critique my 2200 x 500 router design?

I'm a newcomer, currently designing my first CNC router for the purpose of ski-building. This requires a fairly long work area (2200mm travel along X direction). The tool would find use in cutting and shaping the wooden cores of skis, as well as long MDF pieces used to form a mold for shaping the ski during vacuum pressing.

I'm limited by my lack of machining skills and tools, so I've chosen to use aluminium extrusion (80x80 for the frame, 40x160 for the gantry and 30x150 for the remainder). My design includes a handful of connecting aluminium plates which I will probably try to have milled for me. For the movements I have selected the following:

X direction (2200mm travel, 40kg driving weight): 20mm profile rails, 4 bearing blocks, racks with module 1 pitch and 30-35mm pinions

Y direction (550mm travel, 14kg driving weight): 15mm profile rails, 4 bearing blocks, 20mm ballscrew

Z direction (150mm travel, 8kg driving weight): 15mm profile rails, 2 bearing blocks, 12mm ballscrew

I selected flanged bearing blocks which are able to be mounted directly to the t-slots in the extrusion.

For the X direction, I chose a rack-and-pinion drive after learning about the problems ballscrews have with buckling and whip as the length grows. I selected 1:5 planetary gearboxes for the pinion drives, with backlash specified as 15 arcmin or less. I managed to design a tensioning mechanism for the motors using M4 threaded rod and rod-end bearings for the pivot. (I'm not sure how it would work in practice but I can actually build it). With this arrangement, a single step of the motor would correspond to around 0.1mm, and the backlash from the gearbox around 0.07mm (on top of backlash from the rack itself).

The design uses a gantry undercarriage, so the frame is unsupported in the X direction except at the ends, which I anticipate could be a problem. Using a beam deflection calculator, I think it could easily sag 1mm or more in the middle. I'm also not able to incorporate any cross beams in the interior of the frame, although the cutting table (probably 16mm MDF) would serve in this capacity.

I've not properly sized the motors yet, but I think I'm looking at around 1.5 Nm holding torque for each X and Y motor, and 0.9 Nm for Z.

I made the model using Sketchup, and uploaded it for viewing here with a web-based 3D viewer:

https://3dwarehouse.sketchup.com/mod...9-73010da7dea6

I would greatly appreciate commentary and criticism on literally any aspect of this design! I educated myself as best I could, but feel that some aspects would really benefit from working knowledge which I don't have.

Don't try to reinvent the wheel.
Move the steppers to the outside, and fully support the frame.

I only saw one image? DO you have more pics?
Also, you'll want larger motors than 1.5Nm.

Quote:

---

Originally Posted by ger21

Don't try to reinvent the wheel. Move the steppers to the outside, and fully support the frame.

---

Excellent, thanks! A reality check. I had based my design on raptor, which uses an undercarriage. Now I've reworked the X axis to have the rack mounted on the outside of the frame. A better link to the new model (direct link to 3D viewer):

https://3dwarehouse.sketchup.com/emb...e-d471db23b088

I scrapped the tensioned motor mount too and it all seems a lot simpler.

Quote:

---

Also, you'll want larger motors than 1.5Nm.

---

I used a spreadsheet from another post on this forum to get a rough idea. I will look into motors more closely once the mechanical design is finalised, but hopefully NEMA 23 mounts will be adequate. There are two X-axis motors, so I suppose that helps. I don't yet have much intuition about the cutting forces at play.

You probably want to spring load the pinion into the rack. The slightest deviation in rack straightness will result in either backlash, or a snapped gearbox shaft.
You also want the teeth of the rack pointing down, or they'll fill up with dust and chips.

Get rid of the gearbox on the gantry. Use a 2010 ballscrew, direct drive.

EDIT:
Nevermind. I see that it's a motor mount, not a gearbox.

Quote:

---

Originally Posted by ger21

You also want the teeth of the rack pointing down, or they'll fill up with dust and chips.

---

I considered the rack orientation. I couldn't see how to mount both the rail and rack on the adjacent outside t-slots with the teeth facing down. (Unless I switch to a 12mm x 12mm rack, and used a smaller gear - say 20-25 teeth - which might be able to squeeze into the gap between rack and rail.)

Quote:

---

Originally Posted by ger21

You probably want to spring load the pinion into the rack. The slightest deviation in rack straightness will result in either backlash, or a snapped gearbox shaft.

---

I also thought a lot about spring tensioning of the motors. (I'm unsure of the significance of the issue - opinions on the forum seem to be divided!) I was able to come up with an arrangement with the motor mounted vertically of the gantry side and the rack teeth facing outwards, using a hinge to provide the pivot. (Yes, I am mechanically challenged and am hard-pressed to build much of a mechanism...) But no such luck for a horizontal arrangement.

My solution to both these problems was to mount the motors underneath the table on a gantry undercarriage, as per the model in my first post. But this brings about other issues (lack of support along the frame.)

(Thanks to anyone who takes time to look at my design and offer an opinion.)

With upside-down rails, a tight fit with a 20T gear; a bit risky I think?

Attachment 297542

Can't you just reverse the rails and rack? I'm definitely in the spring loaded camp too and like the cnc router parts design. I built the ones for my present cnc but next time I'll just buy them from cnc router parts and be money and time ahead. Hoping to have the design work done on my next cnc soon so I can do just what you are doing and get feedback from those who know a lot more than I do about this madness that is cnc.

BTW.... I'm in Steamboat Springs so I'll be looking forward to seeing some of your product. Opening day is fast approaching and we finally started getting snow so I'm looking forward to getting back on the mountain again.

Put the rack on the bottom, and the rail on top. Then you have all kinds of room below the rack.

Rack on the bottom is the answer, of course. I've come up with a solution for both fixed and pivoted motor mount. I will try fixed first; rack and rail are close so I can use a spacer between them when mounting to keep them as parallel. Learn by doing I suppose! I can swap in a tensioned mount if it's needed. I'm otherwise fairly happy with the design. Thanks all.

I'm planning on building a 3000 x 3000 x 2000 cnc.

I am going with ball and screw for accuracy at the best price.
You can support the screw with sliding mounts that travel along the screw along with the ball which enable you to eliminate whip.
Just add another track and slider(small sizes will do) and have your screw supports move up and down this track,
Cost to make the sliding mounts is an additional ~200$.
You lose the width of the slider and bearing support to your overall travel along the screw but your design(2m) would only need a single support each side of the ball screw.

When your ball nut is it at the end of the screw. The support at the other end will be pulled into the middle of the screw by a spring attached to the slider. The support at the other side of your ball is pushed out the way against its spring. When the ball moves back the other way down the screw the spring pulls the slider(the one pushed up against the stops) and support along with it until it reaches the middle of the screw where it will now come into contact with the other slider and support at the other end of your machine which is now pushed out the way while the following support reaches its centre stop and remains there.

I plan to use beefy springs to move the support sliders into position as the servo already has heaps of power needed to move the 150Kg gantry beam and motor.

I borrowed this idea(sliding screw supports) from looking at how many of the commercial high accuracy large travel linear actuators worked.

Here's a good article on sliding screw supports. They even have a solution without needing another track or sliders to support the screw.

https://www.assemblymag.com/articles...-linear-motion.

I am going for an extra track and sliders for my supports instead of this apparently simpler design because I, like you, do not have access to a milling machine or lathe.

My solution just needs the beam to be machined flat on one side after it leaves the mill. Which I would have needed machining flat for the screw anyway. Then I can fix my screw and bearing support system to that side of the beam without needing any more machining or custom devices. Just need an extra track and slider and bearing support fixed to that side of the beam. Just drilling and alignment.

You should also consider that those common cross section Aluminium extrusions only come in at half the weight of box section steel, and not the 2.5 or more you would expect, because of the internal webbing. I made the mistake of comparing the outside dimensions of aluminium with comparable box steel assuming the aluminium was also a box section.

Weight of aluminium, super expensive P40 extrusion and only 200 x 80 x 4, for my moving beam = 41kg
Same beam in steel at 200 x 100 x 6 = 94Kg

As a ski maker you should also know that the outer skin has the highest tension when it is bent. So if you have deflection on your beam it is because the outer surface stretched under tension.

Tensile strength of 6061 Aluminium = 290Mpa
Tensile strength of Steel Box sections = 400Mpa

No need to guess why I switched from Aluminium to steel. I get an extra 25% tensile strength and a wall thickness of 6mm instead of 4mm along with 100mm on the tops and bottoms of my beams of steel instead of the 80mm of aluminium resisting the torque in the direction of whatever motor I use.