Gantry mill

[shedbob]

This is a fascinating thread of a hugely ambitious project.Thanks for sharing your technical knowledge,insights and challenges.I wish you well with this or any future build. I will certainly continue to follow this,or any other build thread you start.Thanks for posting.

[jsheerin]

A small update - I found an optical square on ebay this week. It's a Farrand type 88305, made in 1962. It's specified to be 90 degrees +/- 2", but hopefully I can calibrate it a bit more accurately than that. I'll use this with my autocollimator to check that I have things at right angles.

[nateman_doo]

I So I don't think it's localized imperfections in the surface giving me problems... Then I tried dusting the surface. I had previously cleaned it, but it's possible stuff had settled on it since then.

Seriously? Dust throws off your measurement? Your thread blew my mind so much I started doing what my 3 year old does and just flip through looking at the pictures. The finite analysis are something I would like to do one day, but this is NASA grade. Anything were dust/FOD throws off a measurement is not something done at home, but in a lab.

Good thing im not looking for that kind of accuracy in my build... LOTS of dust in my house

[jsheerin]

Glad you enjoyed it. I'm hoping I can get back to working on this eventually, but with where I've moved to it's going to have to be scaled back.

[jsheerin]

I've been thinking about this machine a bit more. I still have lots of time before I can do anything on it besides design, but I found an interesting newer design from Mori Seiki. See:
http://www.senga-eng.com/pdf/Catalog...%20DCG%20J.pdf

[Winterwatchers]

I have been looking at the mori seki design for some time and thought about building one based on that design, however my pocket isn't deep enough for it... yet. Its a beast of a design reserved for ultra high precision.

[andypugh]

Its a beast of a design reserved for ultra high precision.

I wonder if it is more or less rigid than a hexapod like this:
Octahedral Hexapod Machine | Flickr - Photo Sharing!
In some ways the hexapod design is easier to make, you just need to get lucky and find a cheap set of 6 matching servo-hydraulic struts. :-)

[jsheerin]

Holy thread resurrection, Batman!

I won't say I've been working on this project again, but I have been thinking about it off and on. I still have basically the same set of problems I'm trying to solve - I want a mill, I can't fit a commercial mill where I am, and I have a bunch of mill parts. However now add to the list: work keeps me really busy (and has for the last 4.5 years), and I don't own the place I live so anything I build has to be somewhat mobile. I also still don't really like any of the small mill options that are commercially available, so I still would like to build my own mill. Anyway, to more interesting things:

If you go back to post 298 where I was talking about trying to get the measurement capability to measure the flatness of my straight edge in order to correct it to be flat enough to replicate hydrostatic bearing surfaces off of, I was basically missing my measurement accuracy / repeatability goal by about a factor of 10. I was reading a bit more about laser interferometry, and I think I can use my laser interferometer to measure tilt by just adding one more retroreflector to my setup. In the picture, I have the interferometer, beam bender, and lower retroreflector - just need the top reflector and then a block to hold both reflectors a set distance apart and a block to hold the beam bender a matching distance above the interferometer. Then I end up measuring the tilt of the reflector block instead of the distance of a single reflector from the interferometer. The interferometer does this much more precisely / accurately than the autocollimator does from what I've been reading and calculating. It also doesn't look like the tolerances I would need to hold on the height of the reflectors would be that critical. If I wanted to increase accuracy after I built that block, I think I could calibrate it by measuring a known height block under one end of the reflector base. I could measure the height of the block using the interferometer first. I think this would get me to my measurement accuracy goal when measuring flatness.

[Hezz]

I'm kind of in the same boat. But I think your best solution is to design a custom mill that is semi mobile and will fit in your space. Either with steel or epoxy granite or a combination of the two. You could make a nice machine either way. One thing you could do is design it as a semi heavy roll around enclosed mill with pads on the bottom designed for lifting with a pallet jack or small fork lift to move it when necessary. I think with this approach it is more likely to have a design approach of weighing 1-3 tons instead of the heavier weights of commercial machines. However, with careful design of the structure you still should be able to attain a machine with pretty high repeatability and tolerances.

One approach that I have thought about is to make a small EG machine structure that fits inside of a self contained steel roll around space frame which is very stiff. And then the EG structure is bolted to the space frame to reinforce it and make it stiffer. The outer space frame is then skinned with aluminum sheeting to become the containment structure for the coolant system.

[jsheerin]

I got tired of not having a mill and perpetual planning to make a mill, so I caved and just bought a bench top mill, a Precision Matthews PM-30. This is a ~500lb bench top mill with an R8 spindle, variable speed belt drive motor, and it's the biggest one I thought I could easily get delivered and eventually move in my truck using just an engine crane in the future. Here's a link: http://www.precisionmatthews.com/shop/pm-30mv/

My plan is to use the mill to make the parts needed to convert it to cnc control. I bought Granite Device Argon drives to run some of the Fanuc servo motors I had sitting in storage. Link: https://granitedevices.com/digital-servo-drive-argon/ I'm going to try to use two 0S motors on the x and y axes and a 5S on the z simply because those are the smallest motors I had kicking around. The 0S is a 750W motor and the 5S is a 900W motor. They have a mix of 2000 or 2500 pulse per rev incremental encoders on them. They are probably overkill for this machine, but as long as I can physically get them mounted (which I think I can - I've quickly mocked them up in CAD so far), I'll use them. I tested all the motors with one of the drives and they all seem to work fine. It was also pretty easy to tune the servo loop with just a motor sitting on the bench. Obviously I'll have to retune once the motors are on the machine. Part of the process of tuning the motors in the GD software tool is inputting the coil resistance and inductance of the motor. That made me realize that the inductance of the motor will also control how fast current can rise in the coils given a max voltage from the drive which will control how much force can be created vs time and thus how much acceleration. Checking various AC servos that I have, the smaller motors were lower inductance, larger motors higher which makes sense - bigger coils in the bigger motors. However comparing the Fanuc motors to some smaller Yaskawa motors I also have, the Yaskawa motors have higher inductance so the Fanucs should perform a bit better. I might use one of the Yaskawas on a harmonic drive I have for a 4th axis, but that's down the road.

Anyway, I now have the mill sitting on the floor in my garage along with a pile of steel tubing that I'm part way through cutting up to make a stand from. After getting the mill mounted on that I'll need to do a bit more disassembling and measuring to figure out what size screws I need. I'll see if any of the ones I have will fit, but I'm guessing I'll have to buy at least two for the X and Y axes. Then I'll mill some motor mounting / thrust bearing blocks. Eventually I plan to add a Tormach style quick change tooling system and as already mentioned a 4th axis. I've been playing with Fusion 360 which will do cam for a 4th axis, so that could be fun. I don't really plan to make a lot of future posts here on this process because it's so far away from building a mill from scratch, but I thought some of you who were previously following along might enjoy seeing what I've decided to do for now with some of the parts I had accumulated.

[jsheerin]

I built a stand for the mill over the last two weeks. It's 3" square tubing, 3/16" wall thickness with a 1/8" plate shelf and a 1/2" plate top which seems to be fairly flat after I finished welding it on. It should weigh about 250 pounds. I put a lifting lug at the center of gravity so it's easy to move around with my engine crane. I put the front legs closer together so my engine crane, which has angled legs, can get close enough to easily set the mill down on it. I just need to paint it.

[mwmkravchenko]

Looks like a great welding job John. And a nice and stable base for your mill. Can't overestimate the benefits of a stable platform for a mill if you want quality machine work. Vibration is a huge problem in many DIY setups.

[jsheerin]

I got the mill mounted on the stand and leveled, wired it up, and made some test cuts. I modified a steel block for a new quick change tool post on my lathe and then did some drilling and tapping on a bracket to hold a VFD (also for the lathe). The mill seems to work pretty well. Drilling is very nice compared to using a hand drill. I have the mill mounted a bit low for using manually, but for cnc it should be great. Then I started pulling apart the machine to measure and plan my conversion. The grinding job on the saddle ways is a bit lack luster - they didn't actually smooth out the complete surface of the dovetails on the saddle. I'm debating scraping in the dovetails. I have a dovetail straight edge although I should probably do a bit of practice first before I dive into working on the saddle. I also want to grind some slots for oil distribution on the ways before I put it back together.

[mwmkravchenko]

Looks like a nice sturdy little Milling machine.

You going to do a zig zag pattern to distribute the oil?

[jsheerin]

Yes, I was thinking of a zig zag pattern.

[jsheerin]

Here's the 5 year update on my mill project. The take away is 'life gets in the way'. But over the last 5 years, I learned (kind of) how to hand scrape, built a lot of scraping and measuring tools and started scraping in my mill. Then I moved again at the end of 2022 and all my stuff is back in storage. Here are some pictures I pulled out of my photo album.

[jsheerin]

More test fixtures and scraping setups. I need to get a bigger surface plate if I want to scrape in the top of the table and possibly the face of the column (I could probably do the column without it, but it would make things easier). I also need to either make some new gibs or build up one of the surfaces on each one with turcite or something similar or do the same to some of the other moving surfaces.

[mwmkravchenko]

John you are the first person I've seen do the most obvious thing to me on a raw straight edge. Use a flycutter! Complete sweep and beautiful finish. You put some serious time into scraping. Seem to have the hang of it!

I was out of a shop from 1999 to 2004. So I do know what that feels like. It sucks!

Mark

[mwmkravchenko]

More test fixtures and scraping setups. I need to get a bigger surface plate if I want to scrape in the top of the table and possibly the face of the column (I could probably do the column without it, but it would make things easier). I also need to either make some new gibs or build up one of the surfaces on each one with turcite or something similar or do the same to some of the other moving surfaces.

The attention to the underside of the milling itself is admirable. Everything starts at the foundation. If it a kilter nothing above it will ever be stable.

Turcite is a great idea. although much softer and easier to scrape, it's also really easy to bugger up for the same reasons.

I have a nice Large Colchester Mascot 1600 that I have to grind and scrape in. Including new Turcite. Theres near 0.012 wear on the ways. It will be my adventure in this machine rebuilding business.

Mark

[jsheerin]

Thanks Mark. Unfortunately I didn't start off scraping the bottom of the mill - I thought it was sitting flat on the stand I made, but either I was wrong or something shifted / warped / relaxed along the way and I eventually noticed I was warping the base when I tightened the mounting bolts. So I went back and made the mill fit the stand. It doesn't have much contact along the back edge but I got most of the sides and front to fit and the little platform under the front of the column and it's way better than where it started. Then I had to go back and touch up some of the scraping I had already done on the base. That was a bit demoralizing but did go faster the second time around.

Fly cutting the straight edge worked quite well. I don't remember it taking too much work to get it flat after milling.