[Com]

Im curious to know why a heavy iron base it vitial for a sturdy machine? Is it possible to make a machine that will cut largish parts within an accuracy of 0.0001" that is constucted of a steel/aluminium frame? Whats the advantages of a heavy frame if any?

[Al_The_Man]

Rigidity.Anyone who has tried converting a drill to a small mill will tell you.
Al.

[mxtras]

Resonance frequency.

The larger the mass, the less likely the machine will resonate, or vibrate.

Aluminum, even in bulk, will still resonate - it's kind of like a rigid spring. Iron doesn't like to vibrate, aluminum does. Vibrations create chatter which can start one heck of a destructive cycle.

It's not 100% dependent on the material, though. Shape has a substantial impact on resonance. The use of iron is a safety factor to make shure you don't build an expensive, automated tuning fork.

Scott

[Karl_T]

FWIW,
It is possible to design a machine with a steel weldament and fill with concrete. The concrete provides mass and vibration damping. I've seen it done a a HUGE scale custom machine. Built in place, it will never be moved.

Karl

[DareBee]

[QUOTE=Com] Is it possible to make a machine that will cut largish parts within an accuracy of 0.0001" QUOTE]

Anything is possible (so I am told).
But IMO it is VERY EXTREMELY UNLIKELY.
That's why you pay over a million for the big accurate machines.

[mxtras]

Concrete is a good dampener, for sure. Low shrink grout is excellent for this kind of thing. If you do it, use a fairly dry mix and compact it with a vibratory settler to remove the air - you can rent one pretty cheap. Intimate contact between the steel and the grout after the grout is cured is essential for best results. It is a good idea to at least clean the surfaces that will contact the cement or grout and if possible, prime. Don't fill when the structure is cold - warm it up and let it 'soak' first. Get the filler material and the structure at about the same temp and keep them there to avoid internal condensation.

Scott

[CJL5585]

In the late 70's & early 80's the Japanese had some machines at the National Machine Shows and the frames were pre-stressed concrete. Other than this, I have no idea as to how they worked. I know that these units were advertised as being accurate to 0.00001 inch.

Jerry

[mxtras]

"Interesting" stuff them foreign folks come up with, huh?

.2 micron tolerance? Wow. That's ....uhm....pretty good.

Scott

[Com]

Looks like concrete could be the anser to my next project then, just a quick one, is bolting to a cast iron plate no good?

[DieGuy]

That is nuts, there are to many external factors to really get a large machine to have a .00001 accuracy, Now resolution is possible but accuracy, the thermal characteristics of the room, the floor isolation the load, the application of power and the friction of the system will cause more inaccuracy than that. Most gage block are only certified to around that. I work in a dimensional metrology lab that was one step from NIST in the certification chain from about 1983 -1988. I do know what I am talking about. We had samll single axis micrometers that would do about 0.00001 accuracy on a good day.

[mxtras]

The Japanese are famous for ignoring reality! I'm an engineer at Mitsubishi, so this is not a slam - it's an observation.

Very tight environmental controls are mandatory - all the way down to the way the site was prepared for the isolated foundation the machine would have to be set on.

OH - CJL didn't say how big the machine was, DieDude - I almost jumped to the same conclusion since that is the topic of this thread ("largish parts") - that the machine CJL mentioned had to be a monster, but heck - the machine may have only been a couple of feet long - who knows. Either way, I agree than sub-micron machines are not commonplace, for sure.

Scott

[DieGuy]

In my last big machine shop gig I did in the 80's and 90's we had some lathes made by Moore Machine that would actualy turn contour parts to about 0.00005 in. accuracy in about a 1 foot cube. They were on isolation foundations, in there own room with no operator and a precision air flow system much like raised floor computer rooms, used roller v ways, had air bearing spherical spindles, use laser for positional feedback and yaw correction, had 4th axis for the tool post so the cutting edge remained normal to the workpiece and used precision diamond tools. They literally cost millons way back then and were state of the art for there time.

Guess what the physics ain't changed since then.

[CJL5585]

Sorry I didn't mention what type CNC machine was built of concrete.

It was a very large Concrete lathe. Weighed thousands of pounds. I did not see the unit in operation, but was told that it would be doing demo metal cutting.

Other than this........ I don't know.
Jerry

[unterhaus]

A knee mill has a lot of wasted weight to get the table up to where you would normally use it. So if you don't mind squatting you can do away with some of that weight.
Stewart tables (hexapods) use the interaction of the forces in their arms to get away from heavy structures. It does seem that to get the same accuracy you'd have to go to very expensive components. EMC has the ability to drive a hexapod.
I've recently seen some concrete machines, I don't think the idea is going away.

[damae]

Interesting thread, about using concrete to dampen vibration. What about filling with sand?

[mxtras]

Dampened and tightly packed - I suppose that would work pretty well.

Scott

[damae]

I was thinking of a deadblow hammer. They're filled with loose sand.

But if you needed to dampen the sand, you could use oil instead of water, especially if you're packing the inside of a steel beam. The main advantage of sand would be that it could be removed easily if you ever needed to move the machine.

[ty1295]

Hardinge I think has used concrete in machines also.

[wizard]

In a nut shell iron dampens vibrations in a way few other materials do. The heavy part is more debatable but the machine does have to be stiff. Heavy does help though when strange loads need ot be handled.

As to cutting large parts to 1/10 thou that is a whole different sort of machining and a whole bunch of stuff comes into play. The question is what do you mean by largish parts? In any event my limited CNC experience has been with diamond turning on optical parts. We could hold tolerances to a few microns but that was over very small distances.

Just so you know once you start getting excessively tight tolerance wise everything comes into play. The air conditioning ducts and the rock music playiing on the radio to the quality of you lubrication system for the ways. When you say largish parts one would have to wonder why the tight tolerances.

Dave

Im curious to know why a heavy iron base it vitial for a sturdy machine? Is it possible to make a machine that will cut largish parts within an accuracy of 0.0001" that is constucted of a steel/aluminium frame? Whats the advantages of a heavy frame if any?

[wizard]

Back a couple of years ago when I was working on diamond turning equipment the machines, in part, where made out of a composite material that I called synthetic rock. Not sure who or how they where made other than the obvious that they where casted and then cleaned up.

Recently I've gotten interested in building some CNC hardware, mostly out of interest as a hobby, but you never know. To that aim I'm thinking about researching materials to make my own synthetic rock. So I geuss this is a good thread to ask this question in, has anybody tried this?

My thought is to go down to the local boating supply shop and get some epoxy resin. In fact I've been reading up on the west system. The big issue there is that they explictly reccomend agianst large or thick sections due to heat. A fire hazzard! Since the machinery I'm familiar with had very thick sections (6") this might be a problem.

The thought is to use what is commonly referred to around here as stone dust, throw in some fibers and mix it like concrete. The hope is that this would be easy and stable.

Thooughts?

Dave

Dampened and tightly packed - I suppose that would work pretty well.

Scott