View Full Version : Punch and die for thin matal (motor laminations)

01-03-2014, 09:16 PM
I'm still thinking about ways to make small experimental motors and it may be best to use thin steel laminations. But I need a way to make a lot of them to stack, and it will probably require a punch and die (unless I would have them laser or water-jet cut). I found some EI transformer laminations that are 0.0135" thick and have a 1.5" x 3" center portion. I found a simple hole punch set for smaller holes, and I thought a similar method should work for what I want:

http://thumbs2.ebaystatic.com/d/l180/pict/120767872729_12.jpg http://www.ebay.com/itm/120767872729

Here is the most recent idea I have for a motor, and in this case it is a Switched Reluctance Motor (SRM):

This actually shows a very thin motor where the stator is made from a USS 1-1/8" steel washer in which I will drill six holes 1/2" diameter and a 1/8" slot for the windings. There will be six windings as shown in blue, and they will be energized sequentially to pull the rotor along and rotate. I show the rotor made from a USS 7/16" washer with four 1/4" holes and then wide slots cut to form the four pole pieces. I also show an initial idea for a punch that could make the rotor pieces from the transformer lamination steel. The first operation would just punch a 1.25" round disc. Then there would be a center 1/4" hole and four more around it. Finally the triangular pieces would punch the slots. But that seems like a lot of work and probably not the best way to do it, especially the last operation. I also thought about making a narrow flat punch that would essentially cut slits, and that might be better.

Actually, for the SRM, the rotor could be machined from a piece of bar stock, while the stator should be made from laminations. I may get some shim stock which comes as thin as 0.001, but I think 0.005 to 0.008 would be good. It would just need to punch 2.75" diameter discs and then a 1.25" hole in the center, and then six 1/2" holes around the periphery and finally the six slots.

I think I will make the first proof of concept prototype from just the washers, and not try to make a hole punch. The punch from eBay will work for 0.5" to 1.0" and probably fine for the thin material I'll be using.

Here is a video of a very crude SRM that I made a while ago:


Here is a good video of the basic principles of the SRM as well as the Reluctance Synchronous Motor which has certain major practical advantages:


01-03-2014, 09:29 PM
Are you aware that most motor laminations are made from glassy steel? It performs much better magnetically. However the motor shops used to burnout the windings in bad motors using an oven. The high temperatures would cause the laminations to revert to a normal solid state and the motor would loose much of its power as a result after it was rewound.

01-03-2014, 09:56 PM
I know that high quality transformer laminations are made from Grain Oriented Silicon Steel (GOSS), but I was unaware that high temperatures destroy the magnetic properties. Perhaps the grain orientation is a result of the rolling process aligning the molecules of iron in the preferred direction of flux, and the heat allows them to reorient more randomly. Interesting.

I added another video that was among those suggested, and it shows another type of motor (RSM) similar to the SRM. With either type, the rotor material does not need to be laminated because there are no eddy currents or losses, but the magnetic properties are still important so as to provide the most efficient magnetic paths.. It looks like the RSM rotor could be made from a tape wound toroid cut into four quarters and turned inside out. I had another idea in which the stator consisted of mating toroid sections, which may provide the best properties. Something like this:


01-04-2014, 06:21 AM
The laminations do not need to be punched , they can be machined electrolytically using similar masking methods to home shop pcb prodution.

Bruce Simpson used electrolytic machining to produce intricate reed valves for one of his pulse jets.



01-04-2014, 12:24 PM
That is good to know. But I think it would be tedious to use that method to make several dozen or upwards of 100 pieces as would be needed for a motor, even a small one, using the thinner material. Even with the 13.5 mil transformer lamination steel it would take 100 pieces for a 1.35" thick rotor or stator.

I found several good videos of the process of making rotors and stators and other motor components, but mostly highly automated industrial processes. This one seems particularly good:


One thing I learned is that perhaps I should make punch and die sets for only a single shape, and then use an indexing mechanism to punch repetitive patterns. Like this "Minster Notcher":


Zahnrad Kopf
01-04-2014, 12:34 PM
I do a bit of this from time to time and more oft than not we simply stack the lams as thick as necessary and WEDM them out. They're then placed on an orienting mandrel, compressed with end caps, and welded along the stack in a few places. Lams have ranged from .004" to .060" in thicknesses, and from 1.000" - 2.500" in height. depending on application. Hope that helps.

01-04-2014, 12:55 PM
I think I need to learn a lot more about tool, punch, and die making before getting too deep into this. Probably best to start with a simple hole punch as Tony has illustrated in his video. I haven't found anything nearly as detailed or useful for the home shop. I did find some videos about the subject of tool and die making, such as this:


I think WEDM is beyond the range of what is available to a hobby level home shop machinist. I have thought about making the lamination blanks and then bolting them together, followed by a manual drilling and milling operation. This might be OK for features that are not too much smaller than the stack depth, and I may do that for a proof of concept prototype motor.

Thanks for the idea.

01-04-2014, 01:20 PM
Cant tell you anymore than whats out there on punches and dies but i do remember the steel we used to make for motor laminations, it was 1.3 % Silicon steel, hardly any carbon, the aim was 0.0035% carbon, bugger all really, not even sure if it qualifies for steel!, theres more sulphur and phos in it so silicon iron was more appropriate, we used to kill it with titanium and a little aluminium (scavange the excess oxygen out of it) it was a bugger to make, 3 ladles of 350 tons was regarded as the maximum we could make as the secondary steelmaking units like degasers were buggered, vesseles all burnt out, steam ejectors to get vacuum burnt.
the steel for transformers was worse as it was 3% and more Si, difficult to get the silicon into the steel, boiling over the top, temperature running away, too hot then too cold.
Memories! Lol
Interesting link
How about using a pantograph?

01-04-2014, 02:03 PM
Are you aware that most motor laminations are made from glassy steel? It performs much better magnetically. However the motor shops used to burnout the windings in bad motors using an oven. The high temperatures would cause the laminations to revert to a normal solid state and the motor would loose much of its power as a result after it was rewound.

I would suspect that the motor problems were due to the inter-lamination insulation being destroyed rather than the steel properties being affected. With the amount of carbon present (or the lack thereof) there is no chance of the steel becoming permanently magnetized.

Since your parts count is relatively small, the manufacture of a dedicated punch would seem to be more work that it would be to cut the parts individually. The smaller the hole or shape the lesser the clearances between the punch and the die need to be.

The punch you have shown, presumably a HF multi sized punch could be used to punch the major holes. A small sharp cold chisel used on a mild steel block would be used to cut out the angular shapes after the major holes are in. I like the method Zahnrad Kopf described even better.

Using the flat washers is a crap shoot. A36 steel might have enough carbon to leave them permanently magnetized, and the thick cross sections would surely be a problem with the high frequencies this motor would necessarily require to operate. 'Prolly get more heating out of the core than rotation using that stuff.

With the sizes you need, a scrap transformer could be disassembled for the metal you need... preferably an old largish audio transformer... they had better transformer steel.


Jaakko Fagerlund
01-04-2014, 02:17 PM
I think WEDM is beyond the range of what is available to a hobby level home shop machinist.
It is ot out of reach as I and many others have done a working WEDM at home, but getting to know enough of all the parts needed and designing a working machine is a little bit big project. I used a better part of one year building one from ground up.

01-04-2014, 02:31 PM
It is ot out of reach as I and many others have done a working WEDM at home, but getting to know enough of all the parts needed and designing a working machine is a little bit big project. I used a better part of one year building one from ground up.

You could use a sinker EDM. It's easier to build, but projects that need an entire different project to be completed make my head hurt.:p

I'm building the pulse EDM sinker unit designed by Ben Flemming... only been at it about 6 weeks, figured I'm at 90% for the power unit and the electronics... I haven't started the mechanical side as yet.


01-04-2014, 02:32 PM
That link on magnetic domain migration in ferromagnetic material was quite interesting. Here is a video that shows the phenomenon more clearly:


I don't know much about silicon steel except that it is used in most transformers and motors, and that it is rather expensive. Magnetic metals also usually contain a fair amount of nickel. The 13.5 mil material I have is fairly easy to cut with ordinary HF scissors, so I think it should be no problem making punches and dies for it. They may also be commercially available. I would need something rather thin for the slots, and otherwise they would be just round. Here are some I found:


And here is a large assortment, but seem to be only the slotted dies:

Here is a punch set for 1.25" to 2" holes. Gives an idea about how to make a punch/die set:

Zahnrad Kopf
01-04-2014, 03:33 PM
I think WEDM is beyond the range of what is available to a hobby level home shop machinist.

I only brought it up as a working alternative to what you had previously posted. RE: -

... But I need a way to make a lot of them to stack, and it will probably require a punch and die (unless I would have them laser or water-jet cut). ...

If you want to do it with punch and die, you certainly can and it's been done simply enough. Just acquire or build a smallish Danly style shoe set. It's not difficult at all, really. Calculating clearances and tonnages is the easy part.

You could easily do it in an arbor or hydraulic or pneumatic press. But for the time and effort, I think you'd serve yourself better by farming that part out simply because sizes and cut outs will change with designs. ( and it will get old VERY quickly, punching so many lams ) If you standardize on a set OD and internal shapes it's not so bad or so much work, but my experience is that these things change very frequently. ( different magnet shapes and/or sizes, cores, and shafts )

Then again, maybe I misunderstand your intent. We only do prototypes, so it's the natural order of things to be always changing. Maybe your experience will be different. Of course, there's always the "... just wanna do it for the sake of doing it..." reason too. Best of luck.

01-04-2014, 03:37 PM
I'm thinking that I could make a tool to do the slitting by using something like a hacksaw blade or other piece of 1/8" thick hard steel as the punch, and have it fit in a slot through a pair of steel plates separated by the thickness of the material to be cut. Here is my idea:

I may need to adjust the dimensions to obtain a deep enough throat, and it might be difficult to mill a narrow slot. But I could also make the top and bottom pieces as two halves, and simply mill a notch the right depth, and clamp the halves together with bolts horizontally. I also show two pins to keep the cutter aligned, but once it has entered the top slot it should not be necessary. Only the bottom pieces need to be precisely aligned.

Actually I might also be able to make a nibbling tool with a narrow cutter with a notch the width I want. In this case it is 1/8" wide.

01-04-2014, 05:16 PM
'Are you aware that most motor laminations are made from glassy steel? It performs much better magnetically. However the motor shops used to burnout the windings in bad motors using an oven. The high temperatures would cause the laminations to revert to a normal solid state and the motor would loose much of its power as a result after it was rewound.'

I can see this being true. Glassy metals are produced by cooling the material from a molten state at a high rate of speed, thousands of degrees per second I believe, to prevent the formation of crystals. Possible that heating it to a high temp, even though not molten, would allow this to change. Just guessing at this- but yes I can also see burning out the insulation between laminations causing problems with higher eddy current losses.

I also wonder about the statement about most motors being produced with glassy steel, and I wonder if this goes for transformers as well- point being, can you identify those as possible donors for smaller projects?

01-04-2014, 05:40 PM
Yes modern transformers are often glassy steel now. I don't how to identify the metal directly, possibly etching it then viewing under a microscope.

01-04-2014, 06:15 PM
Here is information about silicon steel, showing the grades as well as common thicknesses, and types of coating.


There are also other steels that are commonly used:
http://www.protolam.com/page6.html Cold rolled motor lamination steel - cheapest, least abrasive
http://www.protolam.com/page8.html Nickel alloys - higher performance, more expensive, fragile
http://www.protolam.com/page9.html Cobalt alloys - highest performance, most expensive, and magnetic properties are ruined over 1625 F.

That company specializes in prototype and small production stamping of motor laminations. Some time ago I got a quote from a more local company for a set of laminations for something like a 1 HP motor, about 6" dia and 6" long, and it was about $500. At this point I am just looking at this as a learning experience, but perhaps if it proves successful (especially the SRM or RSM design), I might consider having the laminations professionally made. For most practical applications, it's hard to beat a surplus three phase ACIM. And if size and weight are critical, as in electric cars and bikes, it has already been common procedure to rewind motors for lower voltage and then use a VFD to overclock them and get as much as 4x rated power. Adding liquid cooling can remove heat from the stator to allow higher continuous current, for another 2x or so.

01-04-2014, 09:22 PM
Not exactly up on the glassy production as it was the annealing bays bit, however when the steel is cast into slabs it must be kept hot, above 750C, if a slab cools it will burst, i have witnessed this many times, the crystalisation forms giant crystals that burst the slab, dramatically, i have some chunks of slab that did this somewhere, ill try and find and photograph them, they are tennis ball size!, fairly easy to see the crystal structure.
You cannot reheat a cooled silicon slab for rolling, again bursting will occur.
All GO steels are hot connected from castin through reheat furnace to the roughing mill and coil box that takes the 10" slab down to 1" in about 6 or 7 passes, it then gets uncoiled onto a 7 stand hot mill down to about 1.5 mm and coiled up, thats what stops the stuff bursting, the crustal structure is smaller.
Its like putty when its hot and you have to get the temp right or it will sag inbetween the bars of the reheat furnace, these are all walking beam these days.
The lowSils are reasonably tollerant to fatigue and used exclusivly for rotating parts, not as efficient as the higer silicons, but they dont burst!
The higher silicons are used for transformers and stators as creep isnt an issue.
Link to the stuff i used to do

J Tiers
01-04-2014, 09:43 PM
If you have many to do, punching individual features will be worse than a one-step punch, although they will take a lot less pressure to cut.

Punching with cruder dies will throw up burrs that can give you big problems later. Then you have the burrs to clean up. Might make it even more tedious.

I like the stack and WEDM method... should work fine. You can likely de-stack and varnish the pieces, then re-stack in assembly. The WEDM may mess up the edges if they are already varnished for insulation.

No need to use the greatest steels.... although the stator of SRMs will see a higher frequency input from the inverter.

01-04-2014, 10:34 PM
I had no idea of the peculiarities of GOSS. Quite interesting. It seems that such steels are used almost exclusively for large power transformers and perhaps smaller high quality toroidal transformers. Motor laminations generally use non-oriented silicon steel, which is less expensive and easier to use. Because of the geometry of motor laminations, grain-oriented steels would cause variations in magnetic properties and losses. However, the SRM, and RSM designs particularly, may be suited to GOSS because of the way the rotor is constructed. And if the stator were constructed as I proposed, with cut toroid segments, it would also benefit from GOSS. Here is what seems to be a good explanation:

Their website has a wealth of information on various materials - well worth browsing. The Wiki article also seems to be rather comprehensive:

I have also seen transformer cores made from iron wire, which should result in excellent grain orientation and minimal eddy currents if the strands are insulated. It may not achieve the density of flat laminations, unless a hexagonal cross-section could be used and carefully wound with optimal packing. I have searched again more recently for this and have been unable to find anything. This would be ideal for toroids because it would be easier to produce the ideal circular cross-section as opposed to the square that results from tape winding. I have seen tape wound cores that were stepped so as to make a nearly round cross-section, but it is an expensive machining process and probably not worth the extra cost except in high-end military or audio equipment.

It would be interesting to make a motor using powdered iron or ferrite for the magnetic material. It is best suited to higher frequencies, such as 1 to 10 kHz, and would result in a very high speed motor. A two pole motor is 3600 RPM at 60 Hz, so it would spin at 360,000 RPM at 6 kHz. That may be useful for some applications, such as miniature drills. Otherwise, a higher pole count could reduce the speed, so a 40 pole motor would turn at a more reasonable 18,000 RPM at 6 kHz and 3000 RPM at 1 kHz. It might be possible to make a very small, powerful motor at such high frequencies, as is already done for 400 Hz. But ferrite is very fragile and probably difficult to machine.

J Tiers
01-05-2014, 12:07 AM
A client had 200 kW alternators that ran at 60,000 rpm, gas turbine powered.

They used a very special type material from Japan, the name of which I forget. It had very good magnetic properties, and exceedingly low losses, but was a bit harder to fab into stator lams than regular steel. The rotors were PM.

The whole 200kW alternator was about 12" diameter and 18" long. Obviously not directly grid connected, they had a system of rectificatin and boosting and ran 480V grid-tie inverters.

The alternators worked, their gas turbine was not nearly that good.