Milling machine design and build

[handlewanker]

The side forces for a 75mm diam shell mill with at least six tips and a depth of cut at .025mm will put even the "massive" column of Assuratman's build to a disadvantage.

Looking at the design for this mill and the one in post #1 and I see only end mills and slot drills as the cutting force to be resisted, forget about any slab or shell mills, you won't be able to swing a shell mill slow enough and resist the side force.....75mm diam at 500 rpm?...forget it.

BTW, the square tubing with rounded corners is not the ideal section to be working with, and the added welded sections to make the corners square is bad design.

I stand by my statement that the 3 HP motor and head mass being driven up and down will make some interesting Z movements for small drill ops.

As this thread is for mill design not actual existing mill criticism, I feel justified in expressing my opinion for what it's worth on a new hypothetical build not expressly designed to any particular design criteria.

So, I toyed with the idea (electronically on my graphics pad) of having a milling head on the front side of the column and a motor on the back side, both on seperate slides, but driven up and down by seperate stepper motors synchronised together.

The stepper motor for the head only has to cater for the weight of the head which can be designed to be as light as possible and using only the ER40 or R8 collet system in the spindle nose.

The ER40 collets will handle 25mm cutters so plenty of cutter capacity here, and as the ER system is nose clamped there will be no need to have an overly massive spindle diam to cater for a draw bar, so high speed does not become rocket science design for the balancing act.

The motor being at the rear of the column is to some extent balancing the drive loads, but in actual fact it is the weight of the motor that the head is removed from that is the benefit.....the motor just follows the head up and down on it's own as a slave due to it being syncroed to the height variation of the spindle.

In this design the head does not "see" the weight of the motor so it ceases to be a factor for the stepper motor drive capacity.

This can also be accomplished by the drive pulleys at the top and a keyed or splined spindle as in a drill press set-up, but as the spindle has to be freely moving in the drive pulley there is a tendency for chatter and vibration to appear at the cutter point, also you would have to have a draw bar if'n R8 collets were desired.

Anything that reduces the weight of the Z axis spindle assembly is good design, and counterbalancing the load of the motor and spindle is not going to solve the problem due to inertia.

As a last solution to flexing, I would drill the square tubing corner to corner diagonally and alternately with 12mm holes with deep chamfers and weld in 12mm diam steel rods from the outside, spaced about 50mm apart down the inside of the column.

This will immeasurably stiffen up even the flimsiest tube, as the square tube really has no resistance to torsional stress longtitudenally, and all the force from the cutter is directed to the column as a torsional stress.....no need to fill the column with concrete, it'll be as dead as after that.
Ian.

[christofer]

"BTW, the square tubing with rounded corners is not the ideal section to be working with, and the added welded sections to make the corners square is bad design."

How would you suggest building the column and base? I would like to keep the machining as simple as possible, that's why I went with a steel tube. I like the idea of welding steel rods diagonally across the tube. What is the disadvantage of squaring the column by welding steel bars to it?

I think I may be forced to go with a counter weight or air spring to support the head due to cost and complexity reasons. I suppose I could use a spline type shaft to drive the head. I will have to investigate the complexity of this.

[asuratman]

The side forces for a 75mm diam shell mill with at least six tips and a depth of cut at .025mm will put even the "massive" column of Assuratman's build to a disadvantage.

As a last solution to flexing, I would drill the square tubing corner to corner diagonally and alternately with 12mm holes with deep chamfers and weld in 12mm diam steel rods from the outside, spaced about 50mm apart down the inside of the column.

I put 5" angle steel with 5/8 " bolts and 1" thick steel plate with 7/8 " stud and 2 ea nuts on the back of column as you can see on my mill. They are good enough to withstand with side force and backward force. My table is 450mm X 650mm. It will hold cuttingforce due to 75mm width shell mill with 1 mm cut depth.

[handlewanker]

Hi all, without actual figures to work with that are formulae or history, the best you can achieve is to make it as you feel it needs to be made, that is go for more rather than less.....it doesn't have to fly so weight if'n it creeps in doesn't impact on the eventual machine design.

When it's finished and you step back and admire it, you'll see a dozen ways you could have improved it by a different approach.......who cares.....the mill has to do a specific job and that is to produce a true surface without deflecting and making a bad finish.

Fabrications are rarely as pretty as a shop bought cast iron machine, but it's only a one off, so make it as you have the ability and equipment.

I would rather see a completed fabricated and roughly designed mill than a perfect design still on paper, never to be built.

Fabrication by welding has it's own requirements and a lack of stress relieving will show up sooner or later.

Welding creates heat, which creates expansion and as the structure cools the contraction will do some awkward "adjustments" to the initiial set-up, especially after machining.

I'm not a fond believer in bolted together structures...they can work, but need to be constructed very carefully.

With a welded structure, Mig welding is not an option due to it's lack of penetration......yeah yeah, some people can weld 50mm deep with a Mig....and pigs fly too....LOL....and Tig does not have the OOMPH to do any thing significent.

A 140 amp hobby stick welder at best has a duty cycle of about 25%, so carefull welding times are needed to prevent cutting out in the middle of a seam weld.

You would need to use a stick welder to get any degree of weld integrity, especially if'n you are fairly new to the welding game.

We're dealing with rule of thumb design here, so whatever you fancy then that's your pleasure, no matter how you perceive it, everyone will find their own way to Heaven.

Don't be put off by the "experts" choice of design....they aren't in your shoes when you come to source materials and build it.

There is a very interesting CNC mill designed and built in the USA by Levil Technology and goes by the title WL 400...it has a fixed table with a moving X and Y slides, also the very simplest tool changing system too.
Ian.

[FannBlade]

I would rather see a completed fabricated and roughly designed mill than a perfect design still on paper, never to be built.

Ditto
Build it and use it to build the new revised version.

[handlewanker]

Hi Fannblade, I think the object was to build a mill to do some CNC work, not give birth to another mill to build another mill to build.......whatever....LOL.

Machine building in itself becomes a passion when there are so many design concepts to choose from.

Starting right at the bottom is reinventing the wheel, but nonetheless it gives you free reign to develop your own machine concept.

BTW, just read your 3 in 1 rebuild....you have the patience of a woman darning socks....LOL.....best of luck getting it to final stage.
Ian.

[christofer]

Here is my next revision of the design. I copied the idea of a VMC. I hope to have a mini VMC when I am done with this project. Travels should be about 24" X, 12" Y and 14" Z.

The column is 8x8 steel tube with a 3/8 or 1/2" wall. The base of the machine is a piece of 12x8 steel tube also with a 3/8 or 1/2" wall. In the base I intend to weld bulkheads every 8" or so. They will be 1/4" steel plate.

The rails are HSR25.

[asuratman]

Hi,
Its for cutting light material such as wood, mdf, etc OK. Depend on material you want to cut, like steel, you need more rigid and weight than your design. Just continue on with your design now and try.

[handlewanker]

Hi Chris, the VMC design, as it's aready "up and flying" will probably be an easy way to aquire design integrity without having to do vast sums of FEA...aka pencil and back of envelope scratching while sitting on the toilet, where best ideas originate....LOL.

I see you advocate welding bulkheads in the base tube every 8" or so....inside the tubing?????....not as easy as you anticipate.

The more welding you do the more stress factors occur from expansion and contraction with no where to go....which is OK if'n you can go and anneal the whole structure in a great big fire, which will allow the steel to relax and assume it's final shape, before you want to do any machining.

The final design when machined will rely on the flat surfaces being flat and square to one another so that the twin linear rails will be in line and on the same plane......shimming is a messy solution to untrue surfaces, but adding pads where the rails sit can work for spot machining and hand fitting etc.

In the Fadal design the column (without any dimensions to go by) looks really squat, so it's ideally suited to resist torsional forces, much like a tapered pyramid shape.

BTW, the larger base tube will only give you a more unsuitable flexible base to build on, unless you add the bulkheads you favour.

I would still go for the corner welded in rods in place of full bulkheads, as you can add the steel rods all along the base tube and weld from the outside, making it almost as rigid as a solid piece of steel billet.

In the final design of the head, post #31, there appears to be a myriad of pieces that will tax your welding ingenuity as to how you will assemble the structure.

Pieces fitted together and welded in the corners are not the easiest weld seams to achieve, even for experts, and welding from one side is the most stress inducing join you can introduce to a design.
Ian.

[christofer]

I really am considering the idea of welding rods from corner to corner as you suggest.

Perhaps for the head I should use another piece of steel tube so that I do not need to weld a complex structure together. As of right now the head is 6 pieces of steel assembled into a box structure. I could achieve the same result with a piece of steel tube and two steel plates. I might add some additional bracing to stiffen it up or perhaps weld rods from corner to corner in it as well.

I intend to have the structure stress relieved before machining so I am not too worried about warping it while I weld it together.

I hope to not have to shim anything other than perhaps the alignment of the head. I intend to joint replicate the column to the base after squaring its alignment with adjustment screws.

[handlewanker]

NOPE, definately not....no adjusting screws for the base.

Mate the base to the bottom of the column, revealing the fit between the faces with mechanics blue and filing and scaping if necessary to get a dead flat fit around the outside edges, (25mm wide band) making the centre slightly clear.....then bolt and dowel it down.

Any contact at the centre of the base/column interface will lead to a potential pivot point situation......the column will lean to the point of least resistance under load like a seesaw.

Personally I would weld the column to the base, as I find any bolted joint a potential weak area.

Bolted joints are OK, but not good, if'n you are going into production and wish to have modular construction for ease of manufacture.

For a one off machine the least loose bits the better, and welding makes the most efficient fastening system ever invented.

I find square tube with rounded corners a pain to work with, due to the radius on the outside edge.

I would use two pieces of heavy channel steel 8" X 4" section with the 4" side cut down to 2" and the lot welded together with one seam per side to give you a column 8" X 4" but with square corners.

I used this construction for the box headstock of a lathe I built in the early 60's and still use today.
Ian.

[jsheerin]

Anything that reduces the weight of the Z axis spindle assembly is good design, and counterbalancing the load of the motor and spindle is not going to solve the problem due to inertia.

Just a quick note on counterbalancing - the above quote is not quite correct. If your acceleration is less than 0.33g either up or down, then a counterweighted Z structure will always require less force from the linear motion system than for a non-counterweighted design. If your acceleration is over 1g, then a counterweight design always requires more force from the linear motion system while accelerating. In between these two levels of acceleration, the counterweighted design will require less force to raise up to a 1g acceleration, while the non-counterweighted design will require less force to accelerate downwards at greater than 0.33g.

Basically for the non-counterweighted design, Fs= ma+mg = m(a+g)
where m = mass of the moving parts
Fs = linear motion force (due to a screw being turned by a motor or whatever - we want this to be small so we can use smaller, cheaper motors, drives, power supplies, all of which generate less heat to warp the machine frame, etc.)
a = acceleration of the moving parts upwards
g = acceleration of gravity

For the counterweighted design, Fs = 2ma + mg - mg = 2ma
where m now equals both the mass of the head and the mass of the counterweight

So for example, for g=1, a=0.1, Fs for the counterweighted design will be 0.2m and the non-counterweighted design will be 1.1m. For a=-0.1 (accelerating downwards), Fs for the counterweighted design will be -0.2m and Fs for the non-counterweighted design will be -0.9m, but really we care about the magnitude of the force so we'd consider the non-counterweighted design to need more force.

Having said all that, the other thing to think about is that the non-counterweighted design will require a constant force input from the motor to hold the moving mass in a constant position (assuming you are using ball screws which are capable of being back driven by the weight of the head). This means the linear motion motor, motor driver, and power supply are all being worked harder thermally when the Z axis is not moving which is probably a lot of the time if the parts you're making are anything like mine.