Help Design A CNC Unit

[Joe Wood]

Folks, could you help me design a CNC setup for cutting these various curved rafters only? The CNC shop I usually take these to has been so busy the last few years that it sometimes takes weeks to get them back.

The widest pieces I'm cutting is 2x12, the longest would be 12' so I’m thinking a 2’ wide x 12’ long table. Could the router travel up and down a single strongback above the table? I only work with clear wr cedar and Calif. redwood ~ 2” thick so don’t need a powerful motor. I'm hoping to just set the new CNC table on one of my shop benches when I need to use it.

How do I get started with this design? Are there any good resources I could check out? I'm in the San Diego area if that matters.

Similar Threads:

• CNC Machine Shop American Southest. We Design, CNC Programming, Tool Design, Engring
• Mobile Manufacturing Unit #1: LaserPod by Stone Design. Engraving on the streets.
• Configuring steps/unit.....what unit?
• Need Help!- FANUC 15-MA Battery unit & I/O unit etc
• design grad project: new breed of CNC design, need input

[joeavaerage]

Hi,
you will need a machine that is stiff enough to contain cutting forces without deflection or it will screw the accuracy and finish of your parts.

You may rather imagine that you can use a low power spindle, and you may be right, but I think you'll want at least 2kW. At 2kW you should be able to use single
phase power, with a decent high current outlet. I say this because you'll need a tool about 1/2 inch in diameter otherwise it will deflect, remembering that it will have
to cut two inches deep. That tool will have considerable cutting forces and therefore a stiff or rigid machine will be required.

I would suggest that you keep the gantry fairly short say two- three feet. It so much easier to make something stiff if its small. That would mean that you would
lay the timber long ways. The gantry would then traverse the length of the timber. Rotating ballscrews over that length is not practical, they would have to be large and therefore
expensive. I would suggest that you need rack and pinion, at least for the long axis. The gantry and Z axis could be much shorter and ballscrews would be a good solution there.

What sort of budget do you have in mind? That will determine what you can door indeed if it can be done at all for a given sum.

Craig

[peteeng]

Hi Joe - I suspect that what you need is not what is in your head. To cut 2" deep will require a stiff machine ie big and heavy not being able to be moved easily. Some 2kW plus and ER20 (1/2" tool) it won't be something you can move around and will require a dedicated space. I think it will be a mill type configuration like attached to allow front loading. This is called a moving column design. It will use rack and pinion on the long axis and ballscrews on others. If the arc is always the same radius then a copier will be easier for you vs CNC make a template and your good to go manually. But keep asking questions its not a short journey your on unless you buy an off the shelf machine. Regards Peter

Here's an industrial version of what you need. Biesse call it a pod and rail configuration.

[joeavaerage]

Hi,
Oh yes, wouldn't a front loading machine be nice!

The only fly in the ointment with that design is the required torsional rigidity of the 'backbone'. If you were going to go that way the I would have to suggest something
like a 200mm x 200mm x 9. Such big steel sizes are stocked by companies doing steel fabrication for buildings, and you may find a 12 or 13 foot length.
It will weigh enough that the machine cannot be considered portable.

Craig

[peteeng]

Hi Joe - Since your timber workers and the job is cutting timber which does not require accuracy to 0.01mm I'd suggest you build a timber machine. Big solid laminated parts are strong, stiff and damp. You will need to get an understanding of a few things so keep asking Q's. There are lots of plywood CNCs out there, yours would be a bit different due to its length but all doable. One thing is who will do the CAD work? if you don't have someone familiar with this then the design and the eventual operation will be an issue...Peter

[Joe Wood]

Yeah so I'm wanting to build something simple for a one purpose task, up back and around in one axis, farthest travel maybe 18".

the CNC shop uses a 3/4 roughing bit in a few passes and gets a very clean job for the most part. He was often screwing the pieces down at a point or two and we'd just fill the holes so I could get by without needing any vacuum hold down, just a thought.
I can't get that vg cedar anymore so use the redwood which is very stable stuff to cut.

I could see two parallel metal gantries 18" apart, and building some of the parts with laminated timber!

What sort of spindle and moving mechanism am I looking to get, for these simple up and back narrow pathways?

my idea seems very doable we'll come up with a nice design!

[joeavaerage]

Hi,
I had an idea or two, pictured very roughly below. The long central member is 200 x200 x 9 RHS steel. I imagine all the bits could bolt/screw to it rather than any welding.
Welding is a sure fire way to make things bend and twist. I would suggest to get the stiffness that a 9mm or 12.5mm wall thickness be sought. In addition to the torsional stiffness
you can drill and tap a hole and be secure that there is enough thread to really hang in there. This idea just has the one main member, which may save on cost, but also puts a
premium on it torsional stiffness. It also requires only one gear rack and gear rack drive motor, maybe a cost saving that counters the cost of one solid RHS.

You might find this sort of material is part lengths from a architectural steel fabricator.

The Yaxis, that slides back and forth on top could likewise be made of this material.

This is a sort of peteeng's 'front loading' machine.

Given that you are a timber worker then I can well imagine that building in wood and/or plywood would appeal. I'm a steel man through and through, and I tend to think in steel.
Christ, I think I dream in steel and probably s****t steel too!

I would suggest a 2.2kW water cooled spindle with an ER20 toolholder. It would require a VFD to drive it. The Chinese seem to specailise in making these things and you'd get
a complete kit for under $1000. Pretty good value for money.

Craig

[peteeng]

Hi Joe - The machine will be 3 axis. Longitudinal (side to side) transverse (back to front) and z up and down. If you build a machine base you can clamp the front of the job, cut the back then clamp the back to cut the front, this would be two programs. The base would not be solid it would be slats so you can place clamps nearly anywhere. I would stick to timber everywhere. Mixed media is sometimes hard. A big steel bench is hard to keep straight if you want to weld it, but if you bolted it together, aligned it then tacked it that would be good. A timber bench can be planed flat. If you search for an ER20 spindle 24000rpm. Something like this. but you need to check what power you have available in your shop. Plus I recommend not to get excited and buy anything until you have designed it. Definitely identify where to get things and what they are but be patient as things can change thru the process and you end up with expensive surplus stuff on the shelf....

https://www.aliexpress.us/item/32568...andl_shipto=US

You do need to figure out who is going to do the CAD CAM as this is key to the success of the project. No point having a great machine sitting there and no one to run it (does happen) Peter

an er25 spindle goes from 1mm to 16mm this maybe good for you. ER20 are 1mm to 13mm...

[joeavaerage]

Hi,

but be patient as things can change thru the process and you end up with expensive surplus stuff on the shelf....

I can double and triple vouch for that, you need a rock solid idea BEFORE you hit the <Buy Now> button!!!

Craig

[peteeng]

Hi Joe - I think you also need to find your tool supplier. You need long and thick tools. These are generally not at the hardware store. The tool supplier will also be able to give you an idea of the power required to spin those tools. There are things to figure out like ideal feed speeds for the tools BEFORE you buy anything. You need feed speeds, spindle power to figure out drive trains and gear ratios etc so you end up with a machine that does what it's supposed to do. Plus have some overhead in it for future growth. So its a paper exercise until all the unknowns become known. I expect you'll be making sawdust in about 6-8months so be patient... Peter

[peteeng]

Hi Joe - I can tell your excited so here's my sticking point list:
1) Getting rails form Asia will be tricky as they are long. So you may have to get these local. So start looking for a local supplier of round or square rail. You can put in an enquiry to asia but they have trouble with fedex and similar freighters with over 2.5m rails. Your going to need 4.5m I think. I don't recommend joining rails. 4.5m is a std length here on OZ from local suppliers. They are 2x the price from asia but they fill containers to get them here in one piece...
2) Decide on Er20 or Er25 talk to your tool suppliers and maybe the CNC company you deal with for some pointers. Then decide on kW of the spindle
3) Square rails at 20mm would do the job but lets go 25mm
4) You need to get familiar with a machine control system, do you have an electronics person in your circle? Electronics for this is a big subject and will always be a hot subject. There's plenty of opinions in the forum
5) I'd also consider belts for the long axis vs rack and pinion but you shall have to find a source for both to decide. Connecting R&P is easy, a 32mm belt or bigger would do this job nicely
6) A non expanding urethane adhesive for laminating like titebond is good. If your happy with epoxy probably a bit better but hard to tell in my experience
7) Think about everything as a solid sections no need to make hollows as timber is light. laminate with opposing grains to prevent cupping
8) Use cold forged furniture inserts for threads in timber that need to be removed (not the zinc cast ones). For permanent threads use std bolts, use undersize pilot and then thread with tap. Screw in use PVA glue or titebond will never come out unless a spanner is put on it.

I'm sure there's more but you will need to sketch something on paper then into CAD Good luck on the mission. Peter