Building a CNC router/mill for aluminum and light steel, looking for design critique.

[benhobby]

I designed a small machine to mill aluminum and steel, and I designed it to be as cheap as possible. T8 leadscrews, SBR12 fully supported linear rods, nema17 steppers. I know it sounds WAY underbuilt for steel, but hear me out. The T8s are using compression anti-backlash nuts, and are SUPER solid. The rails are decent quality, not too long, and fully supported in steel bars cast into concrete. Each of the x and y axes has two nema 17s geared down. If thats not enough power I have provisions in place for two more per axis. I have a 1.25Kw spindle motor (RC car brushless motor) geared for a max RPM of 15K. Both the base and the X axis are going to be cast out of fiber reinforced concrete. I have attached a rendering below. Missing from the rendering are the leadscrew mechanisms, and the spindle and motor. Please give advice, this is my first CNC but not my first machine of this complexity.

Similar Threads:

• Newbie- critique my 2200 x 500 router design?
• New Machine Build- Building stainless steel router
• New Machine Build- Steel / aluminum CNC router design
• Bolt Together CNC Router - Newbie looking for design critique
• New Machine Build- New steel mill/router design. Feedback appreciated.

[awerby]

What's the backlash from those gearmotors? The nuts will compensate for backlash between them and the screws, but not for backlash in the gear train.

[peteeng]

Hi Ben - Why are you using round bearings and not square rails and cars? They are stiffer? Peter

[benhobby]

I was able to get all of my linear components, 4x400mm and 2x200mm with 12 bearings for a total of $45, and that was the main driving factor. If I could afford them, I'd be using them. They are bolted in on every single available hole, so I'm doing about the best I can with what I can afford. Thanks for the advice, though, and a true linear rail upgrade is in the future if I use the machine often enough and see the need for it.

[benhobby]

I'm using closed loop GT2 belts for all axes (and even the spindle), and I've actually tested a backlash of 0.05-0.1 degrees. On 8mm pitch leadscrews, we're talking single digit microns for the gearing backlash. I think the cheap chinese anti backlash nuts are the worst link in the chain.

[peteeng]

Hi Ben - What is the width of your belts? and why use concrete? even fibre reinforced (steel or plastic fibres?) will crack over time and change shape? Cutting steel is no mean feet and this is not the usual way to do it, mills are $$$ for a reason. Peter

[benhobby]

Hi Ben - What is the width of your belts? and why use concrete? even fibre reinforced (steel or plastic fibres?) will crack over time and change shape? Cutting steel is no mean feet and this is not the usual way to do it, mills are $$$ for a reason. Peter

Peter, The belts are 6mm wide. They won’t be an issue due to the low torque of my motors. The reason for most of these choices is that the full all in budget, including the entire machine, software, and tooling, is $250. Sure, if I could afford to properly build a mill, I would, but I’m only 18. I mostly came here for design advice, not “let them eat cake”.

[awerby]

Have you cast this concrete frame yet? That seems to be the most critical feature of this design, and also the most unique. Concrete normally shrinks a bit as it sets, so casting steel bars into it may not work as intended - expect some corrosion and displacement. It will need to wet-cure in the form for quite a while to build up maximum strength. Is there a welded frame inside it? Without that, it would seem pretty fragile, even if well-cured. Usually people making monolithic frames use epoxy-granite, not regular concrete, to avoid the shrinkage and fragility issues.

Your choice of spindle will limit your ability to cut metals, since those motors have very little torque, especially at lower speeds. You will probably be able to do some light engraving on steel, and cut aluminum shallowly, but will have trouble doing what most of us would consider "milling" in those materials. But making any sort of CNC engraving machine on your budget is still a pretty good accomplishment; just dial back your expectations for the material it will be able to remove.

[peteeng]

Hi Ben - If you have not cast the concrete I suggest you make it out of MDF. Being the first machine you will want to change it and changing concrete is improbable. Plus concrete I think will let you down. MDF is cheap and you can do all sorts of things with it. You are limited with the 6mm belt and a good MDF frame will be stiffer then a 6mm belt. Once you are through the learning curve you can move onto other materials and morph your machine forward to your needs or build a new machine using the first machine. I'd reconsider the concrete, everything else would look the same. Peter

[xillianto]

He is not the only one who has been thinking about concrete, Mine is reinforced with 12mm rebars every 50mm, with 10mm steelplates inserted for rails.
Then levled the plates with epoxy before installing the rails.



Machine IS under build, wating for some parts on my X and Z axis.
Also changed from 2.2kw to 3kw ATC spindle, running DMM servo's, 2010 ballscrews and 20mm linear guideway rails.

[JoanTheSpark]

..Machine IS under build, wating for some parts on my X and Z axis..

Is there a build thread of your machine somewhere please? Thanks.

[peteeng]

Hi Xill - do you mean portland cement (add water = concrete) or epoxy concrete? If portland what about it cracking as it cures? Concrete cracks all over as it cures? So your just using it for mass? Peter

[xillianto]

Hi Xill - do you mean portland cement (add water = concrete) or epoxy concrete? If portland what about it cracking as it cures? Concrete cracks all over as it cures? So your just using it for mass? Peter

Hi, Have not had any problems with cracking for the almost 2 1/2 year that i have had the frame.
it stays dimentially stable aswell as i can measure.

there is very mutch rebars in it though
but remember to keep itnwet during the first 48h of curing..

Sent fra min SM-G975F via Tapatalk

[xillianto]

Is there a build thread of your machine somewhere please? Thanks.

No sorry, only some progress pic on facebook groups, but here is a fresh pic

Sent fra min SM-G975F via Tapatalk

[JoanTheSpark]

..but remember to keep it wet during the first 48h of curing.

I got a relative who tests concrete samples from bridges and similar for a living.
They cure them at 25 deg submerged in alkaline water for 28 days I think before they break them in a press to get the compressible strength value of the concrete.
The value by then usually is above what the concrete is supposed to be by specification (25MPa, 35MPa, 50MPa etc.).

PS: thanks for the image.

[peteeng]

Hi Ben - Did you get your machine built? Thinking of something similar - Peter

[asuratman]

You can make cnc router using H beam in your design easier. I made several cnc router and cnc plasma with all H-beam 15x30, they are very ridgid and heavy since all steel.