Advise for home built "big boy" CNC mills?

This forum is dedicated to those Hobbyists Interested in CNC machining & 3D Printing in their home shops. (Digital Read Outs are also topical, as is CAD/CAM as it relates to CNC)

Moderator: Harold_V

choprboy
Posts: 322
Joined: Sat Oct 29, 2011 11:23 pm

Advise for home built "big boy" CNC mills?

Post by choprboy »

The CNC forums tends not to have much traffic... but I figured I'd ask. I've been thinking about building a "big boy" CNC mill capable of working in steel and aluminum. Something to ultimately replace my manual milling machine with far great capacity and accuracy. I'm not talking about the aluminum extrusion models with router motor spindles you see everywhere. I'm thinking 800-1000lbs of epoxy-granite superstructure, linear rails, 20"x30" work area, 200-300in/min feed speeds, 700in/min rapids, 400W+ servos, and a 3-5HP spindle.

The actual capacity of these 20-30Krpm spindles to do steel cutting has me concerned. Using a conventional machining cuin/min calc, it seems like a 3KW spindle would be underpowered. But as I understand it, using a high speed machining strategy you are taking very small cuts very fast, so in aggregate you move a higher cuin/min than you would conventionally milling. Does anyone have experience with the high power, very high speed spindles, and what they can do in steel/etc?
User avatar
Bill Shields
Posts: 10459
Joined: Fri Dec 21, 2007 4:57 am
Location: 39.367, -75.765
Contact:

Re: Advise for home built "big boy" CNC mills?

Post by Bill Shields »

You are going to build this from assembled pieces -> or are going to convert an existing manual machine?

If you are building then you want to purchase a spindle assembly, complete and ready to go

3KW is nowhere enough.
Too many things going on to bother listing them.
choprboy
Posts: 322
Joined: Sat Oct 29, 2011 11:23 pm

Re: Advise for home built "big boy" CNC mills?

Post by choprboy »

Building and assembling from pieces: casting epoxy-granite base and column with embedded steel mounting plates, scrapping in and mounting linear ways, ball screws and servos.I am looking for a combined spindle/motor assembly with ATC. Not converting an existing machine.

Power wise, I'm going to be running it off a 240V 50A outlet. So I'm probably practically limited to ~5KW for the spindle (~6.5HP), after taking into account servos, electronics, cooling pumps. etc. There are these Chinese spindles (some which get pretty favorable reviews), but the seem to mostly only go up to around 2.2KW, a couple in the 3-4KW range, and then jump up to 8-9KW with a huge increase in price. When you look at the specs the rated power output is at 10-12Krpm+ with very sharp drop off at lower RPMs. Going from my current 3/4HP manual machine, a 3-5HP (2-4KW) spindle would be a huge increase, but the RPM profile would be completely different than the 800-1200RPM I am use to running at.
User avatar
Bill Shields
Posts: 10459
Joined: Fri Dec 21, 2007 4:57 am
Location: 39.367, -75.765
Contact:

Re: Advise for home built "big boy" CNC mills?

Post by Bill Shields »

A very ambitious project..

Generally -> spindle assemblies are independent of the ATC that feed them.

Think of the ATC as a part of the machine ..not a part of the spindle.
Too many things going on to bother listing them.
choprboy
Posts: 322
Joined: Sat Oct 29, 2011 11:23 pm

Re: Advise for home built "big boy" CNC mills?

Post by choprboy »

By spindle with ATC I meant a gripper and release system built into the spindle (not a threaded drawbar or fixed collet) for quick change, probably for a BT30 with pull stud. If I really went 10Krpm+ I'd love a HSK, but I think tooling wise that's out of my price range. Before I started looking for a spindle I was thinking 8-10Krpm max. But all the new spindles out there seem to be 14-30Krpm.

That's where I think I'm stuck. I was expecting to have a spindle running a 1-2Krpm with a large endmill hogging material at a few dozen in/min, and then small endmills at 8-10K and high feed rate for finishing passes/detail work. But the 14K+ spindles have an effective power of a few hundred watts at 1-2Krpm.

At some point in the future I would add a carousel and tool change arm for fully automatic tool changes.
User avatar
Bill Shields
Posts: 10459
Joined: Fri Dec 21, 2007 4:57 am
Location: 39.367, -75.765
Contact:

Re: Advise for home built "big boy" CNC mills?

Post by Bill Shields »

Ah...that is clearer now.

Spindles capable of what you want are out there..but may cost more than you are interested in spending....especially when you look at support systems for cooling at high speeds.
Too many things going on to bother listing them.
LIALLEGHENY
Posts: 363
Joined: Sat Jan 11, 2014 12:36 am
Location: Bohemia, NY

Re: Advise for home built "big boy" CNC mills?

Post by LIALLEGHENY »

I have to ask why do you feel you need such a high RPM spindle? You want to move from a 3/4 HP , 800 -1200 rpm manual mill up to a 18000 RPM machine with 5X the HP......are you doing production work that would warrant this, or 3d work running small endmills that would warrant this? Do you have so much work that shaving minutes off of each part is a necessity? Next question, is your power source single or 3 phase?
There are plenty of good used 20 x30 travels machines available with 8000 RPM spindles. Perfect example is a Cincinnati Arrow VMC750 , I recently bought one for $5000 . Linear ways , rapids over 1100 IPM, up to 590 IPM cutting speeds, 8000 RPM, 12 HP , 21 position tool changer......I run steel , stainless , aluminum ,bronze and don't need any more RPM than I have.....
If space is an issue, well then you may need to go a different route......maybe pick up a heavy duty CNC knee mill and adapt a higher RPM spindle to it, there are many options . I have a Brute Dahlih CNC knee mill in my shop, been sitting there for years with a bad control unit on it....the machine itself is like new......but I couldn't justify putting new controls and servos on it when the cost was going to be as much as the Arrow, if not more......and the Arrow had bigger travels, a tool changer, higher RPM spindle , etc. I'm not knocking your idea of wanting to build a machine, but when you start to add up the cost of building your own versus what's already out there you may want to change your direction.

Nyle
User avatar
Bill Shields
Posts: 10459
Joined: Fri Dec 21, 2007 4:57 am
Location: 39.367, -75.765
Contact:

Re: Advise for home built "big boy" CNC mills?

Post by Bill Shields »

As I commented previously 'ambitious project'
Too many things going on to bother listing them.
choprboy
Posts: 322
Joined: Sat Oct 29, 2011 11:23 pm

Re: Advise for home built "big boy" CNC mills?

Post by choprboy »

That's just it... I didn't think I needed a 18000 RPM machine. I have several things I want to work on (custom transmission bell housings, adapter plates, maybe one day milling a model v8 engine from a single block) that are far larger than the work area on my current mill (about 6"x18"). The bell housing I'm currently working on is about 18" in diameter. I figure a 20"x30" travel (with about 12"-18" of Z) would do for what I currently have envisioned. And possibly some day I'd like to try my hand at an impeller or other 4D/5D milling where I may need a high spindle speed for small endmills.I was thinking in the 8000 RPM max range, with most use around 1-2K RPM.

But then I got looking at what spindles are easily available and most of them are 14-30K+ RPM and don't produce rated power until the 10-12K RPM range. As I was expecting to do most of my machining at a much lower RPM, I started wondering if I was planning/looking for the wrong sort of spindle and should look more at very high speed machining strategies.

I only have single phase power in my shop. I have a dedicated 100A shop sub panel off a 200A service, with 20/30/50A drops to my machines. So I can't run a 10+HP machine. Floor space is also very much at a premium, so a big machine like a VMC750 wouldn't fit. I was planning on building a horizontal mill and I think I can fit it in a roughly 4'x6.5'x6' space (quick calc). A VMC750 clocks in at 9'x9.5'x9' and several times heavier. I also haven't seen a VMC750, or anything similar, for $5k around here, it is 2-3 times that. Occasionally a pretty clapped out 1990's Hass or a Tormach will show up, but not often. I did a machinery auction today and an Acra 10x54 manual mill converted to CNC went for $9500, the manual mills went for $3500-$4500.

For my machine with a 3-5HP spindle, linear ways and ball screws, probably the all-in-one servos, and LinuxCNC/Mesa controller, I was budgeting around $5-6k total for parts (not including tooling), plus a lot of my labor.. for what should be a very capable machine.
LIALLEGHENY
Posts: 363
Joined: Sat Jan 11, 2014 12:36 am
Location: Bohemia, NY

Re: Advise for home built "big boy" CNC mills?

Post by LIALLEGHENY »

Where is "around here"? Yes not having 3 phase is going to limit what you can do, but 50 A is still substantial and plenty of power for a decent size phase converter. I would still suggest finding an existing machine and upgrading the controls if need be. The knee mill I have for example has a 4200 RPM spindle, approx 16 x 30 travels , 6" quill travel plus the knee , 5 HP, NMTB 40 taper, and weighs in around 5000 lbs. But still takes up more room than you're figuring on. One thing you can't skimp on is "mass" especially as your travels increase. The more mass and rigidity the better. Personally I think you would be better off with a vertical machine as setups will be a lot easier when your not fighting gravity. If you were doing production runs of parts that could change the equation, but one offs would be better suited on a vertical.

Nyle
choprboy
Posts: 322
Joined: Sat Oct 29, 2011 11:23 pm

Re: Advise for home built "big boy" CNC mills?

Post by choprboy »

Arizona. Right now on Craigslist there is a:
- Tormach 770, that's 7'x14" travels, for $11750
- Tormach 440 for $10000.
- Supermax CNC'd 9x42 mill for $11500
- Bridgeport 9x42 manual mill for $5250
- Hass VF4 in nice shape for $23000, that's probably the sweet deal of the current crop
- and various 2'x2' to 3'x4' aluminum framed routers/engravers/plasma CNCs for between $1000 and $6000

... that's fairly typical around here. Occasionally there are some deals on VMCs where you can pickup one or two and a bunch of parts in the $3-5K range, but the are usually quite old, have been sitting outside, and are in need of servo/spindle repairs. I've thought about picking one up, but space is an issue. I was figuring with a cast epoxy-granite I can build it in large blocks (multi-100lbs to 1000lb) and bolt together into larger structure to have a dampening mass.But I'll be doing a bunch of casting tests/measuring bending moments before I get to far down that road. For me this is a hobby machine. I do occasional projects for friends and would love to be able to do it as a job, but I don't have anywhere near a base to live on currently.
pete
Posts: 2518
Joined: Tue Feb 10, 2009 6:04 am

Re: Advise for home built "big boy" CNC mills?

Post by pete »

I've got to say I agree 100% with Nyle's points. And you did ask for advise. So fwiw and the little I'd like to think I know.

Machine design only looks fairly simple on the surface until you start getting into it in depth. And as the accuracy requirements increase so do all the finer details. How much reference information have you studied and how much do you already know? Then there's some fairly expensive additional tooling if you don't already have it. For your sizes you mentioned I'd be wanting a known certified and almost perfect 3'x5' or larger surface plate, cylindrical square, straight edges, 10ths indicators, levels, build or buy a used Kingway tool etc. There's at least 3 references I know of that I wouldn't even try to get a project like this started without some long hrs studying what they have to teach. One I've mentioned on this forum quite a few times before. Run a Google search for a free online PDF of the book Testing Machine Tools written by a Dr. Georg Schlesinger. Print that whole book off and get real comfortable with it. That would be the bible for machine tool alignment and not optional to study in detail for sure. The second would be the roughly $150 book from Moore Tools titled Foundations of Mechanical Accuracy by Wayne Moore. The explanation and diagrams within it about the 6 possible misalignment's on any slide surface are the clearest and best explained I've seen anywhere. Once you start moving towards getting less than + - .001" expected accuracy things really start getting complex and there's a lot more to pick up from that book. Again not optional to fully understand and appreciate before even starting imo. The third would be the roughly $100 book from Industrial Press written by E. Connelly titled Reconditioning Machine Tools. And yes that one is mostly about rebuilding worn machines by scraping them back into new or better accuracy. But the test methods, checks and maintaining that square and geometrically correct accuracy in all 3 dimensions while the components are being built is not nearly as simple or intuitive as it might seem. All 3 references I've mentioned expand on each other with the finer details that might be less clear in the other two. For a bit less than $300 you'd definitely save multiple times there cost in errors while trying to build any machine tool from scratch.

Afaik the epoxy / granite idea isn't a real new concept and there were and may be still is at least one high end machine tool manufacturer using it over the last 20-30 or more years. But you can bet they bankrolled, built and tested more than a few prototypes before they knew where to modify and redesign to work around the inevitable beginning errors. With a one off machine build that's not a monetary or viable method for you. There's been some really smart clever threads on the CNC Zone forums I've read with lot's to learn from, although I sort of lost interest and haven't really kept up to date. There's also some epoxy/granite builds on YouTube, a lot of them while I very much envy there CAD/CNC skills seem to be making a lot of basic entry level design and construction errors. I have to assume many just won't spend the time and in reality minor costs for at least those reference sources I've mentioned. It's literally impossible to know too much about this or the basic engineering concepts behind building any machine tool. There's some real complex engineering formulas to understand even just for the possible ball screw shaft whip when there being driven at high speeds. I know just enough about some of this to understand I very much don't know nearly enough.

It's my understanding as good as cast iron is for absorbing vibration, the epoxy/granite is capable of even better. Besides the capability of doing it in home unlike large cast iron castings, that vibration absorption is a real built in benefit. But no material is perfect. That E/G mix is also a bit like concrete. Good in compression as long as there's an adequate underlying support structure. And concrete is really easy to break without that support. I've broken and loaded out 10's of thousands of tons of even steel reinforced concrete. And like concrete, that E/G mix is fairly weak in tension or with any impact shock loads. Yes it's strong, it's also real brittle on it's own.So you have to build in additional support to work around it's natural weakness and pre plan and then cast in any metal threaded attachment points extremely well. After studying all those E/G machine tool build threads, I think if it were me I'd go with an underlying structure of a fairly heavy welded steel plate much like a egg crate assembly with the E/G mix cast around that. I'm sure by now you also know that E/G casting has to be done under vacuum to remove any trapped air to end up with a solid monolithic casting. I think I'd build up and weld that whole structure much like an I beam works, the steel allows threaded connection points, properly supports the E/G mix while that absorbs the vibrations. Even then I wouldn't hesitate to add actual concrete reinforcing rod anywhere I thought would help.

No matter what you try for machine tool design it's guaranteed to have been tried somewhere else multiple times before. So steal ideas from what others have already done and save reinventing the wheel. Go with what's proven to work in the past. And I'd also be studying every exploded view cnc machine tool parts diagrams I could find online and borrow heavily from those. I'd also very much agree that a vertical mill design will be much better than any horizontal spindle arrangement. Horizontal with an outboard arbor support wins over a vertical for absolute metal removal rates, but in any home shop without volume part production that removal rate difference is almost incidental.

Spindle rpms? There's no such thing as a direct drive electronically controlled spindle that can do it all without some form of an additional mechanical speed reduction that also increases that torque at lower rpms. End mills are obviously just one type of milling tool. Unless you dedicate this to just that type of tool, your still forced into that lower rpm requirement for some tooling. Yes you can do helical interpolation of holes. For some work that's still not good enough and even more so unless you go with very expensive ball screws and nuts that would instantly blow most of your budget. So one way or another single point boring isn't going to be optional for some work. Run the numbers for boring even 4" diameter holes in steel, and even with carbide your low rpm requirements makes that low rpm high torque range real important.

Just like the almost universal world wide standard of 5C collets, the Bridgeport dimensions are still fully entrenched in tooling design today. There dimensions and because almost every shop had at least one for the last 70 + years became that now universal standard at least in North America. So one design choice your almost locked into is using the same tee slot size and pitch spacing between those slots every real Bridgeport and most of the clones use since a lot of table mounted tooling is still designed around those dimensions. I didn't understand that at the time when I bought my 3/4 sized BP manual clone. Luckily for me they did use exactly what the BP's use for size and spacing. Later when I bought a matched pair of Glacern 615's that use 3 hold down bolts on each side there bolt pitch spacing matched perfectly with my table. So that's one design detail worth keeping in mind.

Physics don't scale well either up or down, the larger the spindle is the more the expense and difficulty it's going to be to get that +10k high rpm. The various 40 tapers are now probably the most common today, so if it's made at all it will be made in at least 40 taper. 30 taper is much less common, but most tooling can still be found if you look hard enough and whatever head, spindle and bearings you decide on are going to be a whole lot cheaper at that size than the larger 40 taper. Threadexpress on YT did a great comparison and analysis between BT 30 and R8, and while the 30 taper looks on paper like it's far superior to an R8, it's really not all that much better. https://www.youtube.com/watch?v=m9rMbSLyAvM That high rpm also starts getting you into the requirement of high rpm balanced spindles and even the tooling. None of that is going to come cheap. Adding an ATC to your mill build almost dictates your going to have to go with a 30,40 taper though. And I highly doubt there's anything that will be even close in price today than the Tormach TTS quick change set up. Unfortunately I don't know if it's available in a 30 taper yet. And I don't know of any real high rpm off the shelf R8 spindles. That's a couple of items you'd have to do your own research about. Your size and table travel numbers would be much easier to do with a knee type design. What I don't like about that knee and it's built in unavoidable defect is that it has real accuracy issues at larger table travels. The non optional clearances that allow the table to even move, also allow the table to slightly rock as gravity affects that table during larger moves. An actual bed mill design would be the way I might go. So with that design your almost then locked into moving the spindle on the column. There is one non fixed bed design that might work around some of the issues. I don't know who invented it first, but a whole lot of the Euro manual mill designs use it extensively such as Deckel to name just one. The X axis ways are attached to the knee in a vertical orientation and are the full length of the expected travel. That works much like the same idea as that I beam concept. And the tables made for these are simply a fixed tee slotted table bolted to that X axis with the Y axis as a ram the head moves in/out on. There far more rigid than any BP and don't have the same built in issues with that table rocking. Some machine tool designs use a fixed bed, some with only a Y axis movement, and traveling X on the column as well as having Z. Any I know of designed like that are far larger and heavier than what your thinking of. Building in adaptability is another issue even if some of it has to be done with manual adjustments on a full 3/4 axis cnc. That Bridgeport design is so common not because it's even close to the best, it's turret, ram, tilt and nod features are what makes it adaptable to almost any job that will fit on the table. But I can't visualize how it could be easily added to a fix bed mill. So that Deckel/European mill design might still be the best overall design choice. Maybe some of these thoughts might help anyway.
Post Reply