History of Machine Tools

Ed,

Maybe you should give a lecture. Danbury isn't that far.

John

Well, I'm sure to the degree that the people who researched and wrote the rest of what I posted about it made a very big point of it around 25 years ago. American Machinist was deeply involved in reporting on it ever since Parsons got the ball rolling. The people who wrote that material above knew him well, and were in on the whole thing.

It was a modest computer they say, but a computer nonetheless, that the MIT/Air Force/Parsons research machine was controlled by.

Incidentally, there were tape-controlled machine tools going back to 1906, but they never went anywhere until Parsons got the Air Force involved. They were sort of machine-tool versions of the old punch-card-controlled weaving looms.

Ed Huntress

No, I don't think so, John. What you're describing sounds like some of the commercial machines that came out in the mid-'50s, like the J&L Binotrol, and the G&L Numericord.

I don't have access to the old reports anymore, but the MIT machine was described as a computer-controlled machine. It had a vacuum-tube computer controlling it. I don't recall if I ever read just what the functions of the computer were.

Ed Huntress

Not a celebrity, but it does make me certifiably old.

Ed Huntress

contributions

Nah, I'm no expert on it. I was just one of a team of editors who researched and put that history together. We each took a few slices of it and studied them for a year or so.

A great deal of the research material was right in the building with us -- the collection of American Machinist magazines going back to 1877, and the McGraw-Hill corporate library, which contained every important book published on machining up to the early 1950s. McGraw-Hill published most of those produced in the US up until that time, anyway.

And some of our editors knew the key people very well. Colvin had been an AM editor himself at one time. Dick Moore used to come in for lunch. Parsons was around from time to time. Andy Ashburn, who wrote much of the WWII training material for machinists, was our Editor when I was there. And so on. We had the living history right there with us.

The curators at the American Precision Museum in Vermont know more about it than anyone alive. If you ever get up there, stop in. It's worth it.

Ed Huntress

and...

Just for the heck of it...

Scan around in these newsgroups for a few minutes, and look at the number of responses, and number of different contributors, that make up an average thread. Now compare that with the number of responses, and contributors, that THIS thread has generated in just over two days.

Then do some Google searching, and see if you can find a thread or two about the origin of letter size drills, or about some of the machine-tool builders that have gone bankrupt or closed their doors during the last few years. How much interest did those threads draw, and from how many different people, and over how many weeks and months?

Then think a minute about what the answers to those questions might mean, and about how interesting history is to people who already work in this profession. And, if history is interesting to us, then maybe, just maybe, it'll be more interesting to your students than you might have estimated.

If the classes you teach are like most that I'm familiar with, then it's possible that the members of just a couple small newsgroups have generated more bytes of typing, and more real thoughts and ideas, and more honest, sincere debate on this topic, than all of your students put together will produce from ANY topic in an entire year - even though the newsgroupies don't have to be here, don't get graded for either attendance or contribution, and do this only for fun and personal enjoyment.

And if history IS that interesting, then isn't it just possible that it offers you a way not only to inform and train your students, but also to draw them into their work and studies, to capture their attention rather than just hoping for it, to help them WANT to excell in class, and beyond, for reasons that are entirely their own, and that might make everything else you teach them just a bit more compelling, and important, and worth learning?

I get the impression from one of your earlier posts that your heart isn't really in this, and that you're doing it because you've been told to. Your students will sense that, of course; and they'll probably follow your lead.

Lead thoughtfully and carefully, please. There are careers at stake, and lifetimes that will be spent in factories and machine shops, and food and homes and security and stability for families that don't even exist yet. And maybe a great deal more.

KG

Just in case it helps any, you seem to communicate pretty well. And you've obviously figured out that being a machinist means working with people, and not just machines. That's excellent.

Don't sweat the oral presentation. In fact, don't bother DOING a "presentation" at all. Just talk. Just say what you want to say, as if you were saying it to friend. Enjoy the chance to share what you've been studying or preparing, and invite everybody else to enjoy it with you. They will, if you give them a chance.

And stop back here as often as you can. Communication takes practice too, just like sharpening a drill or running a CAD system. And, if you don't mind the fact that some of us get crazy and cranky once in a while, you might just find the group to be a fun place to hang out.

KG

...

Ed and others, thanks for the interesting thread. Like many innovations there clearly were multiple paths involved. Here are a few related ones at least worth mention.

The Jaquard looms go back almost 200 years, punch-card controlled weaving machines. Interesting topic in itself, along with Babbage and his mechanical computing machines. Side note, you might enjoy Gibson and Sterling's sci fi book "The Difference Engine".

Few folks in this group under 40 may know what an automatic screw machine is, how they work or that they go back at least 120 years. Not that I know more than from reading a book. Bunch of change gears to set feed rates and cycle timing, and hand-cut cams to cycle through changes of tools, feeds and stock advance. Net result is loading a

20' bar, then coming back in 20 minutes to empty the bin of whatever part it just made 300 of. Need a turned length of .750"? Set up change gears for a cycle that feeds at .010" per rev and cut a cam that dwells on that cycle for 75 spindle revolutions. Something like that. Pretty darn close to numerical control. Efficient enough overall they are still in use all over the world today. The older Machinery's editions have sections on programming them.

While I know almost nothing about them, some pretty fancy cam-controlled milling machines were developed either side of WW2.

I'm not arguing that any of this is in any way the equivalent of direct numerical control, just that it was a long evolution with many overlapping concepts.

Hydraulic tracers also evolved roughly in parallel with NC and co-existed for a few decades before becoming "mostly obsolete". Still used here and there. Tracing goes back around 200 years, probably to the gunstock duplicating lathes in Whitney's factory that produced government muskets in the early 1800's.

Hydraulic tracer duplicating mills and lathes were a mid 20th century innovation, the new concept being the servo mechanism that allow a stylus to follow a template using very little pressure, while rigidly controlling a lathe carriage or mill table applying hundreds of pounds of cutting force.

Tracing and NC were both driven by the need to mass-produce parts having complex contoured surfaces that could neither be expressed with simple specifications nor be milled or turned using straight feeds or specially shaped tooling.

Some of the first practical NC machines were based on hydraulic tracer machines because the feed control mechanism was so close to what NC needed.

Bob

Second prize, *two* free cerfications!!

:^)

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

And along the same lines, cam operated spring forming machines. Often very complex programming via specially machined cams. Of course this has been replaced by CNC spring formers which are more suited to quickly changing jobs rather than set-ups for long runs.

However, is there really that much difference between holes punched in a card telling a machine where to go and profiled cams telling a machine where to go? I guess one may have a mechanical linkage where the other may have an electrical linkage. Either way, they are both effectively numerical control.

So where does the line cross to being true NC? Is there a definition of this point? I remember that my Comp Sci teacher once defined a computer as being such because it could modify it's own program if needed, rather than a set of codes that were simply repeated from memory. I wonder if there is such a (historical) line or point where mechanical becomes defined as NC becomes CNC?

Koz

Bob Powell wrote:

I like your support Kirk.

Good on you!

Stan-

Hi,

"-half a day of math per week - our book is Mathematics for Machine Technology"

Back in 1993 we used the same text. Ours was the third edition. We never got to Unit 60 Machining Compound-Angular Surfaces: Computing Angles Of Rotation and Tilt because many in the class could not keep up and unfortunatly the Connecticut adult education system leaves a lot to be desired and teachs to the lowest level.

IMO this book could be a lot better but it's at least a move in what I consider to be the right direction. I never do trig that long way like they have you do it in the book. Much easier for me to look at one of those charts that you can find in something like a Carr-Lane book.

Lately I have been doing a ton of automotive reading, downing a 120 page book about every 2 weeks. When I get done reading what I want to I might just get motivated to do Unit

1. ;>)

"but I've met a few machinists before I started school who would have been easier to work with if they'd learned a few of those things and followed them."

LOL. You will meet quite a few more. ;>)

BTW, what CAD are they using to teach you ?

AutoCAD ?

Enjoyed your post. Hope you stick it out because IMO you do have the right attitude to make it as a machinist. Have you decided yet what area of machining you would like to go into ?

jon

Ed

I just found this site which goes into Parsons and MIT.

John

and this site

John

Ever consider that because the future of machining in the U.S. is so dim that studying the past feels much better than studying what's current ???

jon

Haha! Ask Dave ("Why"). He still runs them. Belt-driven ones, at that.

Cams give you automatic control, but not numerical positioning instructions. They're a kind of proportional control, as opposed to numerical control.

The big difference, and the line of demarcation, is between control systems that are numerically programmable, or programmable with a discrete set of symbols or numbers, and those that require making a proportional-control pattern or model.

Could you guys maybe learn something ELSE new ? Such as how to snip the 180 lines of nothingness when all you want to do is say, "Best wishes ! hi to Mom and the kids !"

It's NOT that hard. Really, it isn't.

Right, the paper tapes are real programs, the punches denote ones and zeros, I think they're either ascii or G code. G code = excess grey?

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================