Ahoy all! My name is Christian.
I've embarked on an arduous journey to design build a 5 axis CNC router.
I have been following the CNC world for about 5 years now, and only recently got my own CNC Mill about 5 months ago. It's a CNC Fusion converted Sieg X2. So I'm not super experienced with machining, but I think I'm a quick learner on all things machining after being a keyboard warrior for so many years.
I've already started building it, and I've already got the vast majority of parts to build it, I just have to actually build it now.
Parts purchased so far include:
Drive gear and bearings:
- 12x SBR16UU 16mm dia. round rail bearing block
- 2x SBR16 16mm dia. 800mm long linear bearing rails
- 2x SBR16 16mm dia. 1200mm long linear bearing rails
- 2x SBR16 16mm dia. 1800mm long linear bearing rails
- 1x (1605) 16mm dia. 600mm long machined ballscrew
- 2x (1605) 16mm dia. 1150mm long machined ballscrew
- 1x (1605) 16mm dia. 1800mm long machined ballscrew
- 4x (BK12) 16mm ballscrew driven end support
- 2x Holdren Astra TS 4 stud non-ABS front wheel hub assembly 1998-2005
- 6000RS rubber sealed deep groove ball bearing (10mm x 26mm x 8mm)
Total: AUD$1388.51
Electronics:
- 1x Warp9 6 axis ethernet SmoothStepper motion control board with terminals for Mach3 and Mach4
- 1x C41 PWM variable speed control board with 2 relays
- 1x 5V/2A switching power supply unit (KL-10-5)
- 1x 48V/7.3A switching power supply unit (KL-350-48)
- 1x Unregulated linear 1440W/72VDC/20A toroidal PSU (KL-7220)
- 2x Geckodrive G251X
- 4x Geckdrive G201X
- 2x NEMA23 280ox-in 2.8A 1/4" dual shaft stepper motor (KL23H276-28-4B)
- 3x NEMA23 570oz-in 3.5A 1/4" dual shaft stepper motor (KL23H2100-35-4BM)
- 1x NEMA34 1200oz-in 1/2" single shaft stepper motor (KL34H2120-42-8A)
- 1x 2.2Kw air-cooled spindle
- 1x 2.2Kw VFD
- 75' of 4 conductor 18AWG shielded motor wires
Total: AUD$3320.46 (USD$2267.47 :O )
Material:
Length Size Material 16m 89mm x 89mm x 5mm mild steel square hollow section 8m 89mm x 89mm x 3.5mm || 4m 125mm x 75mm 5mm mild steel rectangle hollow section 1m 150mm x 50mm x 5mm || 2m 75mm x 75mm x 6mm mild steel angle 12m 50mm x 5mm mild steel flat bar 300mm 25mm x 50mm aluminium 6060 T5 flat bar
Total: AUD$749.07
Screws:
Qty. Head Size Length 16 socket hex, cap head M5 25mm 136 || || 20mm 24 || M6 50mm 48 || || 20mm 32 || M10 20mm 16 hex head || 15mm
Total: AUD$95.00
Grand total: AUD$5,553.04
$500 over budget so far... Who saw that coming? :P
I still have to buy a computer, I'm thinking an old Dell workstation, something with buffered memory and a xeon CPU and all that jazz to keep her reliable. Opinions?
I also need to get more wiring and such.
I was going to build the whole thing, then post a completed machine thread here, but I then figured I'd get more out of an on-going thread. I also hope this thread will serve as a wealth of information for someone building their first machine of this type, because there's little information out there on DIY 5 axis machines. I spent months scouring the internet for every picture and thread I could find to get ideas to make the design as easy for me to build as possible. I hope you can all input a lot of information to make this thread one of those 'gold mine' threads for the topic of DIY 5 axis.
The Details
I need this machine to be able to carve moulds for my vacuum forming machine that I also built just before this (pictured below). "What type of moulds?" you say? Well I'm glad you asked. I'm going to be making RC car body shells, so I need a machine that can make nice smooth forms and also reach minor undercuts. I still haven't sorted out what material these moulds will be, I'm not too familiar with the 'foam and plastics carving' side of CNC, so any advise would be GREATLY appreciated on what material would be best for a very smooth surface finish, is easily machined, and is dimensionally stable both over ageing and when being used to form a body shell i.e. having a -13PSI vacuum of force on top of it.
The size of the machine is much larger than most RC cars would call for, but I wanted to be able to make a body for all sizes of cars 1/32 all the way up to 1/5. I also love full size cars, so making fender flares mould or whatever else for full size cars was a consideration, but not critical factor in the design.
The total machine dimensions are approximately 2143mm x 1289mm x 2083mm, however that is just what the CAD design, it already change a bit since starting the actual build process. Because things seldom go according to plan.
Pictures and Videos
So with that out of the way lets get into some photos, because that all you really want to see, isn't it? :P
I'm just going to talk about some design philosophies and other things that I feel are worth mentioning as I go through the photos of what I've done so far.
Firstly, as promised, the vacuum former in question:
It's heavily based on the Protoform design by Doug at Build Your Own Vacuum Forming Machine.
Here is a drawing that I quickly made from the CAD design, I've changed the size of the gantry structure to something a little smaller because I felt it was a little overkill for the rest of the machine. Some other small things are changing, but I won't bother to model them unless I need to.
And here's a rendered view.
On the topic of the overall design, particularly on the gantry, I'd like to talk about an observation I've made on different styles.
My machine is a type where the gantry is basically a just a straight bar across from one side to the other and the cutting table frame raises the linear rails up to the required height.
The other type of gantry design is what I believe is often called a 'moving bridge' gantry where the rails that the gantry ride on are about level with the cutting surface of the machine and there are two large columns that go straight up and hold the cross beam section of the gantry rigidly. A bridge gantry is picture below.
This is a moving bridge gantry design.
There is also a very common bridge gantry design where the gantry itself stays still and the table moves on only one axis. I believe that design is to keep the weight of the entire gantry off rails and just moving the table. This is a good design, however it requires a lot of machine for not as much cutting area.
The design of my machine where the rails are raised to the height of the gantry (lets call it 'straight gantry') makes the potential for a rigid frame easier to achieve because you can use two smaller sections of (in my case) steel to hold the general gantry in place. Unless you have the tools to work with 200mm x 200mm sections of steel accurately, in which case I recommend you go for a moving bridge or a fixed gantry where the table moves, or a compromise of my machine's design and a moving bridge gantry (more on that in a moment).
On that "observation" I've made - when a tall machine like these is cutting material, the gantry will be handling forces from varying directions, the most significant one to me is when cutting in either the positive or negative direction along the X axis (the one that the gantry travels on) because that's typically where the least bracing is. The machine will never be cutting a piece of material that is level with the rails on my machine, therefore there will always be a leveraging action on the bearings that the gantry rides on, which can significantly multiply the forces exerted on them. This will be just fine if you have the right rails and bearings to handle this, in fact it's no problem at all. However if you (like me) have had to cut the budget a little and buy rails that may be a little under-sized from the machine, I think this is something to consider.
In this quick diagram I made is a side view of my machine design.
The orange rectangles are the linear bearings that the gantry rides on, the blue area is the most common space that my machine will be cutting, thus the length of leverage over the bearings. The arrows obviously indicate roughly the direction of force on each part.
The advantage of a moving bridge design is that the X axis rails are closer to level with where the tool is cutting, therefore reducing the the leverage on the bearings. Then the entire gantry kind of becomes one of those toy plastic eagles that balances on the tip of its beak and the wings cantilever past the beak to balance it out, and then the rigidity of the eagle (gantry) just has to be rigid enough not to flex under stress.
Where the point of contact with the finger is the cutting tool and the weight in the wings is the location of the linear bearings. (Do I sound crazy yet?)
Another advantage of the moving bridge gantry is that because the rails are closer to level with the work piece, any variations in the straightness of the rails will be exaggerated less in the location of the cutting tool as the gantry moves.
So by now you'd think that a moving bridge gantry is by far better than the design that I've chosen, and from all perspectives that I can think of, it is. However, I don't have the tools to work with the size of steel that I wanted to make a moving bridge gantry out of, plus (and more importantly) I realised this phenomenon after starting production of the machine so it would be too major of a change. Obviously my design still works, there's plenty of huge commercial machines that use the straight gantry design. Also, having the raised linear rails helps keep them out of harm's way.
Now more on that compromise design - a moving bridge gantry has the rails below the most common space to be cutting, so why not just combine a party straight gantry design and a moving bridge by just raising the rails half way? Well you can and I think it's the best way to keep extra stress from leverage off the bearings. There's this Chinese manufacturer that has chosen this design and personally I love the look of the entire machine, it looks awesome.
This design puts the rails level with the most common cutting height, increasing rigidity.
Here's their Alibaba page:
Hot Sale !! 5 Axis Cnc Machine * Cnc Milling Machine 5 Axis For Foam Wood Mould Decoration - Buy Cnc Milling Machine 5 Axis,5 Axis Cnc Machine,5 Axis Cnc Product on Alibaba.com
And that's the end of that obsessively long discussion on gantry design philosophy. I want to reiterate that I believe everything I just said is not important if you have a sturdy design and adequate rails and bearings. Moving on...
Here is a screen shot of how I make sure I can get all the right lengths of steel that I need after cutting it to 4m lengths for transportation. Not something that most of you wouldn't already have figure out, but honestly, it wasn't immediately obvious to me a while ago and I want this thread to be handy for a lot of people. 5 axis machining should be more accessible to the DIY community and I want to try help with that. Autodesk's Fusion 360 will have 5 axis CAM in the near future so hopefully I've come in at the right time to build this machine because Fusion 360 is going to be super cheap for the functionality that it has.
Short of spending 3 hours with a file on each end, this is the only way I can get square ends. This is 89mm x 89mm x 5mm square hollow section steel. So that's why I say I don't have the tools to work with 200mm x 200mm steel.
These plates go on the top and bottom ends of all four corners of the machine. The ones you see here have been faced off with a fly cutter on my Sieg X2 CNC machine for a good mating surface for when they will have another piece of steel bolted down, which will support the gantry rails. The others were just roughly cut to size and had an 18mm hole through the center then an M16 nut welded in place for levelling feet. The 18mm hole was made with a helical interpolation on the X2.
Sorry about the sideways-ness. It isn't sideways on my computer. That plate is to be welded to the bottom all four corners of the main frame with the nut on the inside of the steel hollow section.
To try held those foot plates square for welding, I machined all the faces of a scrap piece of the 89mm x 89mm hollow section including the ends, then clamped it in place for welding like so. It worked quite well.
I'm using a flux cored wire MIG for all the welding because that's the best I have access to right now. I don't have any real good reasons to not have gas by now, but as you can see I've managed to figure out gas-less welding alright now.
Welding the shorter ends of the machine together, only just fitting on my homemade T-slot welding table. The critical dimension to meet here was having the center piece the same distance from the ends nearest the camera because those ends are where the gantry rides. Obviously you want your gantry to be level with the cutting surface of the machine.
I ground a lot of the critical ends with a concave-like shape so it was easier to get a square end by only having to ensure the corners be square, then the plate to be welded on the end just rests on those corners when welded. When all you have is a 300mm disc sander for this job you take any opportunity to make things easier. I simply stuck a magnetic square to the end and measured off that face to get the correct distance for welding that center piece pictured above on the welding table.
If you don't know what I'm talking about, don't worry, it's just me cheating because I have less that adequate tools. Moving on...
To keep heat distortion down, I welded half a side, then welded half on the opposite side, then finished the other half of the first welded and so on... You can see where I stopped and started in the middle of each bead in that last photo.
Welding those longer sections in place. As you can see, I have clamped some scrap pieces to the already welded shorter ends, then rested the longer pieces on top to weld them level and all that jazz.
I wanted to talk about how I levelled and squared it all as I slowly welded it piece by piece so I made this video. Isn't my voice sexy? You should see my face! Haha (shut up Christian).
One thing I neglected to mention in the video is exactly how I'm adjusting the height of each corner to make it level for welding. I have M16 bolts in the bottom as feet for each corner, I showed those plates being made above. So obviously I just turn the bolt as necessary to level out each corner. Actual feet would have been much nicer for this, when you turn the bolt it sort of rolls on a high spot on the concrete and sends that whole corner out of square.
(This photo is also sideways, that center bead is actually vertical.)
Good enough for gas-less MIG? Still welding in half lengths to reduce distortion.
Doing some beginner level scraping for a good mating surface between the X axis rail supports and their mounting plates. This is before any scraping.
2nd pass.
3rd pass.
Final pass.
Same but opposite end.
I know only the outside areas are contacting and that's less than optimal, but I don't think it's that important, it's got the widest points of contact possible anyway, and I also just wanted an excuse to try scraping steel.
The plates are then mounted in their intended place, then I rest the X axis rail supports on top like this, ready for welding. The rest I've done it like this is to give myself the best chance possible of maintaining that large contact area after welding. If I welded the plates independently than just bolted the supports on top the plates would not be level and thus the supports would not have a good mating surface.
The plates are tacked in place then the X supports are unbolted so it's easier to weld.
First welded...
'A grinder and paint makes me the welder I ain't.'
A whole bunch of brakets for holding ballnuts and ballscrew bearing blocks. They're yet to have holes drilled in this photo. I'm extremely skeptical that these will be thick enough to take the forces necessary, but I've already got the material. If they flex too much I'll do something about it.
I then machined the ballnut blocks. These are awesome looking little things, but for some reason I managed to cut my hands four times that day, including one that had a stationary carbide end mill slice through my nail.
All six M5 screw probably aren't necessary, but again... They look awesome.
A bit of a mock up of the X and Y axis ballscrew assembly.
MacGyvering a drill press out of a megnetic drill because my actual drill press sucks. I'm drilling the holes in the X axis carriage here.
I'm using this little template hole pattern for the mounting holes for the linear bearings ignore the four bolt pattern on the right, that one had a mistake. I simply square the template up against the sides of the carriage material and transfer the holes. The center hole that the transfer punch is resting in is a hole for adjusting the set screw in the top of the bearing to maintain tolerance on the rails.
And that's all I have for this introductory message. All future progress will be posted in new messages throughout the course of production. All advise, or just opinions are welcome.