The Work Envelope Problem

In a weightless environment, a machine tool slide can be designed to move a distance equal to its own length, facilitating machine tool self-reproduction (MTSR). However, here on Earth, such an arrangement results in premature wear of the mating slide base, and so we have the Work Envelope Problem preventing MTSR: If we cannot move a part its full length, we can't reproduce its length feature, that is, the work envelope of the machine tool is *smaller* than the size of the work to be produced.

There are sixteen possible subtractive machining configurations: the work and the tool can each be moved or held by hand or by machine, for example: Whittling the end off of a stick is part-hand-held, part-hand-not-moved, tool-hand-held, and tool-hand-moved. Milling the slide mentioned above is conventional machining: part-machine-held, part-hand-moved (in non-CNC machining), tool-machine-held, and tool-machine-powered. Sawing the end off of a 2x4 board to be used in house construction is part-machine-held, part-hand-moved, tool-machine-held, and tool-machine-powered.

The work envelope of some of these configurations is infinite in the direction of the primary degree of freedom, allowing solutions to the Work Envelope Problem. With great care, it should be possible to subtractively machine metal the same we we'd saw the end off of a 2x4 board. I explored this possibility yesterday on a granted Smithy Super Shop just a month old.

There are at least two types of machine tool self-modification (MTSM). In both, a copy of a part is modified. In subtractive MTSM, a feature is added to the machine by removing metal. Subtractively, I mounted a second saw table to the lift columns two weeks ago by drilling two holes in the table and fitting bushings to the column end to accept screws. To complete that cycle of MTSM, I'd use the fitted table to make the same modification to the original table. In additive MTSM a part is added to a machine tool, integrating it into the structure. Subtractive MTSM is usually associated with machine tool learning or evolution, and additive MTSM with growth. Both lend to MTSR.

I did another project using these bushings and the threads showed a chance of tearing out; I needed stronger threads, so I remade the bushings of metal. I had a bit of cast iron but it wouldn't clamp in the four-jaw chuck; it was continuous cast and not very square. Without squaring the bar, I could not proceed.

I could have mounted the milling table and milled the bar, but that accessory had not arrived, so I set up a metal cutting blade (zero rake carbide teeth) and shaved all four sides in both directions using the saw fence to get the part squared up, then switched to a coarse sanding disc to finish. I was covered head to foot in grey cast iron dust. I took a shower and vacuumed the dust.

With the squared bar in the four jaw chuck I was able to finish the bushing OD features yesterday. It wasn't the only option, but the infinite-in-one-direction nature of the saw configuration of the machine did facilitate additive MTSM in this case, leading me to a deeper understanding of MTSR, and opening the possibility of complete MTSR in the future on this or another machine by me or some other researcher.

I'd appreciate any comments on other audiences for this area of interest.

Doug Goncz Replikon Research Seven Corners, VA 22044-0394

Reply to
Doug Goncz
Loading thread data ...

Trimming sci.physics.research because their moderators felt the post was OT for their (moderated) ng....

Get an independent four jaw chuck post haste!

Mark Rand RTFM

Reply to
Mark Rand

Trimming sci.physics.research because their moderators felt the post was OT for their (moderated) ng....

Get an independent four jaw chuck post haste!

Mark Rand RTFM

Reply to
Mark Rand

I don't think that holds: consider the workers and equipment on an oil drilling platform as a "machine tool" -- they can move the drill bit several miles underground, but the "machine tool" itself is only a few hundred feet high. I see no _theoretical_ problem in constructing a machine tool that has work extending outside itself by distances larger than the size of the tool itself. There are many practical problems, of course....

Tom Roberts

Reply to
Tom Roberts

Examples: Railroad-track laying machines. Building bridges.

Nick

Reply to
Nick Mueller

A machine can build a structure larger than itself if it has closed- loop positional feedback that allows it to correct errors. If I understand what DG is looking for, this means the self-replicating machine tool would have to make laser rangefinders as well as duplicate all its own mechanical parts. The basic self-replicating unit may be a diversified company that owns both a steel mill and an IC fab.

jw

Reply to
Jim Wilkins

I can think of one work around, if you have the rigidity: Use a tool which has a work span which makes up for the loss of working envelope. If for example you flycut or face mill a table top, you can cover a larger area than the motion of the table alone. The problem with this is that I can only see it working with flat surfaces, anything with a profile (rabbets, dovetails, and the like) would still be limited to the relative motion of spindle and table, and thus subject to "work envelope creep", unless you were willing to make such features out of additive pieces themselves made up of flat surfaces (ex: a rabbet made by adding a rectangular prism rather than subtracting a rectangular area, or a v-way made by adding a primatic bar) which forfeits many of the self-alignment and rigidity advantages of machining the pieces from solid to start with. I did say it was a work-around...

--Glenn Lyford

Reply to
glyford

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.