CNC Mill Project Underway

Hello all, this is my first post on the forum but I've read a lot of threads here.

The machine I am building has a work area of 300mm x 300mm x 200mm and is designed to cut aluminum with ease and steel with care. It is made from laser cut 3/4" 6061 aluminum plate, and then drilled with my chopped down drill press and tapped for rails and other parts.. 180 watt belt driven (3:1 or 2:1, I forgot) servos on all axes, and I'm using a 9 axis planet CNC board to drive everything..

The spindle is a BT30 assembly, that will be belt driven with a timing belt by a 2500rpm 1.5kw servo. I do not yet know what speed I want to gear it for. I'm currently thinking somewhere around 4500 rpm top speed but I don't know for sure. I don't really want to deal with the super high speed low torque spindles many homemade machines have to deal with, and I do want to tap with this machine so I need to balance the top speed with torque. Any suggestions on the drive ratio and what belt series I should use for this are welcome.

I am building this on a $3000 budget, and the machine should be only 24" wide. Because of budget limitations I'm not purchasing any custom machined parts, and just getting the big aluminum pieces from OSHcut and using the laser and brake at work to make smaller sheet metal components such as the enclosure and motor mount. I am also using SBR20 rails instead of profile rails partly because of cost, but mostly because the mounting height matches exactly to the 20mm ballscrews. That particular linear bearings do have adjustment screws to clamp the bearing in the housing to eliminate backlash, so I think adjusting that properly will make a big difference. Also, the bearings are always very close to the tool end regardless of where the table is in its travel, so the forces seen by each bearing will always be less than the cutting force.

I have chosen to make a fixed gantry machine with a rising table, because in the case of power loss the table will gently settle to the bottom, as opposed to the spindle falling into the work. The Z and Y axes are joined by a precision ground cast iron block intended for measurement fixtures, which are readily available on ebay.

Due to Fedex problems I am still missing many plates, so the mockup is very incomplete.

https://i.imgur.com/VqdkCYa.pnghttps://i.imgur.com/8z4oM7A.pnghttps://i.imgur.com/kl6CzMi.png

Mockup picture links: (high res, don't display well on the forum)
https://i.imgur.com/q0VFsde.jpeg
https://i.imgur.com/Mz5tQj3.jpeg
https://i.imgur.com/njkGt2j.jpeg
https://i.imgur.com/e1OObwx.jpeg
https://i.imgur.com/yHH2noW.jpeg

Target weight is around 350 pounds.

The machine base is designed to be a server rack. All electronics will sit there, divided into computer, CNC control, power, pneumatics, and servo modules. The goal is to have reusable designs for these basic building blocks that can be used in any variety of projects that also include 19" rack space, as well as being able to swap individual assemblies and add functions without major rework.

Due to budget and size constraints there is no ATC functionality, nor is there space to add it later. I might be able to replace the drawbar with a pneumatic puller for faster manual tool changes, but that requires a little but of rework of the lid.

It looks like a very compact machine.It also seems likely to have a lot more gantry stiffness on one direction than the other.Have you considered the possible deflection parallel to the T-slots?

Quote:

---

Originally Posted by routalot

It looks like a very compact machine.It also seems likely to have a lot more gantry stiffness on one direction than the other.Have you considered the possible deflection parallel to the T-slots?

---

Yes I have, which is why the plate is doubled up at the columns.

The machine as shown has the table all of the way down. during typical use you would typically try to have it as high as possible, which shortens the columns and reduces the cantilever deflection. It is fairly easy to bolt on more stiffeners but I don't think it will need it. The different axis stiffness is an inherent limitation with this design of machine when very thick plate isn't an option.

I agree with what Routalot is gently suggesting. The geometry of the upright members is not going to give you much stiffness front-to-back. With some big triangular braces, it would be a lot more rigid, perhaps even rigid enough to cut steel to some extent.

Quote:

---

Originally Posted by awerby

I agree with what Routalot is gently suggesting. The geometry of the upright members is not going to give you much stiffness front-to-back. With some big triangular braces, it would be a lot more rigid, perhaps even rigid enough to cut steel to some extent.

---

I have decided to route my cables differently so that I can machine a beefier reinforcing plate and swap it out later. There is a limit to what I can do with just that one area, as after 2.75" stackup there (currently 1.5") I worry I will run into beam and bed twist being the next limit, and I am a bit too deep in building this to accommodate reinforcement in those areas. Though I'm sure if I did that the rails would be inadequate and so on and so forth. I am being very strict on this project to avoid feature creep, which overcomplicates and ultimately kills most of my projects.

I calculate 0.004" deflection, ignoring beam twist, at 100 pounds of force with the axis all of the way down (worst case). While this isn't great I think it's on part with other capable CNCs. Machining loads will obviously be the lowest, but I admit that resonance is unaccounted for here, and I am sure that 10x displacement amplification is possible with 6061, so chatter is for sure possible.

The other axis has 1/14 of the deflection, and likely less as racking isn't the same as canitlever.

After running the numbers I may review the cost and difficulty of getting a right angle gusset cut.

You don't necessarily have to add huge chunks of metal to brace the machine.in fact,I'm constantly surprised by how few people use the kind of tubular bracing commonly found on aircraft engine mounts or racecar engine braces.A bucket bush each end and a piece of tube goes a long way.


Attachment 472002

Quote:

---

Originally Posted by routalot

You don't necessarily have to add huge chunks of metal to brace the machine.in fact,I'm constantly surprised by how few people use the kind of tubular bracing commonly found on aircraft engine mounts or racecar engine braces.A bucket bush each end and a piece of tube goes a long way.


https://www.cnczone.com/forums/attac...d=472002&stc=1

---

That is a good point, but my machine has no base to reinforce too. It's held from the sides by the 3mm aluminum sheet metal enclosure.

Having established that you are using the enclosure to provide some triangulation,might I suggest you give an equal amount of attention to the mounting plates at either end of the pneumatic/hydraulic rams?As it is,it looks lie a perfect way to bend some fairly slender pieces of quite thin angle section as it will probably distort the metal more effectively than it will lift substantial chunks of metal.

Quote:

---

Originally Posted by routalot

Having established that you are using the enclosure to provide some triangulation,might I suggest you give an equal amount of attention to the mounting plates at either end of the pneumatic/hydraulic rams?As it is,it looks lie a perfect way to bend some fairly slender pieces of quite thin angle section as it will probably distort the metal more effectively than it will lift substantial chunks of metal.

---

Doing some figuring I'm going to up the sides of the enclosure wall to 1/4" aluminum, which will add significant rigidity to the weak axis of the machine.

Those cylinders only lift 50 pounds each. The angle is formed 1/4" aluminum plate and should withstand that without issue. A chunk of aluminum filling the entire work envelope of this machine weighs only 110 pounds, and that would be impossible to machine.