For Al Belli - Questions Re Stamping Dies & Production

[larry cottrill]

Al -

I saw in your latest response on the Valveless Forum that you have done stamping dies. This is something I'm interested in, and there is no one I have asked locally that knows of anyone around here who does stamping. Also, there is of course the problem of contracting for small production runs.

What would it cost to make a set of dies for a Dynajet size [or slightly larger] engine shell? Are the dies produced by casting and case hardening, or what? If so, I could make sugar pine patterns myself, with a little guidance from someone who knows what he's doing. [OK, I don't have shrink rules, but I can easily figure out how to oversize the patterns properly, since the shape is simple.] What would I need to do to get this done, and what kind of cost would be involved in producing the dies?

Also, do you know anybody that would be willing to do small runs [say 100 pieces at a time] in, for example 316 stainless sheet?

What is the useful life of a set of dies for this kind of production? What pitfalls are there to stamping [for example, do you KNOW you're going to get good stampings when you design a set of dies, or do you expect to do some tweaking after you see the first units that come off the press]?

L Cottrill

[Al Belli]

Hi Larry,
You are describing the dies that are called drawing dies. The metal blank is held very tightly by a heavy spring loaded platen, in contact with the hollow ( die ) face. The forming punch pushes the metal into the die, and is then retracted. The radius at the die face, where the cavity is, is very important in producing a wrinkle free drawn part. Larger radii usually give better results, but most diemakers start with a smaller radius to give a sufficiently high draw tension to the blank. Most draw dies are produced from lower grade tool steel that is heat treated for strength and wear. The radius area of the die is the most highly stressed area, and is subject to the most wear. Most dies require some minor modifications during proofing.
The cost of a Dynajet sized die to produce half shells would probably be in the $9,000 to $10,000 range.
Set-up cost to get the die into the machine, and adjusted properly would probably be around $400. Running labor cost would probably be $2 per part (including trimming ), plus material.

A stamping house will make any quantity that You want, since You are paying for the set-up.

Labor breakdown:
2 parts ( for one tailpipe ) = $404

10 parts =$420 or $42/part
100 parts = $600 or $6/part
1000 Parts = $2400 or $2.40/part
10000 parts =$20400 0r $ 2.04/part

Don't forget to add the die amortization cost to the above figures, as well as some die maintainence cost.
With all costs including die amort. and material for 1000 parts ( 500 tailpipe assemblies ) figure about $18,000 Total; or $36 for each assembly. Your next run of 1000 parts will be about $9,000 total; or about $18 for each assembly.( and all subsequent runs )

Al Belli

[larry cottrill]

Al -

Thanks very much. Gee, it's no wonder everybody isn't doing it this way -- you have to really know you've got something good before you'd be willing to lay out that kind of front-end cost. On the other hand, if you do know you have a winner and can sell thousands, the overall production cost looks really good. This is exactly the answer I expected, except that I had no idea the creation of small dies would be that expensive.

There is just a bit of that radius observable on a finished Dynajet shell at the welded seam. It looks to me like somewhere around .25 to .50 inch radius -- a little hard to estimate because so little of it is left after the pieces are trimmed and welded up.

Thanks again! Very interesting to finally learn a little bit as to how this is actually done.

L Cottrill

[Pieter van Boven]

http://www.geertsmetaalwaren.nl/index2_eng.htm

Always nice to have some pictures!

Pieter.

[Tom]

Right up Marks street I think...



From that site Pieter posted.

Tom

[Al Belli]

Hi Pieter ,

Nice information! The dies used in the type of deep drawing shown are substantially more expensive than those that I described in My last post.
The nice parts in the product pictures that almost look like pulsejet tailpipes will require a series of draw dies to get to the final shapes shown. This is what drives the cost up substantially.
To draw a Dynajet tailpipe using the conventional drawing process, as a seamless part like the ones shown on the referenced website, You will require about 6 draws; meaning about 6 die assemblies. The cost of the dies would be very high. That is why the two piece welded construction is used.

Thanks,

Al Belli

[Mike Everman]

Nice little videos of the processes, too.

[Pieter van Boven]

Al, Larry and others,

shouldn't it be possible to use an old lath and turn it into a deepdrawing machine? ...no time to discuss this right now, I wiil be back later!
Take a look at the picture to get an idea what I mean.

Pieter.

[Mike Everman]

That looks like automated spinning.

[Al Belli]

Yes, that is automated spinning. Note the auxilary vertical slide for facing the finished part to length. This machine looks like it was built, and tooled to make this part only. I might be wrong but that is what the machine looks like.
The length to diameter ( L/D ) for a complete tailpipe assembly is not compatible with spinning. I would consider spinning the combustion chamber and transition as one piece, and then adding a tailpipe.

Al Belli

[larry cottrill]

Yes, that must certainly be a spinning lathe from a few years back. I especially like the privacy curtains carefully placed behind the machine in this view.

Spinning would often be nice for a lot of the things we do. I know Bruce was experimenting with it for a while and had some pictures at his site. I'm not exactly sure whether it works well for long, narrow work, though. Also, it is done over a lathe-turned mandrel that has to be carefully prepared. If the shape is narrow in the middle and more open at the ends [say, a combustion chamber and a flare], then what do you do? I suppose you could have a 'split mandrel' that breaks in the middle so both pieces can be pulled out the ends. It seems to me that you also have to have some taper for 'relief' to be able to get it off the mandrel; i.e. no part of your spun shape could be truly cylindrical. In many cases [picture the classic Lockwood, for instance] even that wouldn't bother us.

Even if such a thing is possible, it wouldn't work well for what I have in mind, though, because the thing as a whole is not a plane figure rotated around a single longitudinal axis.

L Cottrill

[Mike Everman]

I've been noodling the spinning of long cones, and came to the conclusion that you just start with a tube. You may or may not need a mandrel if you use a multiple of pinch rollers that close in uniformly in a floating carriage, like cam follower bearings. I've been wanting to rig this up on my lathe. The wall thickness will grow as the OD shrinks, and it will grow in length a bit while you work it, but it won't be ultimately any heavier than my Kazoo exhaust tubes.
Need the tube to be non-work hardenable, of course, or your second pass will be a b*tch.

[Bruno Ogorelec]

Guys, don't forget hydroforming. It's superb for long narrow pieces with modest changes of section.

[jmhdx]

Sombody somewhere is probably already producing exactly what we would like, chamber and tail pipe in a seamless high finish allowing us to play with the front end, if you can wait ten years I'll find it. The Dyna-Jet dies are probably still in existence and the right approach to the right man at said manufacturer may yeild dividends. I cant find their web site but they still produce fogging machines and such and I'll bet inside all the gubbins is that same old tube from the same dies.
Mike.