Talk:Numerical control/Archive 1

Industry standard protocols

Does anybody know about the standards used for the files to control the CNC machines? ISO standard?? noooooooooooooooo

The simple answer is that there is no Standard. There are a couple of areas where there is an approximation of one.

• Since the mid 1980s, the Electronics Industries Association (EIA) has published RS-494. It is a multi-part specification and has been partly adopted by the IEEE. This is probably closest to an industry standard.
• ISO has only ISO/IEC 9506, Industrial automation systems - Manufacturing Message Specification which deals more with headers and encapsulation.
• ISO also has ISO 10303, Industrial automation systems and integration - Product data representation and exchange, but this deals more with specs for the part being machined, while NC wants to look at the steps needed to machine it.

Sorry I can't get you a clear answer. The are a number of semiproprietary software solutions based around the RS-494 specification. --Lou I 15:02, 15 Sep 2003 (UTC)

The G-Codes are standardised in ISO 6983.

The newer ISO 14649 offers a feature based alternative.

ISO 14649 has been incorporated into STEP (ISO 10303) as the ARM for ISO 10303-AP238.

Cheers. --Dinian 21:15, 17 Apr 2005 (GMT)

The STEP-NC page discusses ISO 14649 / ISO 10303-238 and also briefly mentions BCL (RS-494)

Loffredo (talk) 18:37, 29 October 2008 (UTC)

"The" Machine Tool Company

There's an external link to Mori Seiki subtitled "THE Machine Tool Company". That's a slogan from a marketing campaign and seems to be free advertising since there are literally hundreds of machine tool builders. In fact, Mori Seiki is one of the world's largest five or six machine tool companies, certainly not "THE" MT company. —The preceding unsigned comment was added by 24.223.223.235 (talk • contribs) .

Point taken, adjusted accordingly — Graibeard 05:59, 19 January 2006 (UTC)

Commercial links

The question has arisen about which commercial links belong here.

• Mitsubishi Electric Manufacturer of CNC controllers
• Siemens AG Manufacturer of CNC controllers, drives and converters
• Aswiss Machines
• Industrias Romi S.A., machine tool builder in Brazil
• Haas Automation, Inc., machine tool builder in the United States
• Mori Seiki, machine tool builder in Japan.
• GE Fanuc Automation Europe S.A.
• Fundamentals Of CNC
• Introduction to CNC Metal Spinning Techniques
• Flow Forming and Shear Forming w/ CNC
• Matsuura - Japanese Builder of High Speed 5 Axis, Horizontal & Vertical CNC machines, & ultra high speed Linear Motor CNC Machine Tools

Are there too many of them here? -- ArglebargleIV 20:47, 14 August 2006 (UTC)

Commercial sites sometimes have interesting information on them, links should point to those parts of those sites. —Preceding unsigned comment added by 196.2.124.251 (talk • contribs) 16:37, 27 August 2006

I'm about to go through and remove all linkspam, which is really bad on this page. I figure I'd give 24 hour notice for anyone watching this for comments/concerns. - Toastydeath 22:29, 22 November 2006 (UTC)

Please do! This article suffers from advertising. Perhaps a good idee to list below the removed advertisment links and the IP adres that posted them. This way repetitive spammers can be detected and warned. --Jurriaan van Hengel 07:46, 23 November 2006 (UTC)

I went through everything, and most links were outright spam. Going back through the history and trying to find all the edits that added the links would be difficult and time consuming. I'm content to just remove them now, and nail the spammers when they return. That being said, here is some removed stuff I felt had no real bearing on the topic of CNC, but may be useful in other articles:

• Build Your Own: Part 1
• Build Your Own: Part 2
• Build Your Own: Part 3
• Machine Tool Help and CNC resource center, Wealth of information on nearly every topic related to machine tools

And conversely, here is stuff I left IN which people might object to, with an explanation of why I left it:

• CNC Zone, Large DIY CNC forums - I thought keeping this might be a fairly good resource to anyone looking for CNC help/etc that wikipedia does not provide.
• The Enhanced Machine Controller, open source CNC control software - It has a machine emulator, so someone wishing to learn in-depth CNC/G-code programming can use this for free, hands-on experience.

If anyone objects to the flow-forming/metalspinning links, I'd vote to remove them. They have basically nothing to do with the article, but I felt bad leaving just 2-3 links in the section.

-Toastydeath 18:42, 23 November 2006 (UTC)

Simplification?

I just came across this article via various random links, and noticed almost immediately that, in the way it is worded, to the average non-manufacturing/metallurgical/industrial engineering layperson, it is actually almost impossible to easily understand exactly what a CNC does without having to go much further out of one's way than should be required when looking for an encyclopaedic yet succinct account of a specific subject.

Assuming, as I do, that CNC's take and record all the required measurements of a master shape for sending to a tooling machine, would it not be useful to introduce the subject with a simple statement to such an effect - in laypersons terms - eg "a CNC is a computerised programming device allowing measurements from a master design which have been inputted into it to be fed into a cutting/tooling machine that will reproduce a specifically shaped/sized part accurately" or something to that effect - as it currently seems to be such a densely worded article thus making it almost entirely useless as a quick reference to anyone who would not already be quite clear as to what it is?!

Godgirl 02:32, 8 June 2007 (UTC)Godgirl

Who buys them?

It would be very useful to include a section on who buys them, maybe with some generic examples. Thanks —Preceding unsigned comment added by 75.33.96.254 (talk) 00:04, 29 October 2007 (UTC)

Photo Caption

the caption for the Haas CNC reading "a low quality"... doesnt seem NPOV "We're just going to have to use our brains."............."Damnit." (talk) 13:23, 12 May 2008 (UTC)

You're right. Kevin (talk) 07:28, 13 May 2008 (UTC)

The example program on CNC

That program is a joke. Not to be mean or nasty or anything. A true programmer should do something about that, it has a Lathe Tool change Code and g93 environmental code for a milling machine. Some one might try to run a program based on that very basic"Logic" A more comprehensive program should be submitted. —Preceding unsigned comment added by Operist (talk • contribs) 13:48, 31 October 2008 (UTC)

Rather Dated Article

"Computer numerical control (CNC) is a computer "controller" that reads G-code and M-code commands and drives a machine tool". This seems over specific and dated. I use a CNC machine which is controlled directly from job designs constructed using a graphical user interface. No G-code or M-code is involved at any stage. I know this because I wrote both the PC software and the embedded software that runs in the machine. The software could easily be extended to produce or accept some standard code format but there is no current requirement for it to do so as in many environments designs are purely for local use on a single machine. —Preceding unsigned comment added by 86.139.109.135 (talk) 21:22, 19 November 2008 (UTC)

Request for clarification (2006)

I don't understand this sentence: "His concept was to machine to setpoints as guides for subsequent manual finishing, that is, to speed up a manual process so more points could be included." I find the author's statement ambiguous, and I hesitate to try to clarify it because I cannot get the sense of what is meant.

I believe the original author was explaining the crude steps taken 60 years ago to create an elliptical or curved shape on flat material by intersecting short lines through a dot to dot pattern representing the desired shape. The more points used, the smoother the curve would become. Today's modern equipment still only cut straight lines. They simply cut a large amount of very small lines that make up a circle. 12/13/06 AD

I believe that the author means that the machine would only cut the dots, and a machinist would later connect the lines. Later, MIT improves on the design by connecting the dots with lines. --JB Gnome (talk) 13:35, 26 August 2009 (UTC)

C programming language?

Can some1 guide me with respect to hw programming language C is used to automate these machines. i hear one can use turbo C to write assembly level programs and then connect it to automate the machine.Where on web can i get further details about this. 12/1/07 AD

This is false. Modern machine tools have complex, multiprocessor controllers. G-Code, as an interpreted language, fills all the needs a machine tool is capable of without the need for a higher level language. A couple open source CNC projects expose the axis controls directly, but you will not find this on production equipment. - Toastydeath 15:03, 12 January 2007 (UTC)

POV question

This passage strikes me as particularly biased in its wording. It may be a summary of the thesis of David Noble's book "Forces of Production."

Overall I'd like more citations on this page.

Record/playback is different from numerical control in that the program is produced by the machinist in the process of making the first part. The Air Force wanted numerical control and not record/playback because 1) the latter put the machinists who were union members in charge of program production, thus union strikes could result in unacceptable delays in military production, and 2) numerical control demonstrated the capability of producing complex parts that were not possible by the conventional manual methods used in the record/playback technique. The Air Force used its deep pockets to get its way and while American manufacturing may have been better served by the simpler Parsons concept or by record/playback, today this is a moot issue.

159.53.46.141 20:43, 3 July 2007 (UTC)

Yes, this is not only biased, but flat out wrong. Record/playback was only "viable," if you want to call it that, in the absolute infancy of NC controls. Tracer systems are more practical and useful (even today), and those predate record/playback. It's a factually baseless passage meant to curry favorable public opinion in an old argument that is no longer occurring over CNC and the future of manufacturing in America. - Toastydeath 00:15, 8 November 2007 (UTC)

I am not sure that I am the one who actually did the entry for Numerical Control, but the text is certainly mine from other materials I have written. I don't see the bias. Also, with regard to the statement that record/playback was only viable in the infancy of NC controls, Mr. Nobel's book explains that record/playback was one of many alternative technologies that never got out of the lab because the U.S. Air Force used its deep pockets to promote by-the-numbers. Record/playback was not a NC concept; it was a different concept. Today we have so-called "teach lathes" that are a variant of record/playback that is achieved with CNC. Siemens, who I work for, has an HMI called ManualTurn. The operator uses handwheels and dialog boxes to make the first part. This process produces the program that can be run to make the 2nd, 3rd, and subsequent copies. If we want a historical precedent for this, record/playback as it was developed by GE in one of their labs, seems to be as good as anything. —Preceding unsigned comment added by 208.5.44.21 (talk) 15:16, 8 November 2007 (UTC)

Record/playback DID get out of the lab, and it did get into production. Whereupon it was found to be a huge waste of money and time for the businesses attempting to use it. You can still pick up old R/P machines, but most have been converted to numeric control. The problem is simple - record/playback is akin to using training wheels on a bike. The union aspect is very much overemphasized in the current wording of the article, as NC/CNC is a technically and practically superior control methodology. Even "Teach" lathes do not rely on the record/playback principle. They rely on conversational programming on top of a CNC, which has nothing to do with record/playback. Older Dixi horizontal boring mills can be had with 'true' record/playback, and it is nowhere near as useful or fast as the conversational CNC variant. Record/playback was never viable on technical or practical merits, and it has been given a chance. It just failed miserably. - Toastydeath 22:26, 8 November 2007 (UTC)

I just changed that sentence and then came here and saw this discussion. ("may have been better served" etc.) Maybe I misunderstood the original author's exact point about record/playback, but I have to ask, perplexed: How could anyone ever use record/playback to do the wicked 3D diesinking that AutoCAD and Mastercam software team up to do nowadays??? And even if it was humanly possible, and a few amazing people could do it, how would that have anything to do with practical use in the tool and die shop down the street? That doesn't make sense to me. — ¾-10 02:00, 6 February 2008 (UTC)

Noble 1986 (Forces of Production) looks like an interesting book, regardless. Maybe I should read that before editing more here. — ¾-10 03:49, 6 February 2008 (UTC)

Update, 1 year later: I read Noble 1984 (Forces of Production, hardcover edition) last summer. Fantastic book; important reading for anyone seeking insight into the history of technology, regardless of your politics. You don't have to be a Luddite or a syndicalist or any other kind of -ite or -ist in order to appreciate, and learn from, that book. Anyway, given Maury's recent helpful merging and reorg of NC and CNC, I further developed the pre-NC history section, and I found a logical place to state the theme of Noble 1984 and give an inline citation to it, but avoid opening any can of worms that is beyond the scope of this article. — ¾-10 02:41, 10 February 2009 (UTC)

What is numerical filing

72.27.54.23 (talk) 17:59, 8 April 2008 (UTC)can you please tell me what it is

Maybe you were looking for info on filing methods. — ¾-10 02:41, 10 February 2009 (UTC)

Recent cleanup was well-founded (drat!) but article development continues

Hi Maury! I have to admit regarding the stuff cut away in the recent cleanup that I can see things through your eyes, and I must accede to stetting your changes. Drat! :-) The info is central to a circumspect understanding of the topic, but its presentation and structure has to fit our medium, alas. I feel strongly that the sentence that cited Noble 1984 needs to "rise again" (or "live to fight another day" perhaps), but the appropriate section of the article will be a not-yet-existing "Societal implications of control technology" or "Criticism" section of one kind or another. That's a critter that I'll have to properly incubate in future. It also occurs to me that maybe we'd have to unify that discussion on a superset level above both machine tool control and industrial control systems. This is still blue-sky thinking as of today. IRL I am busier than a monkey on a banana plantation, so it may take a while to end up living the dream. Anyway, in the meantime, my best regards to all who find control technology, from its development to its social impact to its everyday application, to be one of the most fascinating topics that almost no one cares about. — ¾-10 01:08, 14 February 2009 (UTC)

Well first off, don't think I didn't like your edits - I think they really improved the article, and didn't mean to remove them because they were "wrong" or "unimportant" or something like that. In fact I couldn't agree more; Noble's work is important to include and the current mention is definitely too small. The fact that so much has been written about NC and it's effects on society suggests that this whole topic is much more important that the article would currently suggest (the article is, currently, technical). However, I think it's very important that articles be balanced, and leaving just Nobel in it I think violates NPOV. Bear with me here, this isn't some wikilawyer BS...

Many authors present what I would describe as a purely social view of the topic, and downplay the technical side. Given the astonishingly complex and expensive machinery in question, the technical side needs to be at least discussed with the same level of detail. When you do that, it becomes clear that a purely social view - that NC was an attempt to kill the unions - is unbalanced. The AF didn't invent NC, Parsons' did, and he did so in order to solve an extremely complex technical issue, one that Lockheed failed to solve. It was that technical failure that led to the AF funding Parsons' work. Their reluctance (and Lockheed's) to fund development of NC before the failure suggests that the union-centric view is inaccurate. And that's just Parsons. GM didn't introduce NC to get rid of unions, their problem was in the white collar side of the company, where huge amounts of resources were wasted pushing paper. DAC-1 was the solution, and the fact that DAC-1 could be turned into a CNC system fell out of the project.

It is also important to note that older works present what is essentially an older viewpoint. Automation, of which CNC was one facet, was widely presented in the 1960s as a social evil. One might consider Future Shock as typical of this viewpoint, and I remember watching a US news-show on the topic in high school (some years later). More recently, however, the reverse theory has started to become popular. Right now I can sit at my computer, design any part out of practically any material, click one button and then it will arrive in the mail a few days later. For a little money, on the order of an expensive oven, I can buy a machine and do all of this in my den. When you're discussing the social implications of CNC, this is, IMHO, much more important to point out. And we're only part way through this revolution; it has been widely commented by futurists (notably James Burke) that as the cost of CNC falls that in the not too distant future practically everything you buy will be customized in some way, because the idea of mass production that defined the first industrial revolution will be so diluted that the incremental cost of customization will be less than, say, packaging the resulting product. I think it's vitally important that this view be presented. Perhaps it is true that CNC had a negative impact on society, at least some small part of it, but it is equally true that it has had enormous positive benefits for all of society, something that histories from the 1980s would not be able to comment on.

My original idea was to have a separate section on this (entitled what though?), but I'm worried that what we end up with what is basically a book review in the middle of the article. What we need is alternate views, especially critiques of the book, but I haven't found any -- but York is right up the road from me, so this is not impossible. I think a discussion of this in the NC proliferation area does make sense though.

Soooo.... I'm all for a new section, perhaps just after the history section, that puts all of this as a "first class" topic within the article. As I said earlier, if this much has been written on the topic, then it's absolutely a requirement that it be talked about here (consider any article on GM foods that didn't talk about the social backlash, for instance). If you can summarize Nobel, and add in similar writings from other authors, I'll add more recent comments like Burke's.

Maury Markowitz (talk) 23:48, 15 February 2009 (UTC)

I agree completely with all of the above. To oversimplify the topics for the sake of saying something from the 30,000-ft level, I think that so many of the problems that come across (between the lines) in Noble 1984 as "this is why automation is socially evil and, BTW, also technically defective" can now be seen, a quarter-century later, in a different light, which is something like, "Yeah, those growing pains sure did suck, and development goals and capital weren't always applied smartly, but those problems were speed bumps, not brick walls, on the way to practical, pervasive CAD/CAM and instant/direct manufacturing, which are already realities today (although instant mfg is still primitive compared to what's coming) and are only going to improve with every passing decade." One high-level truth that can be stated rather simply in this yet-to-be-written "social impact" section (or whatever it is eventually called) is that automation increases labor productivity, which means that fewer human-hours are needed per unit of manufactured output. That fact is both a blessing and a curse, depending on who's looking at it and how. The "social impact" section will look at that from various sides and cite various sources. Of course, it may take a while to fully develop because we volunteers (you, I, et al) are working in our "spare time" (I like to live under the beautiful delusion that that actually exists). In the meantime, my regards! — ¾-10 01:09, 17 February 2009 (UTC)

Well do you think you can whip up three of four paras on Nobel that will do the work justice? I can do the same for modern works like Burke. We'd need to have a lead-in section on thoughts in the 1960s and 70s which would be more generic and include automation in general. But that sounds like a complete "level one" section to me. Maury Markowitz (talk) 12:45, 18 February 2009 (UTC)

I don't have the book at hand currently because I borrowed it via ILL when I read it. Will definitely return to this when I get a chance. — ¾-10 00:19, 22 February 2009 (UTC)