Critique my CNC Column Mini Mill

[henriksson]

Hi everybody,

I'm in the process of planning my first milling machine. I have previously built 3D printers, so I know the electronics, but I'm new to the heavy machinery. I only own hand tooling, so I've designed the mill to not need any special tooling, this is also the main reason why I didn't go for a steel construction. The overall budget is about 2000 euros.

The work envelope is about X=500mm, Y=200mm, Z=300mm. The intended use case is various hobby projects, in particular aluminium, brass and (slow) cuts in steel.

All ball screws are 1204 pitch screws, the rails are HGR20 with HGH20CA carriages.

The profiles are 90x180mm 10mm-slot Bosch-Rexroth type aluminium extrusions. The bed is 16mm and the rest of the plates are 12mm (all AW-5083 precision cast aluminium).

I have a chinese 2.2kW spindle in the CAD, but I'd ideally want something with lower RPM range (say 1000 - 10k rpm), so I that can run non-balanced tools like boring bars and fly cutters.

Is there anything obvious which I have missed from a rigidity point of view? Anything else I should be aware of?

Similar Threads:

• X2 mini mill column gathering dust?
• Build Thread- another mini mill column flex project
• Market Research: Bolt On X2 Mini Mill Column Support
• mini mill DOC and column question
• mini-mill z-axis/column rigidity

[awerby]

How are you attaching the column to the base? That's a critical joint there, and those aluminum extrusions don't give you a really good way to do it. I'd suggest a thick piece of angle iron on the outside corner to beef tt up, connected firmly to both sides.

[joeavaerage]

Hi,
aluminum extrusions are not adequate, they flex too much, and at the right angle junction, its really going to let you down.

Your description and desire for a low-med speed spindle suggest that you want to cut metals, in which case this design is just too flexible.

I would suggest that you get some real heavy duty universal column (steel) and cut the flanges to height, and either weld or bolt the column to it,
and the column would need to be steel also. This is the sort of material I mean:

https://www.trademe.co.nz/a/marketpl...B&gclsrc=aw.ds

Note that the flanges are 18mm thick and therefore thick enough for a drilled and tapped holes to mount the linear rails.

Craig

[strostkovy]

Hi,
aluminum extrusions are not adequate, they flex too much, and at the right angle junction, its really going to let you down.

Your description and desire for a low-med speed spindle suggest that you want to cut metals, in which case this design is just too flexible.

I would suggest that you get some real heavy duty universal column (steel) and cut the flanges to height, and either weld or bolt the column to it,
and the column would need to be steel also. This is the sort of material I mean:

https://www.trademe.co.nz/a/marketpl...B&gclsrc=aw.ds

Note that the flanges are 18mm thick and therefore thick enough for a drilled and tapped holes to mount the linear rails.

Craig

I think this machine would be capable of gently routing out 1/8"-1/4" aluminum to make side plates to reinforce the joint. It still won't be amazing, but should be adequate for light work with patience. I assume the design currently calls for long through bolts into tapped holes in the end of the extrusion.

I would advise against using an I beam unless you can machine it, as they are notoriously warped and distorted. I thick chunk of aluminum bar stock would be my choice

[joeavaerage]

Hi,

I would advise against using an I beam unless you can machine it, as they are notoriously warped and distorted.

Over a full length yes there will be very considerable stress left over from the rolling process, however in the short lengths that OP is talking
its not an issue. None the less it would still be considered wise to ALLOW for post fabrication machining in your budget.

I priced making my mill in solid 6061 Aluminum, but the cost was about three times for the same rigidity. Yes, aluminum has a lot of nice properties
but its stiffness is less than a third that of steel, and stiffness is the absolutely most sort after requirement in any CNC machine. Damping....piffle....it comes
a very distant second to stiffness.

Bang for your buck steel is still the best, cheapest and stiffest material available to us. Yes, it is inconvenient to stress relieve and post machine.....but is still the
cheapest way to go.

Craig

[peteeng]

Hi Henri - Does not matter if you cut slow or fast, if you want to cut steel then it has to be very stiff. The Al construction extrusion is a poor choice. Heavy steel sections is what your going to need. If you look at mills with X500, Y200, X300 you will see they are massive. They are massive compared to yours and they are that way for a reason to achieve rigidity. If your stuck on construction extrusion maybe double them up and add corner braces. Or investigate UHPC for the base/column easy to cast a big L solid section and cheap. Then adhere steel or aluminium bits too it. Lots about cast parts in the forum (look up UHPC, CSA grout, epoxy granite, polymer concrete, mineral casting) Peter

[strostkovy]

Hi,



Over a full length yes there will be very considerable stress left over from the rolling process, however in the short lengths that OP is talking
its not an issue. None the less it would still be considered wise to ALLOW for post fabrication machining in your budget.

I priced making my mill in solid 6061 Aluminum, but the cost was about three times for the same rigidity. Yes, aluminum has a lot of nice properties
but its stiffness is less than a third that of steel, and stiffness is the absolutely most sort after requirement in any CNC machine. Damping....piffle....it comes
a very distant second to stiffness.

Bang for your buck steel is still the best, cheapest and stiffest material available to us. Yes, it is inconvenient to stress relieve and post machine.....but is still the
cheapest way to go.

Craig

The flatness across the flange is not adequate for profile rails. They will not be coplanar.

My mill is aluminum because the cost of the material plus the cost of work to get it flat was cheaper than steel. I also live on the second floor and have to make every pound count of my 350 or so practical limit count.

I agree the machine shown will not cut steel.

My personal recommendation for a budget and back friendly machine is to keep the extrusion, but dial it way back on the work envelope. to equal the width of the extrusion. If the extrusion is 100mm wide and 50mm deep and you want to cut aluminum, then I would reduce the travel size to a mere 100mmx100mmx100mm. Halving the length of the cantilever gives 16 times the rigidity, and you can really shrink this up to keep the deflection down.

[joeavaerage]

Hi,

The flatness across the flange is not adequate for profile rails. They will not be coplanar.

Yes, they will never be perfectly coplanar....but you might be surprised how good it can be. When I was doing my research I got a piece of 310x 137, about 1m long
I measured how straight the flange edges were and they were both within 0.1mm over that length and measured the coplanarity, fairly crudely, with a machinists
level and found that the height difference between the flange heights to be less than 0.04mm.

In the event I got the axis beds cast in grey iron. Casting gave me the flexibility to design to the dimensions I wanted as opposed to the dimensions of available sections.
It also allowed me to make the 'web' of the channel sufficiently thick (50mm) that I could machine T-slots without huge loss of rigidity. Casting in iron is not cheap,
and rough machining, stress relief, final precision machining is especially not cheap...but the result is superb...and that is after all what I paid for.

If I'd not had that budget I would have used 310x137 steel for sure. Yes it requires stress relief after welding and finish machining, but I was paying scrap value
for the steel and is still the most budget friendly way to produce a rigid frame.

Craig

[henriksson]

Thanks for all the responses, this is very valuable.

How are you attaching the column to the base? That's a critical joint there, and those aluminum extrusions don't give you a really good way to do it. I'd suggest a thick piece of angle iron on the outside corner to beef tt up, connected firmly to both sides.

I plan to tap the center holes in the column extrusion to M12, drill the base extrusion and use 8pcs M12x120mm bolts from below. I could find some 200x100mm wide 10mm thick angle iron from my local supplier, the question is if I would manage to drill mounting holes with a hand drill?

Hi,
aluminum extrusions are not adequate, they flex too much, and at the right angle junction, its really going to let you down.

Your description and desire for a low-med speed spindle suggest that you want to cut metals, in which case this design is just too flexible.

I would suggest that you get some real heavy duty universal column (steel) and cut the flanges to height, and either weld or bolt the column to it,
and the column would need to be steel also. This is the sort of material I mean:

https://www.trademe.co.nz/a/marketpl...B&gclsrc=aw.ds

Note that the flanges are 18mm thick and therefore thick enough for a drilled and tapped holes to mount the linear rails.

Craig

The problem is that I couldn't come up with a steel based design which doesn't require a welder and a sufficiently large steel mill. My local supplier does have 10mm thick 150x50mm box section (or even 200x100) for about the same price as the extrusions. The two complicated parts are:

1. Connecting the column to the base. Normally, you would weld a flange on the column and screw it into the base, but I don't own (or have space for) a welder.
2. Milling the box sections flat for the linear rails. Looking at projects such as PrintNC, maybe this isn't strictly necessary?

I think this machine would be capable of gently routing out 1/8"-1/4" aluminum to make side plates to reinforce the joint. It still won't be amazing, but should be adequate for light work with patience. I assume the design currently calls for long through bolts into tapped holes in the end of the extrusion.

I would advise against using an I beam unless you can machine it, as they are notoriously warped and distorted. I thick chunk of aluminum bar stock would be my choice

Light work with patience sounds about what I would expect. I'd fine with machining taking a long time, if I'm able to get good results. Given that the outer dimensions of those side plates are not critical, I might be able to cut out some 1/2" aluminium plates with the angle grinder and only use the mill for drilling. Do you think that would make the design viable?

Hi Henri - Does not matter if you cut slow or fast, if you want to cut steel then it has to be very stiff. The Al construction extrusion is a poor choice. Heavy steel sections is what your going to need. If you look at mills with X500, Y200, X300 you will see they are massive. They are massive compared to yours and they are that way for a reason to achieve rigidity. If your stuck on construction extrusion maybe double them up and add corner braces. Or investigate UHPC for the base/column easy to cast a big L solid section and cheap. Then adhere steel or aluminium bits too it. Lots about cast parts in the forum (look up UHPC, CSA grout, epoxy granite, polymer concrete, mineral casting) Peter

I was looking at something like a china import mill (e.g. Sieg SX2) which can do steel if you take it slow. I tried to search for epoxy granite / UHPC but the results I got was 1000kg+ monsters (like Alex CNC's machine for example: Alex CNC - YouTube), and that doesn't sound like a beginner project to me. I also couldn't find any project which didn't require milling the contact surfaces after the epoxy / concrete has hardened.

The flatness across the flange is not adequate for profile rails. They will not be coplanar.

My mill is aluminum because the cost of the material plus the cost of work to get it flat was cheaper than steel. I also live on the second floor and have to make every pound count of my 350 or so practical limit count.

I agree the machine shown will not cut steel.

My personal recommendation for a budget and back friendly machine is to keep the extrusion, but dial it way back on the work envelope. to equal the width of the extrusion. If the extrusion is 100mm wide and 50mm deep and you want to cut aluminum, then I would reduce the travel size to a mere 100mmx100mmx100mm. Halving the length of the cantilever gives 16 times the rigidity, and you can really shrink this up to keep the deflection down.

My understanding is that the main cantilever action on a column design comes from the length Y axis, the length of the X/Z axes doesn't really matter as long as I don't use it (for touch work, keep spindle close to the table and be in the middle of X axis). The Y axis of 200mm vs 180mm wide extrusion comes quite close to your suggestion. I chose 200mm because it means I can fabricate upgrades on the machine itself.

[ardenum]

Your frame is relatively simple and straightforward, I'd get a slab of precision granite and cut out the L shape from it. You'd need to buy some additional equipment like a lapping plate. With your current design you wont be cutting steel, it's not stiff enough.

Maybe consider a router instead. This guy 'cuts' titanium on his router.

[joeavaerage]

Hi,

Connecting the column to the base. Normally, you would weld a flange on the column and screw it into the base, but I don't own (or have space for) a welder.

Quite understand, I don't have a welder either. To weld thick sections like this really requires a much bigger welder than most of us can afford anyway. I hire a welder if I need one,
about $100/ day gets me a three phase MIG of 300A or more.

I prefer dual shield (CO₂ and flux) wire. 1.6mm at maybe 250A, 1.8mm at 300A or 2.4mm at 340A would weld these 10mm-12mm to full depth penetration. This is not the sort
of welder you'd have lying around! The last time I needed this done I took it to a company that do certified structural welding using exactly these processes, cost $50....too easy.

There is no doubt that you have to consider stress relief and machining using welded steel construction....but that sort of skill and capability is widespread in any industrialised country.

Craig

[peteeng]

Hi Henri - size does not matter for the casting its just as valid small or large. The two processes are 1) embed steel inserts into casting then have them finish machined 2) cast the part then bond on steel or aluminium parts that are finish machined. In the bonding process they need to be registered correctly using squares or jigs. Worth a think... vs where your headed. Peter

[joeavaerage]

Hi Henri,
you seem to have an aversion to getting something milled flat. Why?

Are there no large format mills in your town/city/district/province/country/planet?

It rather bemuses me that so many people go to such extreme lengths to avoid out-work machining, by use of materials or techniques which obviate the need,
but only at a marked increase in cost an/or complexity or compromise in the final result.

I can certainly attest that out-work machining is expensive, but with some thought in the design stage you can reduce the cost/complexity of the required machining
and have a very cost effective result.

Craig

[ardenum]

Hi Henri,
you seem to have an aversion to getting something milled flat. Why?

Are there no large format mills in your town/city/district/province/country/planet?

It rather bemuses me that so many people go to such extreme lengths to avoid out-work machining, by use of materials or techniques which obviate the need,
but only at a marked increase in cost an/or complexity or compromise in the final result.

I can certainly attest that out-work machining is expensive, but with some thought in the design stage you can reduce the cost/complexity of the required machining
and have a very cost effective result.

Craig

That is not the case in europe, judging by OP's username he's probably from a nordic country, where machining is insanely expensive. Same here in germany, to give you an example, a 200x120x60 aluminum heatsink, which has very simple geometry, I had quotes into EUR 2xxx. A cast column in EG size 1200x200x360, one quote was EUR 6xxx and that's just face machining. Upload a part in xometry, that's the cheapest you'll get, if you chose anything other than the standard machining, the price skyrockets.

[strostkovy]

My understanding is that the main cantilever action on a column design comes from the length Y axis, the length of the X/Z axes doesn't really matter as long as I don't use it (for touch work, keep spindle close to the table and be in the middle of X axis). The Y axis of 200mm vs 180mm wide extrusion comes quite close to your suggestion. I chose 200mm because it means I can fabricate upgrades on the machine itself.

This is true, though other axes can slightly change that. 100mm in Y and 200mm in X and Z seems like a good compromise to me.

I wouldn't sacrifice rigidity for upgradeability, as the frame will probably be the first thing to upgrade.

You can get custom laser cut plates pretty cheap. If you use 1/4" steel for all of your plate parts you can make full size L brackets for each side, and I feel like that would nearly double your rigidity.

[joeavaerage]

Hi,

You can get custom laser cut plates pretty cheap. If you use 1/4" steel for all of your plate parts you can make full size L brackets for each side, and I feel like that would nearly double your rigidity.

While I have cast iron axis beds, 115kg each, I did not have the budget to have the frame cast and so its made of steel, two large L shaped pieces, 1500mm x 1100mm and 32mm thick.
I think its a very useful style of construction. There is no right angle joint to design and engineer. The axis beds are through bolted to the frame on approx 5mm steel shims. It means that
no machining is required as the shims allow you to level and/or square the axis beds.

Craig

Craig

[henriksson]

Hi,



While I have cast iron axis beds, 115kg each, I did not have the budget to have the frame cast and so its made of steel, two large L shaped pieces, 1500mm x 1100mm and 32mm thick.
I think its a very useful style of construction. There is no right angle joint to design and engineer. The axis beds are through bolted to the frame on approx 5mm steel shims. It means that
no machining is required as the shims allow you to level and/or square the axis beds.

Craig

Craig

That's an interesting idea, also because it is a simple design change. Since it doesn't require much precision, I can cut the L shaped pieces with a plasma cutter if I can't find anyone to laser cut them for me. Do you think two 12mm (1/2") steel L plates would suffice for my intended size and use case? (This is the largest dimension my local supplier carries of sheet metal)

How do you separate the side plates? Also, what are the axis beds made of?

[joeavaerage]

Hi,
I had a plasma profile company cut them for me, there is really no need to grind or anything, they are usable as is.

I would recommend a thickness such that you can drill and tap holes in its edge. You could probably get away with 6mm holes in 12mm thick material, but not much more.
With 20mm material you could use 10mm holes to attach the axis beds.

In my case I used 32mm thick side plates, the big 'L's' and then 20mm thick rectangular plates, six of them, bolted between the side plates to hold them apart.
My intention was then to weld rectangles of 6mm, maybe 8mm plates between the sides making what amounts to a tube section. I needed to get my mill running
as soon as possible as I use it for work daily. If I had welded all the rectangles in I would have to stress relieve it when done. The nearest heat treating oven big enough
is in Auckland and I am in Christchurch, 1500 km's and on another island away. I elected not to weld the rectangles in, at the moment. I could not afford the time,
nor really the expense of sending the frame to Auckland and back, not to mention the cost of the stress relieving, $6NZD/kg, and it weighs 250kg.

Most of my work is making PCBs, and small plastic, aluminum and brass parts for instruments, for which my mill is way WAY overkill. Only occasionally do I load
steel or stainless parts in the mill big enough that I even notice that the frame is not really complete. I'll get there one day....but it seems every day I get more work
for my mill not less! When I can find a fortnight to take it to bits, finish the frame and send to Auckland for heat treating seems to bet getting further away....

My axis beds are cast iron. They are 700mm long 'U' shaped channels with the uprights and web 50mm thick. I had them cast for me locally. They weigh 115kg each.
I had thought about getting a length of continuous cast bar 250mm x 150mm x 3m from Europe, and then milling out the axis beds I wanted. The cast bar was about
$1000NZD, and shipping to New Zealand about another $1000 NZD plus 100's of hours on a mill turn 50% of it all to chips! So I had them cast at a cost of $3500NZD.
An expensive exercise but the result is all I could have wanted.

Craig

[awerby]

That's an interesting idea, also because it is a simple design change. Since it doesn't require much precision, I can cut the L shaped pieces with a plasma cutter if I can't find anyone to laser cut them for me. Do you think two 12mm (1/2") steel L plates would suffice for my intended size and use case? (This is the largest dimension my local supplier carries of sheet metal)

How do you separate the side plates? Also, what are the axis beds made of?

If you cut a couple of "L" shaped side plates out of 12 mm steel and through-bolted them together with that extrusion in between, that would make your design a lot more rigid. If you used some more of that steel as a bed, that would be stiffer than using aluminum only. You could add some T-slot extrusion on top of the steel to provide a way to hold things down.

[joeavaerage]

Hi,

If you cut a couple of "L" shaped side plates out of 12 mm steel and through-bolted them together with that extrusion in between, that would make your design a lot more rigid.

Yes, that sounds like a plan coming together.

Craig