Gantry CNC mill capable of mild steel? (warning: heavy flaming inside)

Let me just start off by thanking you in advance for any input, negative or positive, that you can supply. I am preparing myself for a fire storm of negative comments, lol.

We're building a gantry style CNC mill right now, the thread for it can be found in the wood machine section. In the beginning we were hoping for a wood capable table that might see some light duty aluminum cutting. But now we're wondering if our rigidity might not suffice for light steel work? It was never our intent to build a gantry style mill for steel so the design is probably all wrong for precision milling. I've attached some photos of the construction so far. I would be thankful to anyone who could point out any problems with our design.

For those interested in the design specifics the bed is made out of 6" I-Beam and weighs close to 1000 pounds as a very rough guess. The gantry is a piece of 6" by 10.5" I-Beam capped by end plates of 1/2" plate metal. The gantry weight is currently around 350 pounds without the Z-Axis carriage which we expect to add another 200-250 pounds including the X2 milling head and motor (it will be counter balanced). The z-axis will be mounted on 2" shaft running through Rulon bearings. Both the x and y carriages currently are mounted on 4" sawmill carriage wheels that temporarily run on angle iron (until we finish casting and scraping the cast iron replacements)

Thanks again,

There must be something I can't see. What's keeping the gantry from jumping up? All I see is wheels rolling on the X-axis beams.

Z-axis on one 2" round rod? 2 would work a whole lot better. Still, light milling only.

Just curious.

Dick Z

Richard - the only thing keeping the carriage down is gravity. It was designed for wood routing originally and we thought the carriage weight (600 pounds now) would keep it from jumping up. Is this not the case for milling steel? Do we need to redesign our long axis linear guides?


The z-axis does run on two 2" shafts. I wasn't very clear in my writeup.

Are there any improvements we can make that would increase our ability to do heavier milling?

Your 6 x 10.5 I-beam may not be stiff enough under a torsional load. I-beams are intended for loads acting parallel to their long axis and in this direction yours will be very stiff, but I-beams are not stiff under a twisting load and that is the type of load that is applied by a cutting tool that is located many inches away from the centerline of the I-beam. If you are comitted to using an I-beam and cannot switch to a length of 6" x 10" tube you may find it necessary to box in one side of the I-beam to creat a tube and enhance the torsional stiffness.

Great suggestion Geof we already have come to that conclusion but it's nice to hear it confirmed before going ahead and boxing in the backside!

The reason we never went with box tubing is with our current carriage wheel configuration we need the open front face of the ibeam to place the ball screw in. Perhaps we might have been better off going with conventional linear guides and saving ourselves the headache of engineering around something that wasn't intended for our use.

what is your plan for a spindle , to reliably cut steel , aluminum , and wood then your going to need a spindle that's quite versatile (fast , slow , and torque) , probably heavy and most likely expensive

also what drive motors do you plan to use , starting and stopping motion with that kind of weight may prove interesting if the motors are small , the accel/decel can be adjusted to compensate for this but with too small of motors then by the time your axis is up to speed , it will be decelerating

The spindle will be an X2 R8 spindle driven by a large (we have motors anyway from 0.5HP to 20HP) AC induction motor controlled by a 3 phase VFD. We already have the motors and the VFD we're just waiting for the X2 spindle to be restocked. For wood cutting we have a separate Makita plunge router.

To start out with we've got some tiny Nema 34 973 Oz-In stepper motors. The calculations we've done suggest these will suffice but in the case that they don't we will most likely switch to AC servo motors.

I also believe we might be able to lighten the preload on the X2 spindle bearings and cut up to 10,000 RPM with stock bearings. However maybe this is false information.

Geof's remarks on the boxing of the "I" beam are spot on.

As to using the weight of the gantry to hold it down, I wouldn't do it. You could use opposing rollers as on the gantry cross axis (Y?). If the top and bottom are not exactly parallel, a counter-acting load could hold the rollers tight against the top rail (X axis?).

The load could come from a spring, hydraulic cylinder or air cylinder.

Milling steel takes force and remember all that high school stuff about equal and opposite forces. LOL

If you intend to plunge a milling cutter or drill into steel, it will lift the gantry without arresting the X rollers from below.

Dick Z

The steppers are actually much bigger than you need !
From the pics, you are using a transmission with belts, and a moving nut, which is an excellent choice.

What is the transmission ratio ?

FYI, using an X2 spindle, you will never use more than about 60-90 kg force (600-900 N). This would be with about a 8-10 mm end mill.
A bridgeport used to use 600-800 oz-in steppers, and was a much more rigid machine and a better spindle.
These were the original Bp BOSS cnc machines.

A 3Nm nema 23 stepper at 3:1 will easily move a gantry of 300 kg if using ballscrews. It accelerates to top speed in 0.2-0.5 secconds depending on voltage and steppers.
My millling machine is like that, and it works well in steel.

At 3:1, 425 Ozin or 3Nm is about 240 kg push, or about twice what you really need.
With a 240 kg gantry you would have 1G of acceleration which is very high, about on a par with industrial production machines from a few years back. Today they might do 2G at most.
The 3:1 ratio is in the sweet spot not because of power (torque converted to push, at speed) but acceleration and accuracy (to 0.01mm, so build to 0.005 at +/-1 step).
I am using Nema 23s because they accelerate a lot faster than the 34s, and run a lot faster, they are just a better option.
However, I dont really see a need for max speed, and think the rapid speed people look at is often misguided.

You wont have any problems with your current motors.
Depending on voltage ? and drivers ? and transmission ratio you should get about 300-600 rpm usable max speed on the steppers.

You may find that the roller bearings are not rigid enough.

I use 20-25 mm THK rails in a steel frame. My mill is 2.4 x 1.8 m in size, and about 2000 kg.
I am using a 50 cm x 160 cm subtable at the moment (moving table in a gantry/portal design, very rigid).

Quote:
To start out with we've got some tiny Nema 34 973 Oz-In stepper motors. The calculations we've done suggest these will suffice but in the case that they don't we will most likely switch to AC servo motors.[/QUOTE]

Very informative reply Hanermo, thanks!

You're confirming what we suspected according to our calculations but doubted when seeing the motor next to the machine.. If that makes any sense?

Our transmission ratio is 3:1. You may be misguided by the pictures though, that shaft connected to the stepper on the gantry is actually a piece of 5/8" shaft with 10 tooth 1/2" pitch sprockets on the end. The long axis is so long we decided to try out chain drive as we've heard good things.

The z-axis and x-axis will both be driven by ball screws however, and we're planning on using no transmission.

hello can anyone advise me of which nema motors to use on my new mill router table
x=1000mm y=1500mm z=450mm 16mm ballscrews are fitted,
cheers,

One more question Hanermo when you say we might find the roller bearings might not be rigid enough do you mean our v-groove carriage wheels on our long X and Y axis?

Yes, I meant the v-groove bearings.
They are likely to exhibit some rattle. This will (may) mean you cannot, unfortunately, use climb milling.
Climb milling is less noisy, uses less power, stresses the machine less, and leaves a much better finish.
But it needs a zero-backlash setup in all the machine.

However, from experience, the heavier the machine the fewer problems you will have.
My gantry is about 700 kg or 1700 pounds.
The crosswise (ram on a Bp) piece is about 200 kg with teh Bp M-head as spindle and the 3phase motor, and I can rotate the ballscrew with my fingers, and the machine moves effortlessly.
The BS is 5/8" and could be bigger, but works ok (not great, just ok).

Also,
Timing belt is far superior to any other mechanical transmission technology mortals can afford.
Using a good setup, I got reliable 1 micron positioning on my lathe, as measured by DTI.
My theoretical physical resolution +/- 0.0003 mm and actual results always better than 0.001 mm.
And it is not a misprint (of course, I bought a new industrial 12x lathe, and used hi-res encoders with a 30.000 count/turn at 3:1, with a 20 mm screw, this would not work on a 7x for example).

A good choice is HTD pulleys and belts, in 15 mm widths.
I use these for everything, and buy in bulk.
An XL profile is not a good choice, nor are smaller widths.
Some of the other new precision belts are certainly good, but more expensive, dont bring any benefits (you wont need to position to better than 1 micron anyway), are much more expensive and you cannot easily make your own pulleys.

In my opinion, a 2:1 to 3:1 transmission is by far the best choice.
Steppers are not very precise ! They just are not.

Gearing at 3:1 allows real-world precision of 0.01 mm for milling and or routing.

It also allows higher acceleration due to the higher torque.
Using high voltage gives great maximum rapids, but is IMO overrated.
Your limiting factor will not be rapids, unless you are ripping plywood sheats into strips. Acceleration and precision (resolution) are far more important.

When you weld the stiffener into the I-beam (excellent suggestion, imo) use the heaviest plate you can, short of distorting the I-beam.
Something like one inch thick plate would NOT be overkill, if you can keep the I-beam straight.

As an idea and example, my table is made of solid 30x80 mm tool steel slabs, 1600 mm wide. I leave a t-nut sized space between them, and then rigidly bolt them with crosswise members underneath.
The table with 5 slabs is about 240 kg. The machine has space for 12 as a moving table or 22 as a gantry machine.

Using the 30mm thick slabs was an EXCELLENT choice, ime.

This is the exact info I needed! Again, thanks!

So far from your guys suggestions here are the improvements we will make:
* I-beam will be boxed in with as heavy of metal as possible
* Transmissions will be used on ALL axises (3:1)
* Sawmill carriage wheels might get replaced with another linear guide solution


Hanmero I have a few questions to ask you about your setup. First off do you have any pictures of your machine posted? I would be very inspired to see it! Secondly where do you buy your timing belts and linear slides? When we looked into THK linear slides they seemed quite expensive.

I am also looking for any information I can get about spindle heads. We've tentatively decided on an X2 head but we haven't purchased it yet. Would our money be better invested in something more durable, and if so does anyone have a suggestion? Hermone you use a Bridgeport head? I haven't had any success locating one of those myself!

THK slides.
Various forms and prices are available.
I just finished installing a set on my mill with 70 cm usable movement.
280$ for 2 slides and 4 carriages (at 200 kg load capacity/carriage). These were used but in excellent condition.
From a guy in south korea, on ebay, 70$ for transport by ems postal courier, only took 5 days.
What the slides do, they reduce the friction so much that
1. your resolution is vastly improved
2. your max speed and acceleration is improved
My moving table, 2 m surface area, 1.6 m wide, steel, 250 kg, can be pushed with one finger. Linearity better than 0.01 mm precision (local accuracy).

Spindle head:
There are 2 routes.
Milling head:
I bought a Bp M-type head, about 5 years ago.
400$ for Bp, 300$ for shipping and taxes etc. to Barcelona, Spain.
With a 3-phase motor and then a separate Hitachi VFD.
I use a 2.2 kW VFD, of a type that will work anywhere in the world (in 50/60hz, 110/220 volts).
This is for me to be able to use this head heavily. Bearings are available, and it provides me with industrial quality.
The motor is about 50 years old, and almost soundless in running.
I expect to be able to run the machine (many) thousands of hours making parts for me, and to pay for itself in 7 days.
The Bp has a fairly large quill movement, about 10 cm, and around 0.01 mm tir. It is old, and works very well.
Some of the bearings on mine are a little noisy at some speeds, but in the end results are excellent.
Using a face mill on a 30 mm wide piece of tool steel gives excellent results. Almost ground, shiny, flat. No finishing needed. Just laquer (paint, phosphate, black etc). Using 5-8 thousandth or about 0.1 mm DOC. Rectangular inserts.

The other option is a X2 head, minimill head etc.

The third is a hi speed spindle with VFD from china.
Cheap, not noisy, very good for engraving.
Under 400 all included, ebay.

All are good options.
The Bp is the most durable, heaviest, oldest, most industrial. Most expensive.
Will take the biggest cuts, I think, and is well suited for up to 6-10 mm end mills.
My app is making pieces in steel, thick and or large, so the Bp was the right choice for me.
I dont like the MT2 collets too much, but have both a face mill (very good, from cdco in the states, 2.5 inch) and a ER40 collet chuck for it.

Bp M-heads are available.
I would buy a Series 2 CNC head next, for me, and stiffen up my machine for it (easy).
The M-head is small and light, and I can lift it on my own.
The head is about 40 kg, and the motor 20 kg.
The series 2 is about 200 kg.

If you want, and can tell me which is your use, I can maybe help you choose the right spindle.
An X2 as a generic sort, not too expensive, might be a good choice unless you know for sure what you need to do.
The small VFD driven spindle from china will be good for small end mills, and lower material removal rates, with excellent quality, both in resolution and surface.


As an example, I may make a wood-processing line for building wooden houses.
I need to make 2 carriages of about 200 kg, each, running on linear guides, and planing a 250 mm D log 6 m long in under 4 minutes (on 2 sides).
The carriage steel structures will be planed, drilled and tapped on my mill.
The cnc machine will be about 7 m long and probably two lines in series or parallel.

I will also be making large flat steel components, with good edges. Hatches and portlights for yachts. So I need a large area, but relatively modest metal removal rates.


I buy pulleys, belts etc. all over the world.
I got the HTD pulleys at around 6-8 euros for 16-24-32 teeth, and 12-14 for 60 teeth (expensive).
I buy in sets of 20-200 at a time and get industrial manufacturer bulk discounts.

Quote:

---

Originally Posted by RICHARD ZASTROW

There must be something I can't see. What's keeping the gantry from jumping up? All I see is wheels rolling on the X-axis beams.

Z-axis on one 2" round rod? 2 would work a whole lot better. Still, light milling only.

Just curious.

Dick Z

---

Here is a picture of our Z-axis as it nears completion.

Pics

To anyone who has been following this thread thanks for the interest. As of today my machine has made it's first cuts and I couldn't be happier. I will post pictures and video in the next few days when I get my camera.

First I have a quick question to anyone who uses linear guides. I am heavily considering spending some money and upgraiding the X and Y axises (currently on angle iron and v-groove type bearings) to 25mm hiiwin linear rail and guides. These guides are each rated for dynamic and static loads of 3,210 and 5,180 kgf respectively and have a light preload. Would these be a good choice, or should I go bigger? I should also clarify my ultimate goal is light steel milling so I need them to be robust enough to handle those stresses.


So far it's been a great learning experience. Walked into this project not even knowing the difference between a milling machine and a drill press or having any understanding of machining. As it stands today our machine has many improvements that it will require before we are happy with it. Improvements will come as we push the limits of the current design. The next challenge ahead is learning how to properly use this machine. I will continue to post updates on this thread as short comings are realized and changes are made to the design.