[Swede]

Does anyone have any experience using a German "Gol-Matic" CNC mill? The layout and configuration of these machines is very interesting and versatile. They can be configured into a CNC lathe in a few minutes of effort. The quality appears very nice. Even if you'd never buy one, they may give some inspiration for a home-build.

Gol-Matic: http://www.mdaprecision.com/

I am interested in creating a new metal-cutting benchtop CNC mill of a different configuration, possibly borrowing gol-matic design elements. For example, picture a radial-arm drill, except instead of a single upright column, it would have two. The Y-axis would be overhead, combined with the Z. The X would remain at the bottom of the mill. Such a mill would combine elements of a gantry setup with a traditional mill. I guess the reason I'd like to go in this direction would be to create a stiffer vertical structure for a home-made mill, rather than a single column. Also, the physical separation of the X and Y axes might make for an easier build with rails and ballscrews. The X-Y "Sandwich" I used on my first mill works fine but it inherently limits the travel somewhat, as the sandwich plates (think of the bread in a real sandwich) must carry the linear motion blocks. If the blocks are for 15 or 20mm THK-styled rails, the minimum sandwich dimension is ~ 7" or 175mm square, if one uses a ballscrew between the blocks. I'd also like to get away from the classic long and skinny work table, such as my 6" X 18", and go for something more square.

Anyway, a bit of a ramble. I'd like to try something different and versatile, that would allow easy motion hardware installation, yet be rigid enough for metal.

[ViperTX]

Looks like a Bridgeport horizontal mill.....with a moving head....there's a reason Bridgeport didn't make the head continuously moveable....I would look at having 2 gantry like structures with the axis suspended between them....the gantries would be stationary or fixed to the x-axis.....just thoughts.

[Meltdown]

These are very high quality mills with very tight tolerances.
Taig, homier etc are less than toys compared with this.
They are much more expensive too.
I've used one several times at a friends house.
They usually work wit PCdreh mill+ software which i also use on my mill.

[mwp]

Swede:

Good link. Those are really nice machines. Really expensive as well!

Your idea sounds much like a gantry mill with a moving table.

I think that a gantry-type mill could be quite workable as a serious metal cutting machine. And along the lines of what you are saying about stiffness, gantry-type machine structures tend to have higher stiffness than cantilevered ones-- IF the gantry/bridge is designed properly. This is plainly stated in Slocum's Precision Machine Design (good book) and pretty obvious if you ask me. You have good points about the relative ease of mounting rails, ballscrews, etc. due to the separation of the sandwitch in this type of mill.

I am in the process of designing a welded steel metal cutting gantry mill. However, instead of using a moving table, I am considering making a moving gantry (floor space issues). At this point, I am leaning towards a "ring-bridge" design. A ring bridge is basically a gantry that goes all the way around the worktable. This adds considerable stiffness to the gantry structure. I've seen some examples of this type construction on this site but none as rigid as what I am contemplating. I plan on making use of finite element modeling to optimize the design before building. I may throw in a regular "cantilevered" design just for comparison as well.

About what size work area are you aiming for on your new machine?

[Mcgyver]

That is a very nice looking machine, as it should be for $26k. Of all the cnc machines I've designed (in my head as I'm falling asleep) which I could make, none have contemplated the vast amount of machining needed to create dovetails in cast iron slabs, yet all the quality machines seem to have them. that and a large casting support and connecting the z axis. (I'm an amateur so haven't seen everything under the sun mind you)

I'd speculate that home brew designs tend towards linear bearing designs probably because they are a lot easier to make and have less drag so smaller drivers can be used, but would dovetails with their large bearing surfaces be a better way to go to produce a mill capable of doing serious work on ferrous metals?

I don't know. Maybe the starting point should be to quantify the forces involved - actually approach it like an engineer and design something to have a tolerable level of movement for a given maximum cut. Maybe I should spend some time learning engineering – it might be quicker than building it and finding it a boat anchor!

My objective sounds similar - to build mill that will work on the same materials and to the same tolerances as a typical tool room mill, although with lighter cuts as mine won't weight 2800 lbs!

A fabricated structure as merit, but you will have to normalize it after welding. My idea was to weld on say 2" x 3/4 strips where the rails would be mounted and scrape these to the same plane after normalizing.

The idea I've come up with is to build a fairly massive welded structure of plate steel and structural shapes, normalize the whole thing and then fill it with concrete & rebar! Not as crazy as it sounds (maybe), although I want to talk some engineers who work with concrete before pursuing. Apparently concrete has the modulus of elasticity of aluminum and does dampen vibration.

There’s a sketch attached – picture the column fabricated out of 10x10x.250 steel. This is all idea stage, none of put into practice. The danger of course is that if it was a crappy idea and just didn't work, I'd have a very heavy problem to dispose of!

Mike in Toronto

[mwp]

Hi Mike,

I don't think that a mill necessarily needs to use dovetail type ways to achieve the rigidity needed to cut steel. Considering the dificulty of scratch building accurate, stiff dovetail ways, I would guess that for most people it would be more likely that a machine built with THK type roller ways would cut steel acceptably. And the roller ways elminate several problems with dovetail/box ways, most notably stiction (a large difference in static and dynamic coefficients of friction that can cause errors) and high axis drive force requirements. I have seen several large mills that use the THK style ways as well.

You might want to check out some of the stuff on this site:
http://pergatory.mit.edu/rcortesi/portf/index.html
Especially under the industrial design section. Might give you some ideas as far as assigning numbers to your design, etc.

Regarding cutting pressures, Roger Cortesi (the guy's site I linked above) designed his machine for a 30N (7lbf) tool tip force and allowed for a .0013" deflection at the tool tip due to cutting pressure, worst case, in his error budget. His overall design accuracy for the machine is .001". I'm pretty sure that he is using a high speed spindle for such a light cutting pressure. I believe that for slower cuts, the forces involved would be much greater and that drilling steel would definitely require higher cutting forces. As a rule of thumb, I have been told that the forces involved in machining mild steel are roughly 3X that of aluminum. I'm sure someone here knows more about this.

Regarding the welded structure, I think that I might be able to get away without nomalization (I assume you mean vibration or thermal annealing) if I design it to use minimal welds, use strategiclly bolted joints, weld slowly, and don't weld near the rail mounting locations. Maybe not though.

I do believe that the use of concrete has some merit. What is used in machine tools is a special polymerized concrete (or at least a polymer additive to regular concrete, I'm not sure) that has good vibration damping properties. However, care needs to be taken such that the rebar or metal structure keeps the concrete from being loaded in tension (concrete has little tensile strength but good compressive strength). There is a machine shown somewhere on this site that makes extensive use of concrete.

I don't see where you are getting that concrete has the same modulus of elasticity(E) as aluminum- my references indicate that a 5000psi concrete has E=~27GPa whereas most aluminum is around 70 GPa. ? I could be confused though-It's late.


Mitchell

[Halfnutz]

Nice Machines of course, from Germany it is to be expected. Actually looking at the 1210 allready set up for 8600.00 and comparing it to a Ronco allready set up for CNC for around 7500.00 from someplace I saw on the web, I dont think it is that expensive at all. I was suprised it wasn't more. Compare it to a Tech Isel at about 25,000 for a flimsy gantry router with a Porter Cable "Milling Head" and it seems to me like a bargain.

[Mcgyver]

Mitchell, you are probably right on M of E, I was repeating what I read somewhere rather than quoting an engineering source, thanks for the correction.

Thanks for the link - nice site. I haven’t studied it in detail, but I don’t see where he’s accounting for feed rate or depth of cut in figure out the cutting force. Or at least I don’t understand how is power assumption in watts translates into depth of cut/feed rate. Instead of 4mm at 45k rpm, I wonder what a .5" hhs end mill at 300 rpm into free cutting mild steel would be? Graph this against depth of cut and then we'd have something to evaluate against! I've got to bone up my engineering skills before I can attempt that

btw, the picture I tried to attach (but failed to) had a mock up of an nsk rail - so I to am looking for a way to avoid machining and scraping all those dovetails! My notion was more speculation on whether there is a correlation along the lines of amateur = linear bearing, pro = dovetail/ amateur = non-ferrous, pro=ferrous.

Intuitively, from an ability to handle loads, like the lady in high shoes vs. the bulldozer analogy, the dovetail would seem a better. I’ve got learn more about the forces generated in cutting metal to be able to intelligently decide on a linear/way design.

The normalizing I was referring to was heat treating to remove the internal stresses present in the steel both before and after welding. You may be able to avoid this step, but I'd not be inclined to risk it (assuming of course that we have similar use/performance expectations of the machine).

The stresses in steel will move it all on its own without the complication of welding and to hold 1 thou over a complex fabrication is asking a lot of steel, even with normalizing. Second, as I'm trying to figure out how to maximize rigidity, I thought I'd want to maximize rather than minimize the welding. One t&d maker friend said he suspects it will continue to move a bit after normalizing and may require a tune up later on.

I also think a flat (scraped) surface on which to bolt everything is important towards minimizing internal stresses. Scraping is really the way to go – unless you’ve a HUGE mill/grinder and can clamp it without distortion.

The concrete & steel, being inexpensive and plentiful, seemed like a neat way to achieve a rigid, massive structure....but I acknowledge it needs more development! From my limited ability to read engineering texts, the damping loss for concrete is about twice that of cast iron which is twice that of steel. However its a strange languauge for me so I may be wrong. As you mentioned there are other materials that might be used similarly but with better properties (I wonder how much crushed granite and epoxy would cost)

Mike

[mwp]

Mike, check this out- information on polymer concrete vibration damping:

http://demg.penton.com/content/am/out_of_the_box.pdf



Mitchell

[scottsss]

I came across the Gol-amatic on the Wabco Yahoo group. I did a couple of Google searches but could not find any home users who had these machine. Since I figured I could fire off a quick e-mail for some info and what they thought of the machine.

I only found one web page that I think was in Austria. I would be interested in any other links people have found.

As for building a machine....Their is tons and tons of surplus hardware out their. I would find a piece that is close to you needs and start with that. Just my two cents. Large piece of metal will itself help dampen vibration. Just need to modify it to your purpose.

[Swede]

I think there's a lot of merit to using pourable materials into voids to stiffen structures and dampen vibrations. At one point I was thinking good ol' plaster, but the moisture issues may cause grief with corrosion.

Castable urethane resin would be good for a small machine but pretty expensive, I believe. Maybe one of the harder grades of silicone rubber as used for molds? I think I'd be just a bit leery of concrete, although I can't say exactly why. Straight portland cement? You'd want something thin enough to pour well and with ease, minimal to nil expansion, non abrasive in case it breaks down and migrates. You'd also probably want some type of vibratory wand or similar to kick the bubbles out before it sets and help it fill voids in the structure.

[ESjaavik]

Portland will shrink for several months, up to years if thick. It also is very weak in tension. For machine tools a polymer or epoxy concrete is used. A good shake is necessary. A good distribution of particle size is also necessary. You can find a lot of info on the sites of companies selling the chemicals. The ones I've studied are German so maybe not useful here. Casting in steel rails to fit roller guides are probably the best way to go for amateur use. It is still weak in tension, so some understanding of stress formation is necessary. And the machine will probably be heavier than a CI one, not to mention welded construction. But if you don't plan on any more moving, that contributes to low vibrations anyway. ;-)
The ratio of chemicals to stone is low (8-15%), and the stuff gets dramatically cheaper when bought in buckets. So cost might not be large compared to other solutions in low quantity like for example 1 machine. Large production batches use welded steel molds. In qty. 1, a wooden box can be used. If a sacrificial wooden mold is used that also obviates the need for draught angles. The shaker could be cumbersome and expensive, but I wonder if an old truck and a bad road would do?

[Swede]

Guys, I found an example of a bench mill configuration that really appeals to me. Check out these mills of apparently VERY high quality and of a speed and accuracy that are really impressive...

http://www.cmscnc.com/index.cfm?id=0...owse&pageid=40

Nice looking machines. Note the table proportions, more square than long & skinny. I think this semi-gantry style has a lot going for it in that the columns holding the spindle are fixed, hence can be made very strong and stiff. The Y axis is all by itself down below, while the X and Z are above.

A medium-sized bench mill could be engineered with Bosch 180mm X 90mm "heavy" sections, clad in cast iron or aluminum plates, as the uprights. Since the uprights are outside the tables motion, the bases of the uprights can be made with supporting angular braces, etc.

The X axis would need to be pretty stout. One limitation that I can see with this style is that the Z stroke might tend to be limited due to the overall size and mass of the spindle/Z section.

What do you guys think? All I need now is a CI table of roughly dimension 12" X 16" or 10" X 14". I could always buy a precision CI plate from Nolan supply or similar and cut a few T-slots!

[ESjaavik]

I like that configuration too. And I would make the portal stiffer in the Y-direction than in X. Both directions have to carry the cutting load, but Y will also have to carry the acceleration forces moving the Z-axis. The con with this configuration is the space requirement. But X would not be stiffer given a ton of V**gra! And Z acceleration forces have a very good load transfer too. The only one in question is Y.

Using a high power density in the Z-axis servo and the spindle will be more important than the X and Y servos that will not have to move themselves.

Think about what you will use the machine for. For routing this will be fine, as the need for large Z-movements can be overcome by choice of tools/holders. It may also give you a wider usage range if you can raise/lower the Y-movement including the fixed part (servos and all) and lock it in place like you do with anti-snipe woodworking
planers. Large Z-movements are usually necessary for compensating toolholder shortcuts or for deep boring. Do you need the last one, if not put more thought into the toolholders to get all tools at equal height. Yeah, I know that can cost work or money, but so can a machine that will cover all bases.

And regarding your Bosch sections, I think they will be fine clad in aluminum (epoxied?). Iron not so good, as the temperature coefficient is not the same as aluminum. When everything moves in unison, we can live with that when the accuracy requirements are not very high. But different metals laminated can bend and warp the machine and thus create non-linear errors.

[fireball08]

I have seen a couple of the wabeco mills on ebay recently, not sure about the models maybe someone has bought one there. I like that setup on the mill you listed Swede looks like a great machine to me, I found this website a while back and watched some video's on a similar machine made by Datron check it out says the run high speed spindles and us micro tooling. http://www.datrondynamics.com they also talk about feed rates of 1000ipm wow now thats fast, it's the fastest machine I have seen a video of cutting aluminum.

Nathan

[Swede]

I have seen a couple of the wabeco mills on ebay recently, not sure about the models maybe someone has bought one there. I like that setup on the mill you listed Swede looks like a great machine to me, I found this website a while back and watched some video's on a similar machine made by Datron check it out says the run high speed spindles and us micro tooling. http://www.datrondynamics.com they also talk about feed rates of 1000ipm wow now thats fast, it's the fastest machine I have seen a video of cutting aluminum.

Nathan

I feel inadequate every time I visit the Datron site. Click on the link above, then view the video like Fireball says. Amazing. They're ripping through plate aluminum like it was balsa... beautiful sight, brought tears to my eyes.

:rainfro: And it looks like they're cutting it dry.

[skippy]

Hi Swede,
I stumbled on this thread in the reverse order. I had downloaded the video last year some time from Datron's site (our very own "High Speed") and just now I was doing a clean up in my video directory . I played the Datron video and began to suspect that the video had been "doctored", i.e. speed up to look good for sales purposes. I then did a search here and found this thread which I guess means it's for real. INCREDIBLE! I haven't cut aluminium for years but I remember it only too clearly as taking forever, breaking bits, and hands and clothing full of aluminium splinters. How things have changed! Another good one to see is Bob Scott's 500_ipm video Thanks Bob.

http://www.cnczone.com/forums/showpo...84&postcount=8
Although it's short, it's good.

[philbur]

Here's a UK company that sells Gol-Matic Chinese knock-offs.

http://www.chesteruk.net/store/cestr...on_machine.htm

Judging by the Gol-Matic prices they are intended for no compromise commercial applications where portability into the middle of nowhere together with maximum capability is the primary issue.

For home use at that price it's probably cheaper to buy a bigger house.

Phil