‘flats’ on the twist the same dimension as the thickness of the die.
That is you’ll wind up with say a 4mm long twist section and maybe a
1mm (assuming the die is 1mm thick) straight section. A harder die
will tend to dig in at the edges. So while steel may sound like the
answer because you can make it thinner, it will leave gouges. While
brass may sound like the answer because it digs in less, you might
need greater thickness and have a drunk looking twist. You could
massage the thing to get rid of the flats, might entail a lot of
grinding and negate any advantage to using the dies.

It might be possible to make a series of twisted tapered hole dies
but Man, you’ll earn your martini at five o’clock. You think good
azuring is tedious? Without actually trying it I couldn’t say there
wouldn’t be some unforeseen mechanical difficulty.

I’d agree with Peter, make a great master however you can and cast
from that. Drive yourself nuts just once. 'Course that means only
one martini. But you can have two olives!

9

John,

The real problem with it is that your taper is going to be cut
round, not traingular, and your threads will be threads cut into
that - they won't come off the faces and edges of the original
triangular wire, IOW. You'll have a triangular wire that stops, and
then a round taper with graduated threads cut into it - threads
being a generic term here. The shape of the cut is determined by
the tool used.

Once you twist the triangle wire, you also have a round shape,
defined by the outer edges of the triangle. So if you cut a triple
thread into the wire, even if the wire starts out round, then if the
shape of the cutter is right, the cross section of that threaded
portion can be triangular. One problem with duplicating the original
photo, though, is that the twist doesn’t start absolutely abruptly,
but blends from straight triangle to twist. A cut thread wouldn’t do
that. So one might solve this by starting with slightly larger round
stock, allowing one to then file the unthreaded/untwisted portion to
triangle section, blending into the thread. Kind of tedius to do
though, if you need many.

But for a visual sample, take a look at the spiral tapered three
sided milling cutters often used on the CAD/CAM wax cutting mills.
these are ground into the end of round steel (or carbide). the shank
is round, but the ground portion is clearly triangle cross section.
If the tool grinder had started with a straight flute, not twisted,
on the round portion of the wire, transitioning to the twist when the
taper started, it would produce exactly the type of shape desired
here.

Here’s one of those cutters.

Note that here, the tool grinder cuts right into the round shank,
starting the tool with a smaller size than the round shank. But the
principal could apply even if the grinder starts by cutting flats
that just meet up at three corners.

And of course, with this cutter, the spiral is a slow one, meeting
the needs of the tool, rather than some visual idea. But the idea
still holds with a faster twist.

http://www.ganoksin.com/gnkurl/z

Peter

10 
How long is the finished piece to be? I'm sorry if I missed it in a
previous post. If it is not too long and you don't need many
pieces, Taper it after it is twisted. Twist a non- tapering piece
of triangular wire and then file one of the flat sides On a three
sided piece, you only need to file one of them to make the whole
thing taper.If it needs to be 3 incheslong, I'll go back to the
drawing board.

Thank you Tom (and others),

The basic blank is 10 inches long; that pretty much leaves out
casting. Two other amplifications might also be helpful…

  1. By “I need to do many” I mean less than 50; but that’s a lot
    given the effort.

  2. The pictured example is 14K but the majority will be 925

Tom, I am visualizing the idea of tapering the piece after twisting
and it just seems as though there would be twist marks at the big end
but only vestigial, if any twists at the other.

Bob

11

I can think of three possible solutions, each having their own
drawbacks, each require you to taper down the triangle first:

  1. Holding the piece in a vise and use a three jaw chuck to twist
    the triangle wire at the widest section of the tail, moving down as
    you go adjusting the chuck to hold a thinner section. This can either
    be a three jaw chuck from a lathe or mill held in your hand or the
    three jaw chuck from a electric drill (in which case you would want
    to set the clutch on the drill to it’s lowest setting and work up).
    You’ll need to file away the marks left from the chuck.

  2. Use a triangle wire draw plate as a “wrench”, stepping from the
    largest size down. As the draw plate is significantly harder than
    the triangle you’ll need to take care to only mar the surface in as
    much as you are willing to file down.

  3. Take the tapered end of the triangle wire and push it into
    thermoplastic or epoxy at varying depths. Use this as a wrench to
    twist the end, starting at the thick end and working your way down.

Another alternative would be to do this on a 4 axis cnc mill (easy
and reproducible, although with lots of wasted material). A Sherline
should be able to handle this easily.

Scott Garrison

12 
The basic blank is 10 inches long; that pretty much leaves out
casting. Two other amplifications might also be helpful...

No, it doesn’t. Coil up the long length into a spiral shape. With
proper spruing, this brecomes quite castable, and after cutting off
the sprue, you just straighten it out. That’s easy. The flat spiral
is also easy enough to mold.

Peter

13

Here are some ideas that border on the impossible but may lead to
other ideas…

Encase the tapered triangular wire inside a uniform piece of metal,
twist the combination, and then romove the casing.

Since lost wax casting is stated as not possible, perhaps the
tapered wire could be cast inside pewter or pure tin actually bonded
to the gold or silver core. After twisting remove the casing by
dissolving in acid.

Sandwich the tapered wire between two strips of copper sheet. Rivet,
solder, weld, or fold the two strips together ensuring the wire is
central and tightly clamped inside. Twist the ribbon while under
tension to keep it straight, then remove the casing…somehow.
Separating a twisted ribbon may be a big problem.

Sandwich the untwisted tapered wire beween two half round or
rectangular lengths of copper; each half has a tapered groove cut or
impressed to receive the tapered wire. Bind them together, twist,
then and hopefully the two halves will unscrew from each other and
separate, or they have to be dissolved enough to loosen them.
Binding could be a winding of steel wire along the whole length, and
the twisting could be in the direction that tightens the binding as
the sandwich is twisted…mmm, that could introduce secondary
spirals if the wire is strong enough to bite deep into the sandwich,
and what if the binding wire is wound at the same pitch as the
desired spiral?

Screw threads are roll formed by offset opposing rollers. Three flat
rollers facing each other leaving a triangular gap, each roller
offset to the pitch of the spiral. Now draw a rod of round wire
through the rollers making sure the rollers move together and get
narrower in width with each pass of the wire. This will form the
taper and the twist simultaneously…easy, just like a perpetual
motion machine!

Alastair

14 
This will form the taper and the twist simultaneously...easy, just
like a perpetual motion machine!

Ala’s post is the first I’ve read this morning - I’m sure there’s
more, too.

I’ll suggest an even better way: Don’t pound your head against a
brick wall. If it’s too difficult with the tooling at hand, which
translates in to too expensive/not cost effective/losing your shirt
also, then alter the design to something that IS doable. Cast the
points with a ball on the end where you can solder, thus hiding the
seam, don’t twist them at all, use fluting instead of twisting - etc.
and etc. Just a thought…