View Full Version : Forming a bowl

04-11-2013, 03:04 PM
So I've got a problem. I "need" to form a steel bowl the approximate dimensions of a wok, though with a material thickness of somewhere between 1/32" and 1/16 and I just don't have much of an idea how to go about it.

The end result is going to be used in a musical instrument, so I want it to be relatively smooth and consistent (i.e. hammering it in to form is just out.)

I thought about starting with a positive form and using something like a shop press. But I'm just not sure how that would work. I imagine I'd have to over-bend it and I'd have to deal with material crimping.

The other thing that caused a couple grey cells to fire off was a process I've seen a couple videos of where a brass positive is used on a lathe, and a burnishing tool is used to form the material to the positive. Neat stuff. But I don't see it working with something like 4130.

Any ideas? Or should I just get a hammer?

04-11-2013, 03:07 PM
I believe you're talking about metal spinning -- but I don't know enough about it to tell you if 4130 is out
of the question.

I don't think a press would work -- at least not an HSM press -- I think we're talking serious tonnage to get
a finished part out.

I'm afraid hammer and maybe an english wheel might be your best bet but I'm eager to hear what the experts

Can you buy anything and adapt to suit? like you a say, wok, small sat dish, (metal) parabolic mirror, etc?

04-11-2013, 03:12 PM
Sadly, no. this is the initial "R&D" stage. I'm going to want to repeat this process and make metric buttloads of these. Plus, most of the things I've found in the wild are far far too thick to be useful, though perhaps that suggests a solution a

04-11-2013, 03:50 PM
Fwiw, the discs from plows are often used for wok like cooking devices. Apparently they work very well for that, and probably musical instruments.

One Saturday I watched a couple of guys form five silver goblets on a big lathe by spinning sheets of silver, one heating with the torch, the other forming. Each came off flawlessly and quite quickly. Gorgeous. Stems were to be turned and soldered on later.

They were formed by the owner of the shop, whom I'd never actually seen fab anything but who held a reputation as a fab anything. It was well deserved, and I probably should have paid them for the show.

04-11-2013, 03:50 PM
Check the dollar store for stainless steel bowls.

brian Rupnow
04-11-2013, 04:38 PM
Do a google search for metal spinning. I do quite a bit of design work for a local company that makes spinforming lathes, very similar to the old manual spinning lathes, only with the added twist of CNC. I'm not suggesting that you buy one of these lathes, because they go for about a quarter million dollars each. However there are many spinforming companies who will build a "die" to your specifications, and spin the bowls for you. This is a good approach if you are talking large quantities. The largest spinforming lathe I have designed tooling and automation for was spinning 8" diameter 4140 steel pipe into oxygen bottles. The pipe was preheated to a white heat, and had two oxy/propane torches playing on it as it was spun.--Brian

04-11-2013, 05:20 PM
Hydroform it.

Didn't really understand the process until I saw Mythbusters do it to form a torpedo thing they were making.

They basically clamped the workpiece between 2 pieces of 1/2" or so steel plate..one of which had a hole cut in it to the shape of what they wanted. The used a pressure washer (if I recall correctly) to apply pressure to a fitting in the lower plate that deformed the workpiece up through the hole in the top plate.

Make a round hole in the top plate and this should work fairly well.

Here's a thread on another board with a video of the Mythbusters clip.


04-11-2013, 05:54 PM
I've used that last method to form lens shapes from plexiglass and lexan. Clamp the material to a flat bottom with a ring, then heat to forming temperature and pressurize the gap. To do that in steel, you would probably do it cold, as heating the entire jig would require a pretty big oven, and the pressurizing medium would be subject to the heat.

Not a big deal though- a good solid steel plate could be the bottom, and you make up a ring to clamp the disc down all around it. It would take a lot of bolts and drilled and tapped holes, or you could use a couple dozen strong C clamps. You would likely make a groove in the base for a large o-ring to seal it. Pump it with water- I wonder if a pressure washer generates enough pressure to do the job- quite possible. You are looking at a lot of overall pressure when you multiply the 2000 psi or whatever by the sq inches of the disc inside the clamping ring.

04-11-2013, 09:22 PM
Would the end of a 20lb propane bottle fit your need? Maybe not as deep as a wok but it has roughly the right shape and thickness. There is a thread on one of the knife making forums by a guy who makes drums from them.

04-12-2013, 01:38 AM
Another approach would be to form them on a heavy duty planisher. The bowl in the photos is 1/16" galvanized sheetmetal and was formed with a Japanese Vibro Shear in about 15 minutes. My Vibro Shear takes 5/8" square shank tooling and will planish, shear, bead, joggle, form louvers, stretch and shrink and probably much more with the proper tooling.

http://i738.photobucket.com/albums/xx25/roman52/IMG_1514_zps04d0e154.jpg (http://s738.photobucket.com/user/roman52/media/IMG_1514_zps04d0e154.jpg.html)

http://i738.photobucket.com/albums/xx25/roman52/IMG_1515_zps4d53d2cc.jpg (http://s738.photobucket.com/user/roman52/media/IMG_1515_zps4d53d2cc.jpg.html)

http://i738.photobucket.com/albums/xx25/roman52/IMG_1516_zpsaede3e3c.jpg (http://s738.photobucket.com/user/roman52/media/IMG_1516_zpsaede3e3c.jpg.html)

http://i738.photobucket.com/albums/xx25/roman52/IMG_1517_zps9baca48a.jpg (http://s738.photobucket.com/user/roman52/media/IMG_1517_zps9baca48a.jpg.html)

04-12-2013, 01:50 AM
A/C refrigerant tanks might work too. They're a little smaller than a propane tank. The salvage yard down here is full of them. The problem is that most have dimple feet so they stand up.

I press a lot of parts and a shop press should work if you don't have a punch/press. The problem there is that you need to hold the blank on the perimeter while pressing then shear off the excess. I don't use bolts or c-clamps. I use a spring loaded clamp as part of the moving die assembly. Or you could make it separate. And, you'll need male and female dies but these can be made of cast zinc. They wouldn't necessarily have to be machined steel. And 4130 will be hard to form/draw unless it's dead soft. Is 4130 a requirement? Does it have to be heat treated? Something like 1008 CR would form easily.

Spinning may be easier if you don't have a punch press. The 4130 will still be a problem there too.

04-12-2013, 08:20 AM
Hey thanks for the responses guys.

- Most of the tanks mentioned are just too thick walled.
- No, 4130 isn't a requirement. It's just what I have a bunch of test sheets of, in different thicknesses. My metallurgy is for shyte. I don't have any real knowledge about tonal qualitites of different grades of steel at a particular thickness, so whatever I can work the easiest I'll go with I suppose.
- I'm really encouraged by a couple of the ideas, an English Wheel seems like the right idea (read: the easiest for me to wrap my head around without pictures) if I could rig up something that would hold the center and eek it in a bit as I cranked (with a die/anvil wheel set I'm sure "even I" could turn out on the lathe.)

Spinning was the first thing I thought of, but the size is just too big for a normal lathe (the target diameter at the widest point is pushing 22".)

04-12-2013, 08:25 AM
I haven't heard anything "home shop" to make "buttloads" of units. After you make your prototype, however you do it, You might look for a local outfit with a punch press to form these things. After initial die cost, it isn't very expensive to hammer out a hundred or so at a time (if that is what buttload means).

04-12-2013, 10:13 AM
I did a search for "tank ends" and found this;


There are a bunch of shops who make nothing but what you are looking for.

Also, 20lb propane tanks have a wall thickness less than .090". And then there is the question, if you need a wok shape, why not just use a wok?

04-12-2013, 01:11 PM
Maybe you can adapt a tank head. Here's one supplier: http://www.tankheadexpress.com/ There are many more.


04-12-2013, 09:19 PM
Perhaps do a search on gazing balls ... cut them in half?

04-12-2013, 10:20 PM
Two companies-


04-14-2013, 12:02 AM
You could also use an English wheel.

04-14-2013, 11:38 AM
yeah, someone mentioned an English wheel up top. I'd never heard of one, much less seen it. Me and a buddy of mine (a tool & die guy) went in to Horror yesterday and checked out theirs and spent a lot of time nodding at each other. Sure looks like the right tool for the job, though I'm going to burn a LOT of material getting the English right (to abuse a term.) Turning a curve like that might be a 3 hand job.

Lots of neat videos on youtube on using an English wheel.

04-14-2013, 01:12 PM
You might be surprised at how quick the chape comes in. They used to make aircraft cowlings that way. The most difficult thing you will need to do is work out the method to periodically normalize the disk as it is cold worked. I ran some quick numbers and to make a bowl 10" diameter and 2-1/8" deep, you would need to stretch the center of the bowl about 13/16" across the diameter.

04-14-2013, 02:31 PM
Not sure I understand what you mean by normalizing the disk.

04-14-2013, 08:45 PM
If you've got a lot of these to do, you don't want an English wheel. You want a wide and deep throat bead roller. A bead roller is sort of an English wheel, but it's motorised. With an English wheel, you run the thing round by hand, so it's better for mudguard shaped stuff, where you run it forward and backwards. Trying to get your piece true and circular would be hard.

But the solution all depends on the quantity. The more you make, the cheaper the tooling is per item. To press this kind of shape is quite straightforward for those in the know (not me), but if you have a lot of shrinkage to handle, you would have to use several press passes, with the multiple formers. Like I say, the more your buttload means, the more you can spend up front on tooling.

If I recall, back in the eighties, the tooling to press a new vehicle design was about 10 million. Assuming 100 parts, that's an average of 10,000 tooling per part. For a simple spherical form it would come down by a factor of five say, but costs are up by that since then. So I'd extimate up to $5000 for tooling for your 'woks'.

On the other hand, a trip to China to a wok factory and back might be less than that !

You could even set up a decent home 20 ton press with multiple tooling. Shape up two prototypes with a shrinker and a hammer and dollies, out of material only a little thicker than your target material. Then stiffen one by welding ribs inside, and the other with outside ribs. A couple of frames to hold then in the press, and off you go. Again, depending on how much shrinkage you plan - i.e. how deep the bowl is - you might need two or three stages of pressing.

The home route is a lot of work, but the pro route is expensive. The wheeling, rolling or spinning routes are labour intensive in the long run. You takes your choice.

04-14-2013, 09:28 PM
Not sure I understand what you mean by normalizing the disk.

When you roll the disk, you cold work the steel and cause it to harden. Depending on the material you use, you heat it up to dull orange and slow cool it. That relieves the residual stress and softens the steel so that it can be cold worked again.

04-15-2013, 08:49 AM

04-15-2013, 09:36 AM
Sounds like the equation I have to balance is nicely summed up here
The home route is a lot of work, but the pro route is expensive. The wheeling, rolling or spinning routes are labour intensive in the long run. You takes your choice.

The first few iterations of this are design/prototyping. I need to make sure the material, thickness and final shape are going to work for the application. Once I have a successful prototype then I can graduate to thinking more seriously about production runs. I just didn't want to shoot myself in the foot on the front end.

Before I'd heard about an English wheel, a 20 ton press with a series of progressive dies was my intended approach. Hell, it may be that I end up creating the forms on a wheel, then doing production on a press. I'll probably need the press for the next stage of processing on these monsters anyway.

As far as the definition of buttload, my plan is batches of 25 for as long as demand holds out. We'll see how things go. I've got a LOT of work to do and a lot of steel to ruin before I get to that point.

*sigh* Sometimes I think dating really would be the cheaper hobby.

04-15-2013, 11:02 PM
One possible advantage of the English Wheel or the planishing machine is that you could do non-circular bowls.

04-16-2013, 11:41 AM
I still think that an English wheel isn't the answer.

On either of the two main metal fabrication sites there are projects where people have made their own deep throat and deep clearance bead rollers. I'm using the word throat to mean a roller that can deal with flat material, but can roll a long way from the edge. I'm using the word clearance to mean the wide material can be way off flat in one direction (downwards, usually).

Given that the bead roller powers both wheels, this means that while the upper wheel is powered simply on a long axle from the rear of the machine, the lower wheel can't have that axle, because of the clearance needed for shapes like your bowl. So the power comes vertically up inside the structure that supports the wheel, the power being transmitted by bicycle chains.

This powered bead roller leaves you standing supporting your bowl as it turns, governing the speed with a pedal, whereas a genuine English wheel means you have to provide the force to move the bowl throught the rollers.

I can't see you being able to control the bowl through an English wheel well enough to get good results.

With rollers made to the right curvature, I could see a bead roller getting the shape you want in a very few minutes. Start at the edge, to get the shrinkage, and work in.

Thre's a fellow who's made some machines that would do the trick. You could start with his 'generation 5' bead rollers. He puts his videos on youtube under the name of courtjester90.

04-16-2013, 12:36 PM
I must not be seeing the right kind of bead rollers in my searches. It looks like the throat is too... shallow, certainly to go from the outside in. There'd be no place for the curvature to go.

From the inside out you'd need to be spinning the stock manually pretty quickly while it cranked away under power.

I'll definitely check the vids when I get home.

Thanks. o/

04-16-2013, 05:44 PM
With the - let's call them 'tall throat' - rollers, the trick is to get the power to the lower wheel. courtjester90 - Tom Verity - and I've seen a video of an earlier model of his that was clearer - he runs bicycle chain to a horizontal axle at ground level, then a bicycle chain up to the normal level again, all enclosed in fabricated box sections.

Another roller I've seen, with less of a tall throat, uses a chain to get the drive down to about a foot below the top axle, and then uses UJs to get the drive to the front. This looked a bit flimsy, but I think the principle's OK. It just needed beefier UJs.

I agree with your thinking, that you need to shrink the OD first, and progress inwards, and that's why you need the tall throat. In fact I suspect you'll need to shrink the OD several times. You might need a sequence like outer, in a bit, outer again, in a bit, in a bit more, back to outer, and so on.

Mittler Bros does one with what looks like a 4" tall throat, but you either use one large and one small wheel, or ywomedium size wheels, both much larger than the usual 2" diameter rollers. With the large and small wheels, you ought to change the gearing to avoid slip, and I don't know if they cater for that. The two medium wheels look thin and flimsy. Either way, I don't think it quite has the throat height you'd need, but I may be wrong.

04-16-2013, 06:16 PM
I'd still check out hydroforming.

04-16-2013, 08:12 PM
The difficulty with a bead roller is that the "belly" of the bowl needs to be stretched. Putting a bend in the metal won't make a stable bowl. I'm thinking that perhaps what you might mean is something called a stretcher roll. It is used in saw filing to roll the middle of a wide bandsaw so that it is longer than the front and back edges. There were some saws that were used on harvesters that were a bowl shape and the curve of the bowl was maintained with the stretcher roll. Otherwise the edge of the saw would stretch and the bowl would flatten out.

Hydroforming should work, but the tooling needs to be able to hold the hydraulic pressure, unless you use an impulse former. The basic machine uses a housing that will hold a fair amount of pressure, but not as much as a regular hydroformer. There is a block with a hollow that is the shape of the bowl. The housing is gasketed to the sheet metal. The housing is then filled with water. The housing has a chamber for a blank shotgun shell and when the shell is fired, the shock wave going through the water hits the sheet metal and blows it into the block mold behind the sheet.

I suppose another approach would be to weld the sheet into the end of a piece of round pipe and pay an 8 year old to pound on it with a small ball pein hammer, or you could go after it with a chipping gun with a round nosed punch against a hollow wood block.

04-16-2013, 08:41 PM
If I'm not mistaken, the material in a propane tank is .075 thick. The tanks are designed to work at 120 psi, are safe at 375 (where the pressure release valve is supposed to open, and tested to 900 lbs or so. I tested one myself at that pressure, but that's another story. (nothing bad happened) That suggests to me that 1500 lbs might be able to stretch the metal, and 2000 psi might be able to form it easily. Since that pressure is easily available with water, it seems a pretty elegant way to 'dome' a piece of sheet metal. In fact you could do 2 at one time if you dispensed with the need for a base, and just used two heavy-duty rings to clamp the pieces together with. I suppose you'd need a third ring to put between the two pieces of sheet in order to have a point where the water could be injected and another point where the trapped air could be let out.

I'm trying to think of the drawbacks to this method- besides the cost of the rings (any other way would cost you initially as well) one would be the time needed to clamp and unclamp the rings. If you use the three ring method, you get two bowl shapes for one round of clamping and unclamping- bit of time savings there-

04-17-2013, 08:13 AM
.....That suggests to me that 1500 lbs might be able to stretch the metal, and 2000 psi might be able to form it easily. Since that pressure is easily available with water, it seems a pretty elegant way to 'dome' a piece of sheet metal. In fact you could do 2 at one time if you dispensed with the need for a base, and just used two heavy-duty rings to clamp the pieces together with....

Given the OP's blank thickness, the pressure required would be well under your hypothetical. 1/4" A36 will form a "doughboy" (WW1 helmet) shape with a 1" depth @ 6"dia. around 400psi.

An unwanted discovery after someone had allowed overpressure during hydro testing a line, It had to be cut apart when the banjo (isolation blank) wouldn't come out of the flanges.

04-17-2013, 10:11 AM
It's starting to sound like hydroforming would be a really great way to go for production level work. But isn't going to get me very far in the home shop for initial prototyping. Unless of course there's some really clever home shop hydroforming rig that I can't quite find online (which I seem to doubt. It looks like a high power operation in relative terms.)

04-17-2013, 11:18 AM
A set of 150# slip on flanges, some bolts, nuts and a gasket.
Pressure washer and some valving.
Flanges can be had with small threaded side port, or make your own flanges.
Yer in business.

04-17-2013, 12:55 PM
If hydroforming is viable, try to get hold of a wellhead company. At 22", you're at the higfh end of size., but you can find flanges up to 15,000# rating at 20". Not free, for sure, but could be modded to work.

04-17-2013, 03:09 PM
Jpfalt: You may be quite right. I'm not very experience - just motorised my first bead roller.

I was thinking along the lines of some of these machines for madw:


I was guessing, maybe wrongly, that with multiple passes, shrinkage would get introduced round the edge.

My main concern was that, with hand feeding through an English wheel, a circular shape would be very hard to control. Back and forth with a fender, yes, but feeding a circle round and round ?

But a small but tall throat bead roller should be able to do some stretching in the centre of the bowl, no ?

04-17-2013, 04:18 PM
A set of 150# slip on flanges, some bolts, nuts and a gasket.
Pressure washer and some valving.
Flanges can be had with small threaded side port, or make your own flanges.
Yer in business.

That's just beyond my ability to visualize.

04-17-2013, 05:04 PM
"That's just beyond my ability to visualize. "

Take a look at piping flanges. They come in a variety of pressure ratings and attachments; threaded, weld neck, slip on, blind and more.
I suggested slip on as I think they're the cheapest.
A rule of thumb, ANSI flanges are good for about 3x working pressure, so a 150# gets you in the middle of 4-500 PSI. Use all the bolts in this situation. "Four bolting" will leak in the larger diameters.
Your blank would need be the flange gasket face OD, not the pipe size. That portion won't deform and would have to be cut off if you want wok shape.
So you have a flange, a gasket, the blank, another gasket and a mating flange. Somewhere you have to introduce your water. If you had a spacer with an entry port in between the two flanges you could form two pieces per cycle. Otherwise, one flange has your threaded, or welded, connection.
Pipe the pressure washer so that a valve can bleed off an otherwise instantaneous increase to its working PSI.(if using a pressure washer)
Your valve tree should have an air bleed off air when "packing", which depending on how you configure the rig might amount to about a quart (less) as a SWAG; and a gauge or DI for replication.

If std. flanges aren't the right size make your own by extrapolation. Mild steel is OK. DO NOT use cast iron.
BTW, if you're into production, an hydraulic clamp rig can replace the bolts.

04-17-2013, 07:33 PM
one other option...20T+press, hard maple "form" of your shape, and a rubber pad/pads, to push it into....many "one off" parts are made this way...you may even get several possibly more...http://stampingsimulation.com/engineering-definitions-details/rubber-pad-forming/