[dimmaz88]

Hi all, I've started planning a big build that will be 5-axis + a turntable.

I looked into buying a used robot but the cost of repairs and software put me off. Anyway, where's the fun in buying something!?

The machine will be used for milling soft materials like polystyrene/polyurethane etc, if it turns out rigid enough for wood then that'll be a bonus.

Attachment 496058

Attachment 496054

Attachment 496056

As you can see she's going to be a big old girl!

I'm in the middle of modelling the 2-axis spindle head too.

The frame will be 100x100x5mm steel box section, the Y axis (normally Z on a 3 axis) will be 100x100x6mm Aluminium box section.


My questions are;

1. What would be best for Z axis motion, ball screws, chains, R+P?
2. Would properly sized steppers lift the Z, or are we in AC servo territory?

Any input appreciated.

Edit to add- Obviously I've made no effort in this model to make this actually buildable, I'll work out the details later.

[Algirdas]

No steppers for axes.
Ball screw.

[ardenum2]

you make 2m tall buddha statues from foam?

[dimmaz88]

Not yet, but that kind of sculpture yes.

[routalot]

Try building a quarter scale version first,the mistakes will be cheaper and you can take the lessons learned on to the big machine.A bit of triangulation might be a useful thing.What software are you planning to use for generating toolpaths and will you be able to customise a post processor to work with that axis configuration?

[dimmaz88]

I'm probably going to use Fusion for CAM, followed by IRB CAM.

The model is just a visual aid, it won't look much like that when finished.

[peteeng]

Hi Dim - 2.5m is too long for reasonable ballscrews. I recommend you go helical R&P on all axes. If your happy with lower tolerance you could us 30mm wide belts. These can be very long and work well. I've used 16mm wide for timber and hard plastics but would use 30mm in future. R&P is stiffer and easier I feel and then you don't have to build your own belt tensioners. Belt tension is really important to get it right. So I think helical R&P is the go... Steppers will be fine. Then there's the Z axis, it will have to be much bigger than you have modelled maybe 200x200. Its a long cantilever and will vibrate badly if its skinny. Your config is similar to what clay carvers are in the auto industry... Since the Z axis will poke out the rear the column should be triangular like a pyramid so its stiffer plus it won't vibrate as bad as a tall parallel column will.... Keep adding detail.. The devil is in the detail Peter

re steppers vs servos. two things 1) budget - servos are twice the cost 2) depends on your speed requirements. Figure out your feed speeds and this will answer your question.... An overhead gantry with walls is potentially stiffer then your arrangement. Plus you need to check that your CAM system will be able to produce a post that agrees with this physical configuration. If it can't you will need to go to a conventional machine config....

I wouldn't have the rails "inside" the column. Chances are that they will jam plus you can't machine them if this is a welded frame (I can think of one way to do it. Really important to think through the entire build process and make sure you have machinists with big enough machines to do whats needed plus heat treatment plus cartage etc etc...). If its a welded frame then you need to have it stress relieved prior to final machining. As drawn it will be very curvy post welding.... I suggest a completely bolted together machine so you can adjust it as you assemble it.. Its a big machine and it won't take much to have sticky or jammed cars if something is not right.....

[dimmaz88]

Hi Peter, I plan to build the Z axis as 2 separate "ladders" and then machine the areas for the linear rails. I'll do this on my current 3-axis router. I had considered welding or bolting, and I may go with bolting like you suggest. I like the idea of R&P if I'm honest, my current machine is and it works great. Would I have to have braked motors on the Z in case of power failure? I assume it would just drop to the ground if they weren't braked.

The idea of building the machine vertically instead of a gantry is purely for space. When the machine isn't in use it will take up a lot less room.

I don't envision needing to have it moving overly fast, it will be a kind of set and forget machine. Should the part allow me I think the majority of toolpaths will be a mill turn, like a giant lathe if you will.

How do you mean the rails will jam if they're inside the column?

[peteeng]

Hi Dim - Being inside you have to have careful control of how far the columns are apart & that they are parallel. In a usual set up you can adjust in two planes. As drawn they can only be adjusted in one plane. But if the column spacing can be adjusted via bolts then this issue disappears. Square rail linear systems are very sensitive to misalignment and will jam if the rails are not planar and parallel.... Use helical R&P not straight cut. Then you don't need a spring system to keep correct engagement, but you do need to have the rack mounted correctly. Helical provides zero backlash.... (or very little) Braked motors or a counterbalance would be good. I use braked motors and they work fine. Dropping to ground is a big problem if the machine is in the middle of a deep cut!! Nasty... I'd stay away from welding. I would use open sections (C sections and I beams so you can do really hard bolt ups) and even consider heavy aluminium sections. If you have a router then you can make these parts in the router....

Whats overly fast? 2000mm/min 6000mm/min 20m/min you need to have a figure in mind to design the drive system.... The working envelope dictates the footprint. I'd expect a BC 5 axis machine with walls to be similar footprint.... You are forgetting that the Z will stick out the back and this space will not be usable. Since its a "lightweight" machine make it on castors so you can push it into a corner?? Inertia loads may want to move it around especially now your thinking about 10m/min feeds. This is possible in styrene, commercial routers cut MDF faster than 10m/min...

[dimmaz88]

Ah yeah, I thought you meant jamming from debris. I plan to build in adjustment for aligning all rails.

I think helical rack and pinion would be the best option. I had considered a counterweight but it didn't occur to me that it would stop the Z freefalling ??, that would be a good option I suppose.

I'm thinking a max speed would be 3000mm/m.

[peteeng]

Hi Dim - 3000mm/min with a 10mm pitch screw is 300rpm so easy for a ballscrew so even 6m/min is doable. With R&P much quicker speeds are possible, I'll do some math. I'm designing a drive at the moment with a 30mm pinion. This gives 41m/min at 500rpm easy for a stepper. I shall use a 5:1 reduction gearbox on the drive so its actual speed will be 8200mm/min and provide 115kgf at slow speeds and around 50kgf at fast speeds plenty for what your doing. This is from a 3Nm Nema23 motor... Peter

[ardenum2]

I have to say I'm very scared by your build, everything I've learned over the years about machine design makes me uneasy about such an approach. a moving column with an extending ram would be much stiffer and not harder to make, I'd say easier to make than a box-in-a-box design(which requires near perfect alignment)

refer to video...or search images for 'traveling column floor type'

https://www.youtube.com/watch?v=sY8TTAETQWU

[routalot]

Could we have a little more detail on the spindle to ram connection?

[dimmaz88]

Hi guys, a moving post type design was one of my initial ideas. I've always liked the idea of using UHPC to cast router parts, and thought this would be a great use for it.

I agree that my "4 post" design could have issues with alignment, and a single post would negate this.

Attachment 496158

Attachment 496160

Here is the first version of my 2-axis head, each axis rotates on a 115x75x20mm deep groove bearing. I wasn't sure if I should double up the bearings, obviously it would ad rigidity but is it needed for this application? I'd try and keep the weight down as much as possible.

Attachment 496162

Attachment 496164

Thanks for the input so far

Scott

[peteeng]

Hi Dim - Your truss column will be much stiffer and vibrate less than this skinny tall design. Your cantilever to the head is way too small as well. The BC head will work but you may want to support it on both sides vs only one. Peter

Here's a clay mill just think about what you need to cut harder materials!

https://youtu.be/TqokK03BJsM

[dimmaz88]

That is a big machine, I imagine they're built so sturdy for the speeds they achieve. It always amazes me how fast they can make them move, but I suppose it just comes down to bigger motors and more power.

"Your truss column will be much stiffer and vibrate less than this skinny tall design". That's the conclusion I came to and why I discounted it, I think I'd have to make the column that big it would be beyond the realms of a self build.

I'm happy that I can machine the "ladders" flat with my machine, the alignment will be the tricky bit.

By cantilever do you mean the post it's mounted on?

[peteeng]

Hi Dim - Personally I'd be making it more like a horizontal mill. See attached. The rotary would be on an axis and move in and out towards the column. This will be significantly stiffer than your current designs. Footprint no different. The image has a trunnion but you would not use a trunnion. Peter

edit - Here are the typical axis definitions in a horizontal mill. Your Z axis is a thin long cantilever. It will wobble. Plus you want to put a BC head on it, these are difficult to make stiff. They are usually very large chunks of metal... The horizontal mill however is a pyramid of stiffness and would serve your purpose well. Peter

[dimmaz88]

I could always make the "ram" out of 100x100 steel instead of aluminium, or move up to 200x200mm aluminium. I'm not too worried about stiffness as it's only for milling polystyrene.

The BC head is a must for doing more complex geometry, I've seen a couple of people build them for this purpose.

I must admit, the more I think about it the more I like the moving column idea. If it has a big enough base and gussets it should be stable enough. I'll draw something up..

[ardenum2]

your proportions are off, best use a square cross section for the column, the proportions on the clay machine are good, just copy those. or something like fleximill column from modig

for a 2.5m foam models, a 400mm x 400mm cross section column and 200mm x 200mm cross section ram as a minimum, or even better to make this simple the biggest square cross section pipe you can get and scale the column to it.

[dimmaz88]

I'm limited to 220mm of cutting height on my 3-axis router, which I'll be using to flatten the ways for the linear rails.

Attachment 496188

Attachment 496190

I think if I use concrete this type of design would be solid enough and dampen vibrations.

The rails are spaced 1000mm apart, hopefully resulting in a sturdy column.