[cnccnc222]

I'm considering building a large format router/mill to carve wood and stone, and would like to beef it up as much as I can, and am curious to know if it's within all my powers to even make it make cuts into steel. I want to make it about 2 by 3 meters, and put a 4th axis headstock and spindle on it.

At this point I'd like to discuss the build just one aspect at a time. Let's assume I'm building it into a perfect rectangle that I've carved out of pre-cambrian rock mountainside, and that the gantry will be made of pure magic. Magic that weighs as much as required to get the job done. And assume that there is a machine shop around the corner that can lathe and outside diameter hone heavy gauge steel pipes to accurate dimensions.

Here is an idea for the rails I'd like to get feedback on.

The rails would be made from high carbon steel tubes filled with epoxy granite. Perhaps 5 inches outside diameter. A slot would be milled in them and two would be joined in parallel with a 12 inch wide epoxy-granite plate. The two joined rails should therefore be quite stiff and act as a frame - supporting each other.

They would have pillow blocks riding at the top and bottom, connecting to the gantry. The four pillow block bearings would be wide and use filled teflon for better damping than roller bearings. Pillow blocks should lower the cost over roller bearings as well. The idea is to make rails with better characteristics for the project than are commercially available, and at a better price. Epoxy-granite filled tubes combined with filled teflon pillow blocks seem like a good design for that. The added friction would be overcome by using heavier stepper motors.

The rails would attach to the sides of the concrete table frame through bolts through the rails epoxy-granite middle connecting slab. Careful attention would be paid to the attaching bolts, perhaps using Moglice to take out any play. The table would have 6 or more legs, so the rails would be supported in at least three places.

For joining the parallel rails with the epoxy plate, a casting form could be built with the rails in place, allowing for precice placement of the rails during the casting process. The rails would be cast together laying on a perfectly flat expoxy-granite slab.

In order to make cuts into steel, what would be the required dimensions of a well made rail as described, if it were to have only three supports?

[hanermo]

The framing of the question is not ideal.
I built a similar sized mill, so I have some experience. The build took 13.000 hours, to industrial quality.

There are any nr and form of structures you can use, but it is not ideal to escape from the common shapes, as seen in modern VMCs, for example.

To do cuts in steel, with a 1 mm end mill, at 12.000 rpm, requires only a few kg of force.
To do the same with a 20 mm end mill, at 4000 rpm, requires perhaps 300-600 kg of push force (there are tables for this, google MRR).

You will not be able to make better stuff than is commercially available, at a better price. This simply cannot be done.
Cheaper or better, yes, but not both.

People do not invest 2M€ in a cnc way grinder, used to make linear guides, because they are uninformed, but because thats the best, cheapest, most profitable way.

What you can do, if you want, is to make a big machine, much lower cost than commercial offerings *in big machines*.
This is because the machine will be less (commercial) in some aspects. Esthetics, power , rigidity, practical etc.

You can convert lots of manual work, instead of investing some $$ in components.

Nevertheless you will end up using profiles linear guides and trucks, as they are cheap and easy to use, to make the proposed machine work well and efficiently.

So frame your question better, please define:
Proposed power ?
1 hp (Bridgeport) or 50 Hp (VMC).
Rigidity to mill steel efficiently (carbide tooling) or floppy HSS like an old bridgeport ?
New tools ? (variable flute mills, large face mills etc) or flycut HSS ?

You can bolt or weld some I beams together, and have a gantry that will mill ok to so-so, in a few hours.
In any case it will be heavy to do any real material removal.

A Bp is about 1 x 0.4 m in size, at 1000 kg. Very floppy (today).
Whenever you double a dimension, multiply mass by 4x in steel or cast iron. By 6x if using somethng else.

A 2x3 m gantry mill will mass about 10.000 kg, if using steel/cast iron, well built, and can absorb maybe 20 Hp(30) of spindle power.
(Ihave seen several from many manufacturers. Most were heavier).
My 1x2m bed is
1. A portal. This is much much more rigid.
2. About equal to a Bridgeport (ie quite floppy). Ok for small carbide tooling, about 12 mm end mill. About 1-2 kW.
3. About 1000 kg.

Concrete/EG/sand/etc can be used as a ballast to reduce vibrations.
Dimensions can only be defined once the required spindle power and thus pushing (moving) forces are known.

A Bp is about 50-100 kg in force (1-2hp).

The stiffness shoud be such that low deflection (0.01 mm-0.02 mm) is seen on a dti at spindle nose at required force.
A bp might deflect 0.01 mm at 50 kg. A VMC (6000 kg) maybe 2 microns.