Quote:
Originally Posted by
steve123
Here is the design for my 3' x 4' x 6" steel gantry mill. Any feedback before I start would be great!
Interesting design! I hope you don't take the feedback the wrong way.
Quote:
This machine will be able to mill steel, aluminum, wood, plastic. This is a hobby (not production) machine for making various one-off parts. I expect to be doing a lot of 3D work. I plan to add a 4th axis at some point.
Will you have enough Z clearance for a fourth axis? Obviously that depends upon your intentions but you can quickly run into issues here with most router designs.
Quote:
Why this design? I started designing an aluminum CNC router over a year ago, and just wasn't happy that you just can't cut steel with a router.
This is also a router is it not? As for cutting steel you need a machine stiff enough for the tooling you will be running and a spindle that can run at the right speed. It is a matter of design, routers can be designed to Handle steel.
Quote:
I had several aluminum designs, but didn't want to build them. After a while I decided to just keep making it more rigid until this 9th design. Gone is most of the aluminum, for rigidity and thermal expansion reasons. I settled on a 3x4' envelope because that's about as big as you can get with ball screws.
You can get some pretty big ball screws so I'm not sure why that is a limitation.
Quote:
The 3' wide cutting area makes the gantry narrow enough to be a reasonable weight for this rigidity. The sides (x rails) will be raised so that there will be virtually no sideways gantry motion.
That is only the case if X support is stiff enough. I'm puzzled a bit by why you went with a moving gantry design if your goal is stiffness. The common reason is space but yo seem to be concerned about the stiffness of your machine.
Quote:
The Y ball screw supports are attached to the gantry bottom beam, so the top beam gets no sideways force. I have incorporated triangles into the structure wherever I can.
Travel will be 37 x 49 x 12". There will be 6" clearance under the gantry, and 3" additional clearance if the bed is moved back. I plan to be able to mount a milling vise on the front horizontal beam. The spindle can move 1.5" past the front end of the bed so that I can cut dovetails. There will be lots of space between the linear blocks, so there will be no multiplication of forces. The Y Rails are 8" apart center to center. The Z rails are 5" apart.
I'm concerned about the Z capability and your interest in a fourth axis. The moving table might help there but then it limits what you can do with the fourth axis. Look at it this way add a fourth axis and you effectively have cut your usable Z in half. Actuall it is worst than that because of clearances. An alternative might be to have several beams under the table onto which you can mount vises or a fourth axis. With the fourth mounted on the table you could find yourself coming up short with capacity and suffering from excessive frustration.
Quote:
Construction: Welded steel (stress relieved gantry), concrete filling inside the beams, self-leveling epoxy.
Rigidity: Horizontal deflection on one of the two 3x5" gantry beams is 0.00043 with 100 lbs force. This is for a bare beam, so it will be better when fully assembled. The top and bottom beams of the gantry will be joined by welded spacers. It will have a sheet metal skin on the back.
Nice but will in vibrate? Deflection is one thing but it isn't in many cases a key issue.
Quote:
Dampening: Concrete in various places. Cast Iron Z if I can.
Accuracy: Linear rail beds will have self-leveling epoxy. The Z plates and gantry support bars will be ground. Linear blocks will have preload. Ball screws will have double ball nuts and AC bearings in the end support. I expect the accuracy of the ball screw to be the limiting factor, so .003 per ft. I expect that repeatability will be excellent. I will be using a KFLOP, so I may do glass slides and feedback later at least for the Z.
Feed back on KFlop will be greatly appreciated. I like the approach they have taken with the hardware and software but haven't seen one implemented locally and have yet to build my own machine.
Quote:
Speed (rapids): around 400 ipm (using 1610 (not 1605) ball screws, and low impedance motors).
Acceleration: Well.. A pair of 960 oz motors put out about 1300 LBS force at lower speeds... I'm not sure exactly how much gantry weight matters.
It matters.
Quote:
Materials:
Base: Front horizontal beam 3" x 5" x 3/8" steel, concrete filled, the rest 3" x 3" x 1/4" steel (750 LBS) + concrete. I'm thinking of maybe filling everything in the base with concrete. There are some recipes that include some aluminum powder to make the concrete expand slightly.
3x3 might be a bit on the thin side especially when it comes to getting good concrete fill. It isn't impossible though but it will be tedious. Think about how you will ram the concrete into long 3x3 sections or even 3x5 sections for that matter.
This isn't so much a structural issue as it is a process issue. You will want good stiff concrete to avoid shrinkage associated with looser wet concretes but you also need to get good fill to avoid gaps. Further all of this crap gets heavy real fast and as such you need to be equipped to handle the parts. If you can answer the question of how do you pack concrete in to a 2.5 inch tube then go for it.
Quote:
Movable Bed: 36x48x3" consisting of steel chanel around the edges, a rebar grid, 1" square stock in a grid pattern (drilled and tapped for hold downs), concrete, epoxy coating. 300 lbs. I haven't seen a bed exactly like this. It may be a crazy idea. This will probably be the first thing I make after I pour an epoxy pad.
Well as stated I'm not sure why you are going movable gantry in the first place. But then you also want a movable bed at 300 + pounds. Something just doesn't seem right.
On the other hand you do have an interesting design. The interesting task will be getting it cast flat. One thing to look into is a book written a few years ago that covers making concrete counter tops. Sorry I don't have the author or title but such a book might give you a few ideas. Honestly though making the whole bed epoxy concrete might make more sense.
Quote:
Gantry (Y): Two 3" x 5" x .25" x 48" steel beams (5" is horizontal), 4" tall spacers welded between, sheet metal skin on back. 150 lbs. I am considering going up to 3/8" wall. I'm also considering pouring 2" wide concrete or EG bulkheads in the gantry beams.
A number of pieces of equipment make use of welded steel beams for gantries. It isn't a bad approach but you will most likely have to have it stress relieved and finished milled for the linear rails. Structural steel isn't all that straight and any welding builds in stress that will come out sooner or later. This is probably the biggest negative with respect to steel structures.
I'd also consider a square beam to allow for easier fill of the beam with concrete.
Quote:
Z back plate: 1/2" ground cast iron or steel plate. Girders up the edges to control flex.
Z front plate: mostly 1/2" cast iron. Box section. Z total weight (back,front,spindle,motor) 45+ lbs.
Linear Components:
CPC AR20 profile linear rails from Anaheim Automation.
CPC-AR20MNBZV1P blocks, light (middle grade) preload, "Precision" (middle) grade.
1610 Ball screws with double nuts from Linearmotion2008 (2 on the Y).
BK12/BF12 end supports (will replace bearings with AC).
Spindle:
There is no perfect spindle. So I'm gaining to suggest interchangeable spindles as A low cost solution. For example one head using an off the shelf router and another that is a more substantial spindle for lower speed work. Or you can co mount the spindles and switch as needed.
I suppose if budget is of no concern you could install an industrial do it all spindle.
Quote:
Ah.. the spindle... That's an issue. I would like one that can do 800-15k with a couple of pulley ratios.
At minimum I will get an x2 R8 mini mill spindle from Littlemachineshop, and upgrade the bearings for 10k.
I'm also thinking of something like Mactec54's ER16 (But ER20). The current design shows this type.
Or dwalsh62's ATC spindles
http://www.cnczone.com/forums/produc..._cartidge.html
Spindle Motor (so far) would be a 2HP treadmill motor
Motor controller: KBMM125
You can't be serious? After all the time and money that will go into this you want to use a tread mill motor. To be polite that is nuts. If you have spent all the time and effort to design this machine, take the time to design in a 3 phase motor at the very least or a more substantial servo / spindle drive.
Yes! That is what I was alluding to at the start of this spindle section. Do a robust spindle for lower speed work and use a generic router for high speed work. It saves the expense of an high performance spindle.
Quote:
Coolant:
I'm thinking of using mist coolant when cutting metal. I'm not sure how to control flood on this machine.
Dust control:
Harbor Freight 2 HP
Noise control:
I'm thinking of building a wall with a door around the machine, air compressor, and dust control.
Electronics:
G201x Drivers from CNC Router Parts
Breakout board
960 oz/in, 2 ohm impedance NEMA 34 motors from CNC Router Parts. I will look into a kit from CNCRP when I get that far.
KFLOP
I have an old Dell computer
Software:
Alibre Design (have it)
Possibly Rhino
Meshcam
Kmotion. Don't know if I can get by without MACH3.
FsWizard
Free software for Vcarving
Tools:
I have a Chop saw, drill press, stick welder, metal lathe, small granite surface plate, height gage, dial indicator, test indicator, digital calipers, Master machinist level .0002 per 10". I will purchase a granite triangle for aligning the machine. I plan on getting an engine hoist. I will pour a 6' square pad of self-leveling epoxy to do layout on, build the bed on, and to put the machine on.
Work distribution:
I plan to have the gantry spacers and side plates water jet cut. I expect to have the Z back and front plates cut, ground, and drilled by someone.
It would be nice if I can get the tubing cut accurately.
I will probably make the smaller metal parts.
My brother and I will do the welding, aligning, drilling, tapping.
Budget: $5-6k US
Time frame: I will start after I get some feedback. I expect this to take up to year to build, mainly for financial reasons.
-Steve
Lots of details that don't deal with the bigger design issues above. What you ultimately select control wise isn't all that important as upgrades here are easy and frankly the tech changes rapidly. The one nice thing about KFlop in this regards is the steady stream of improvements.
From my standpoint I'd look a little Harder at the hard parts of the machine and define better your future wants such as that fourth axis. Six inches of Z axis clearance is great for many router applications but I see that fourth axis requirement as a potential problem. Also why the moving gantry if stiffness is an important goal. Nail down what is import at here before worrying about controls.