Re: crane style cnc machine
An interesting idea. I'll be watching your build thread. :)
Since X is a rotary axis, I think it's going to require a translator program to use standard cartesian G code. I don't know of any controllers that would do this natively, but such a thing may exist. To create a straight line, you would need to calculate the X angle and Y distance for a series of points along the desired path. The inverse of converting arcs to lines in a normal system. Not difficult, just simple trig, but would require a program to make it practical. The translator could output standard G code that any controller could read, but the file would be considerably longer than the original G code file. Overall very doable.
Mechanically to minimize the X backlash, I think you will want to drive the X axis from the outside end of the arm. Maybe a 1/4 circle rack attached to the building structure and the pinion on the motor attached to the outboard end of the arm. Something like that.
Re: crane style cnc machine
Hi Dingo - You will have to find a clever maths person who can do a jacobian transfer of the co-ordinates so a std 3 axis system output can be translated into the cylindrical co-ords of your system. Maybe a lathe output will work. But I think you will find that over such a big radial distance the accuracy required in the X rotation to achieve normal tolerances at the Y end will be quite tricky and the crane beam will be quite wobbly. Both of these factors will be hard to resolve mechanically and mathematically. Plus your rotational momentum of the Y axis will be enormous requiring a very large motor to get it going and to stop it. You will need to look at these numbers very closely to avoid disappointment.
There are machines called SCARA look them up they may be helpful and motion controllers exist for these. Your concept has many challenges and maybe building a small scale one first is the 2020 project....Peter
Re: crane style cnc machine
You're going to have rigidity problems. I would consider adding a linear actuator of some sort to the far end of the crane arm.
As for a control, you will need to do some programming to make this work. Therefore, I would be looking at an open source control platform - either Grbl (my personal choice due to familiarity with the code base) or LinuxCNC. If you choose Grbl, you will be limited in step rate if you use one of the AVR based versions; due to the requirement to do a transform for every point along a move -- so I would consider using an ARM based port, or perhaps the ESP32 based port.
https://github.com/terjeio/grblHAL
https://github.com/bdring/Grbl_Esp32
All in all, this is a non-trivial problem to solve. It's not impossible, it's just that you're going to have to dive deep into the control software to make it work.
If you want to save shop space, and stick with a more conventional machine; how about just making the rails & transport assembly tilt up for storage in a vertical position?
Re: crane style cnc machine
Quote:
Originally Posted by
dingo117
Hello all, i am thinking of a new project to work on in 2020 the picture shows my rough idea, i have made some cnc machines before including a standard medium size gantry style router so i have gained a lot of experience in that style and the simple cartesian coordinate system of motion and location.
My main lack of understanding comes from the non cartesian movement of the crane style i have quickly plotted in the picture. i need to figure out what control setup or work flow will let me deal with this rotary X and linear Y and Z axis so ican operate in a normal 3D cartesian coordinate system. my understanding is that i will use something like a typical self contained stepper motor motion generator which seems to not do any multi axis calculation and edit the outputs compared to the inputs it gets from jogging commands or a nc program generated from fusion 360 say which has your standard linear 3 axis set up. so what are my options if any one knows?
A little bit more about this idea of mine, the arm is 6m long, i intend to at first use it for plasma cutting of large metal sheets i do not intend for it to be a super rigid setup for milling metals but i would like enough rigidity to stop it deflecting under its own weight or from unloaded moderate speed motion, say giving it a -+0.25mm accuracy? preferably +- 0.05mm, i know plasma cutting doesnt require precise accuracy since plasma cuts are pretty varied to begin with.
I would also like to add an a and b axis head to it with a small high speed spindle for cutting wood, plastic and foam sculptures, i will tackle the simple plasma operations first but i would like to know that this sort of setup could be rigid enough to at least do the sculpting of light materials.
The reason i want the crane design is it takes up next to no room on the shed floor, the shed is about 14x7m i can put the main pivot in a corner and use the space under it and just pull in a water table and start plasma cutting sheets, as for anchoring down chunks of wood or foam for 3d carving im guessing a heavy metal plate on lockable wheels with fixtures holding on to the work piece will have enough weight to stay put, if not i will think about drilling some holes in to the concrete to make anchors for the base.
If i make a standard gantry style one with the same work area as this crane it will still easily fit in the shed but i will have the rails and rack and pinion that the gantry runs on sitting in the middle of the shed being an annoying obstacle when the machine is not in use
If you have a floor area problem you can build mount a normal 3 axes machine on the wall or free standing against the wall
Re: crane style cnc machine
Quote:
Originally Posted by
mactec54
If you have a floor area problem you can build mount a normal 3 axes machine on the wall or free standing against the wall
on a vertical bed style machine wont you be fighting gravity then with all your cut out parts falling out and possibly getting jammed up on on the machinery? and i would not be able to use a water bed then for plasma cutting but it is a good for a pure woodworking machine for sure
Re: crane style cnc machine
Hi Dingo - I think you need to rethink the config. To move radially 0.25mm at 6m radius requires a move of 0.00119deg. If you used a std 200step stepper this would require a gearbox with a 1513:1 ratio to achieve that move. The calculations will have to consider the r value and adjust for this delta move at every radius difference. I think it's way too complex plus the structures will need to be enormous to be stiff enough. Keep it simple. Peter
Re: crane style cnc machine
Quote:
Originally Posted by
peteeng
Hi Dingo - I think you need to rethink the config. To move radially 0.25mm at 6m radius requires a move of 0.00119deg. If you used a std 200step stepper this would require a gearbox with a 1513:1 ratio to achieve that move. The calculations will have to consider the r value and adjust for this delta move at every radius difference. I think it's way too complex plus the structures will need to be enormous to be stiff enough. Keep it simple. Peter
i was thinking of using a ball screw like a piston to move the arm but i have already decided to not do a crane and started planing for a standard gantry style machine, just refining the dimensions a bit and what i actually want to do with it. i will be posting soon for some advice!
Re: crane style cnc machine
Quote:
Originally Posted by
dingo117
on a vertical bed style machine wont you be fighting gravity then with all your cut out parts falling out and possibly getting jammed up on on the machinery? and i would not be able to use a water bed then for plasma cutting but it is a good for a pure woodworking machine for sure
Yes that would be a problem for plasma I did not read enough of his post to see Plasma
Works well for a wood / plastic router and for laser engraving
3 Attachment(s)
Re: crane style cnc machine
So i think the direction all the suggestions on here are pointing toward my idea for a crane style machine to be more trouble than its worth for the space saving gains! key issues being rigidity problems on such a long unsupported boom? and having to code my own or edit an open source controller or make some inline pulse editing thing.
The momentum issue of a large boom could be overcome by a big motor as some one stated which i dont mind investing in but the other issues are making me just want to go for a normal style machine now. so given all this i have been tinkering with a normal gantry style design, picture attached. so i still would really appreciate some opinions on how to approach something this size.
first thing first rigidity and i intend to do with it. i was not going to bother with a very rigid setup with the crane design as i knew that would be a waste of time, it was wood/plastic/foam etc light material cutting and plasma cutting, but since i have decided to go gantry i know i can make that a lot more rigid easily enough and may be able to cut metal with it now, so any if my design below would be roughly rigid enough to cut metal... slowly i guess? the second big issue is the depth of the z axis, its got a long snoot i may have to lower that to get the rigidity i require for metal i guess but that depth would be amazing to still do big foam sculptures with! so keeping it would be nice if possible. so with that length of z axis hanging all the way down the torsion it will apply to the gantry would be pretty high so any idea if it could at least cut wood wit ha decent depth of cut at that z size? and what height do you think would be good for metal cutting instead?
as for the vertical supports that hold the gantry up i have yet to figure out what i need for that but i am guessing a similar size structure as the horizontal part.
Re: crane style cnc machine
Hi Dingo- You are now talking about making a mill. That's a really big step up in rigidity over timber and plastic. You say "metal" is steel in those thought bubbles or only aluminium?
aluminium is maybe 5x stiffer then timber then steel is 3x stiffer again. So a steel machine has to be 15x stiffer than a timber machine (V roughly speaking). Then it needs more power to cut much much more power. etc etc
1) The 2mm section although it has a V large inertia is too thin to prevent the section changing shape under metal cutting loads (mainly torsion even for timber I expect). Typical steel gantry sections are 250x250x12 (because of your Z will need to be bigger). There is no gain in having the smaller section attached to the bigger section, you may as well put all the metal into one efficient section. I realise you need room for a drive screw you will need to figure that out but a solid is usually used not a hollow for this purpose. A 3m gantry is going to be a very big and heavy chunk of steel, aluminium will be better to save weight but it will need to be bigger as well either billetted and bolted together or welded sections
2) Your huge Z will be the crux of the design it will need to be steel 250x350x12 or bigger to be stiff enough. Have a look at blue elephant site and other builders to get an idea of their sizes for this sort of thing.
3) If you look up the Maximus thread it details designing an aluminium capable router from the start
4) I'd say your current design is under dimensioned badly for aluminium mainly due to the Z axis height
5) You have your long rails on the sides. Unless you have these surfaces milled its very difficult to get them parallel
6) If you are going to weld these long structures they will warp so you will need access to a heat treater for stress relief.
All of these things are in the many threads of this forum in great detail...
Whether you cut slow or fast does not matter if the machine is not rigid enough it won't cut. Mills weigh several tonnes in this size machine for good reason to get the rigidity...
Keep at it.... Peter
Re: crane style cnc machine
i have been going over your maximus thread, looks like a nicely designed kit, i am leaning towards not bothering with a steel capable mill as i dont want to spent the amount of money 12mm thick sheets of steel for the structure will incur. i think i will bump up the gantry box section thickness so it wont twist as easy and shorten all the axis down, hopefully this will make it aluminum capable, in addition the long z axis i may shrink it to 1m but if i use a table to lift the metal stock higher so the z axis wont be exerting such a long leverage that may twist the gantry
Re: crane style cnc machine
Is it worth thinking about the table as dual-purpose ? ie, when not in CNC use can it double as a workbench ? Or a composites layup table ? Maybe somewhere to prepare drawings or do marking out ?
It would require some removeable covers for the linear rails, but that's not difficult.
1 Attachment(s)
Re: crane style cnc machine
Hello Dingo117
If I understand you correctly, you will need to convert Cartesian coordinates to Polar coordinates. This is very simple to do.
This link will help you to get up to speed.Polar Coordinates -- from Wolfram MathWorld I am not sure if this can be done in LinuxCNC outside of re-compiling with changes to the code.
https://www.cnczone.com/forums/attac...d=433688&stc=1
Hope this helps,
Iron-Man
1 Attachment(s)
Re: crane style cnc machine
Hi Iron Man - converting cartesian to polar co-ords is easy its the extra calculation to compensate for the radius arc change in distance that's the hurdle. eg you are using a motor to control the rotation of the crane that works in steps. So one degree movement close to the centre is different step count to one degree at 5m from the centre. This extra calculation layer is not trivial in terms of motion control and trajectory management.
Hi Zorbit - Haven't heard from you in a while, Merry Christmas and soon HNY!! A client has asked me for exactly that combo, so I'm working on adapting maximus to this config, see image attached. Then he can use it as a packing table and a router. Peter
Re: crane style cnc machine
Neat design Peter, good for a small workshop.
Re: crane style cnc machine
So i have got my hands on the x,y cross slide style portion of a manual mill and since i have never had a go at cnc milling stuff like steel before or even aluminum on something very rigid (i had milled aluminum on my current cnc router with shallow depths of cut and gotten alright results) i will try convert the cross slides to be motorized with c7 ball screws and steppers and make a decent column and head for its Z axis. Then try gain some experience with steel milling if it all works out well then i will use that to cut out components for a bigger router or gantry style mill in the future...what ever i finally decide to go with in the end. hopefully i can get some milling work to help cover the costs as well but i am mainly in it for the learning experience!
Re: crane style cnc machine
Iron Man, that sure seems to be the cast with the coordinates, i wont be using a cranes style any more i am pretty sure, i have been looking in to linux cnc and seen that it can do some pretty unique kinematic models, and lots of other neat stuff as well, i think i will try use linux cnc and a mesa card for what ever design i end up choosing in the end purely for the flexibility and performance it offers... and seems to be a low cost control solution as well, best of both worlds.
