Advice/recommendations about new equipment for our machine shop

[Joe G.]

Originally Posted by pipsqueaker

Thanks for the replies everybody! People have been saying that the laser cutter's out of budget so let's throw that away, i knew it was a luxury anyways.

On the other hand, I'm interested in why you guys recommend the lathe so strongly? My understanding of lathes is that they're good for making shafts and things, but why does that give more utility than a 3-axis mill? I might not have mentioned this but we want to manufacture custom parts with these new machines, and it seems like a 3 axis mill would be much more useful/versatile than a lathe, which seems sort of limited in what it can do

To put it simply, a lathe gives you freedom, moreso than a mill. It singlehandedly took our team away from being held at the mercy of what's available COTs, and gave us the ability to make custom parts and mechanisms tailored to our robots. Sometimes it's things that we flat out couldn't have done before, other times it lets us to something in a cheaper, or faster, or more space or weight efficient manner, which pays dividends elsewhere on the robot. Some specific examples:

-Spacers, standoffs, etc. We make every single spacer we use on our lathe out of raw delrin (occasionally aluminum). We can make them to any length, diameter, and center bore that we want, and save tons of money over COTs spacers like the vexpro hex ones. Our students can hold tolerances within a few thousandths of an inch on our small machine, a more rigid one will give you .001" or better. Being able to make a precision spacer any length we want allow other shaft parts to dictate the exact length of the spacer, and makes designing custom gearboxes, drivetrains, and other mechanisms much easier. The spacers tend to end up ugly lengths, but that's okay, you only have to make them once. And the precision really helps solidify our robots, and allows us to use axial-sensitive parts that we wouldn't have attempted in custom systems before, like small pitch chain and belting. It doesn't seem like that big a deal until you've done it, this alone has dramatically improved our robot quality.

-Custom shafting. The lathe made widespread use of hex shafting practical for us, a huge improvement over poorly retained keys causing failures at the worst possible times. Turn down each end of a half inch hex to half inch round, save a few bucks on hex bearings (which REALLY add up, and are a pain to align), and get foolproof axle retention in the process. Like spacers, you can get it cut to precision length on the lathe, and allow other components to dictate shaft length rather than trying to design around a COTs part. We can also add a 1/4" round to the end of any shaft for encoder mounting, and cut our own retaining ring grooves, which save tons of weight and space over shaft collars. Even for dead axles, it's helpful. We ordered about 30 feet of 3/8" aluminum round for dead axle applications, all of which turned out to be a few thousandths oversize and too big to fit our bearings. We shaved it all down on the lathe, rather than giving it to some poor freshman with a file. We've also made oversized shafts for some extremely high stress applications, in sizes you typically just can't get from andymark and the like.

-Custom boring. Need a bearing bore in a gear that doesn't have it? Chuck it in the lathe and bore it out! We've also pocketed out nearly the entire body of a few gears on the lathe, sometimes for weight savings, other times to nest the mounting of an adjacent part in in space critical spots.

-Custom pulleys. 1687's winch based catapult last year would not have been possible to fit in the robot the way it did without the ability to turn a custom winch and two pulley wheels on the lathe. We've also done polycord grooves cut right into ABS or PVC for conveyors.

-Stepped diameter shafting, allowing what would previously have been a pile of loosely connected parts to become one rigid piece.

-Knurled inserts. Before VexPro, we made press-fit inserts for Colson wheels.

And so on. Being able to carve anything roughly round out to precision measurements just changes the way you think about robot design, it opens up so many doors. I strongly recommend that all teams get one as their very first "serious" equipment purchase beyond the very basics.

A few other specific examples of parts made for this year's robot, which I couldn't imagine doing without a lathe:

• Dozens of spacers and hex shafts cut to precision lengths
• Custom standoffs which allowed us to nest elevator bearings inside of slots in our box frame lift system
• A hexagonal sleeve coupler, which enabled super-quick removal and replacement of our drive gearbox.
• A hex/round hybrid shaft which enabled a reverted gearing system for our elevator tilt mechanism, a HUGE space saver.
• A stepped pin used to lock up our elevator.
• Shaving down the outer diameter of dozens of bolts in our elevator ever so slightly, allowing them to press into bearings.
  

A lot of these systems incorporated milled parts as well, but we could have done a whole lot better at reproducing these without the mill. We couldn't have come close to these turned parts without our lathe. In addition, we've found that parts for these custom systems which absolutely demand a mill are generally few in number, very complex, form the big "obvious" parts of an assembly, and as a result, tend to be very well planned out. The turned parts, by comparison, quickly add up in a huge backlog, are generally individually pretty simple despite the precision demanded, and are occasionally forgotten until assembly time. The former lends itself very well to being produced by a manufacturing sponsor, the latter demands an inhouse machine.

I would also argue that a pretty minimalist lathe would adequately serve for most of the jobs above, whereas with a mill, you need a pretty signifigant chunk of machinery to make many of the parts you likely have in mind with a high level of precision. For three years now, nearly every one of our turned parts has been made on an old Atlas Craftsman 618 benchtop lathe, which has performed like a champ for us.

[MrForbes]

Think about ways to make a gearbox. One simple method is to use plates to hold flanged bearings, and standoffs to space them apart. To make this you need to be able to drill holes accurately. A mill makes this easier, but it can be done with a drill press, if you learn some craftsmanship.

The lathe lets you make the shafts to make this gearbox work.

[BeardyMentor]

A lathe is going to be the most useful thing by a long margin. If it is possible for you, try finding out if any local colleges have a surplus auctions or shops. I have seen colleges sell off machine tools that are in perfect working order for pennies on the dollar. Out side of that, look for a Hardinge or Southbend on Craigslist. Cleaning and rehabilitating a used lathe is a great off season project and a good way to familiarize students and mentors with the parts and workings of your new tool. If neither of these is an option Grizzly and JET have some affordable options that are just fine for an FRC team. For something in your price range, make sure you budget 20-50% of your purchase price for tooling and support equipment.

A lathe and a High quality drill press will take you a really long way. A small mill would also be a nice thing. A laser cutter less than $50,000 is going to be a bit disappointing for an FRC team. They will have small working areas, and very limited materials and thicknesses they can cut.

[mman1506]

It's also quite a bit easier to find a waterjet/laser sponsor than a lathe sponsor due to the employee labor involved with running a lathe.

[The other Gabe]

on low budget, a lathe is probably the best way to spend your money. It's really nice to have a mill, but you really don't need one to make a good robot... you may want to look into getting a waterjet sponsor or something along those lines: then you can design the parts you need and get them done for you (and maybe get a nice tour as well).

if you end up getting a mill, Tormach is good

[themccannman]

The first things you should have in your shop in terms of heavy machinery are a good lathe, and then a good mill. Everything else is dependant on your requirements but I've never seen a robotics team that wouldn't benefit the most from having a quality one of each of these first. Next up would probably be a CNC router so that you can do all of your own gussets/sheet metal in house when you need them.

[protoserge]

What does your CAD look like for your current swerve modules? By the time you get tooling, you'll likely exceed your budget. I would venture to say you will want CNC for making parts, especially once you start removing a lot of material at angles and arcs.

The HF mini mill is too light duty in my opinion. Putting that $500 into another machine would be better spent. It most likely doesn't have enough working envelope to do what you want and will be even more limited when you want to add workholding equipment.

One thing you can do is build a small (2' x 2' or 2' x 4') CNC router table with a water cooled spindle. You can accomplish this within your budget using extrusions, ballscrews, and linear recirculating ball rails. Don't consider the Shapeoko. It isn't robust enough to be considered for doing work in metal.

If you can find a used South Bend / Atlas / Clausing / LeBlond / lathe without uneven bed wear (don't worry about backlash at this point), you can get an excellent platform at a very reasonable cost. Something like a South Bend 9" or 10" (10L "heavy ten" being a heavier duty lathe with larger bore) would be good. A Grizzly G0602 or PM1127 would be a good new machine for FRC purposes.

Buy a bench grinder and high speed tool blanks and you will have most all the lathe tooling you will ever need. You can use the lathe to make boring bars for holding small HSS cutters. You can use a lathe to make keyways. You can use a lathe to make custom threaded rods and nuts. You can use a lathe to make a lathe! Learn how to use a 4 jaw chuck properly and you'll just have fun with seeing how round you can make something. Don't forget to get dial indicators and magnetic bases.

As for mills, you want a knee mill or, at the very least an RF-45/IH (Industrial Hobbies) such as the PM940 or Charter Oak 12z. DRO is nice.

For your current budget, I would think that if you could only get one machine, it would be a lathe. Combined with a hand drill to make mounting points for work holding in a lathe milling attachment, you could actually do most of the precision machining of your swerve gearboxes with an end mill in the lathe in a collet. You can also chuck your sheet/plate part in a 4-jaw and use a boring bar to make bearing pockets.

Find sponsors - you'll be amazed at what you can get! I see you're already sponsored by Pratt & Whitney and UTC. I would bet that they can get some contacts if you need them! They may even have some machines sitting around unused in tool rooms that need a good home (and a tax writeoff).

[Jared]

Originally Posted by stinglikeabee

Find sponsors - you'll be amazed at what you can get! I see you're already sponsored by Pratt & Whitney and UTC. I would bet that they can get some contacts if you need them! They may even have some machines sitting around unused in tool rooms that need a good home (and a tax writeoff).

Also see if you can get some of their older tooling. We have quite a few reground end mills in our shop that are really weird diameters but still work nicely.

[JesseK]

Originally Posted by stinglikeabee

One thing you can do is build a small (2' x 2' or 2' x 4') CNC router table with a water cooled spindle. You can accomplish this within your budget using extrusions, ballscrews, and linear recirculating ball rails. Don't consider the Shapeoko. It isn't robust enough to be considered for doing work in metal.

Last summer we invested in a DIY 4'x4' (ish...) CNC router. We spent the summer/fall building/integrating it. We then decided to replace the MDF it had with 1" thick waterjet aluminum, and man that upped its precision at a reasonable expense. We finally got it calibrated in Week 2 with first production parts in Week 3. It changed everything with how we design now versus how we did in Week 1 & 2. Precision plates that used to take an hour now take 20 minutes and are lighter. Swapping out different materials is a breeze - take one 4'x4' sheet off and put another on. It really cuts down on material waste and saves a lot of time.

We manually apply cooling fluid via paint brush and manually vacuum as the piece is cut rather than after. We're able to reliably get 1/16" depth clean cuts per pass, so 1/8" stuff is done in no time.

Point is, with the right know-how it's easy to not have to get extremely fancy with a CNC.

P.S. no one is a machinist by trade on our team. It took 1 mentor's drive and the rest of us to follow in order to pull it off.

[FrankJ]

To be honest I am not sure which is more important, a mill or a lathe. I think our mill actually had more run time this year, but the lathe did things that would have been difficult to do any other way.

Don't let the search for perfection get in the way of usable. The referenced HF lathe is in a class of 4xX mini lathe, most all are from one or two factories in China. While not production level quality, they are quiet capable & have a large support group to learn from. I would get one that has a longer bed if going that route. Get the heaviest mill you can afford / have space for. Knee mills are nice, but other styles work too. Stay away from the small ones that use a chuck rather than a collet. Look at this as a multi-year project. The tooling often will transfer to the bigger machines. We recently added a Project Lead the Way CNC that was gathering dust in our school. It is smaller than I would like, but it will machine the pieces that will fit on it. It is just slower that a heavier industrial version.

Youtube is a great resource for learning how to (&not to use these machines).
mrpete222 is a good place to start.

[GeeTwo]

Any ideas on how to learn to use these tools should we buy them? Whether a lathe, a CNC, milling machine, or whatever, we don't have any mentors with experience using these tools.
Example: Off the top of my head, I can't figure out what I would do with a lathe that I couldn't do with a drill press and a bit of patience. Based on the posts, there's obviously something else they're good for.
Are there any on-line tutorials that cover use of these tools, or will each turn into a high-priced coat rack unless we find a local person who can teach us to operate it?

[Joe G.]

Originally Posted by GeeTwo

Any ideas on how to learn to use these tools should we buy them? Whether a lathe, a CNC, milling machine, or whatever, we don't have any mentors with experience using these tools.
Example: Off the top of my head, I can't figure out what I would do with a lathe that I couldn't do with a drill press and a bit of patience. Based on the posts, there's obviously something else they're good for.
Are there any on-line tutorials that cover use of these tools, or will each turn into a high-priced coat rack unless we find a local person who can teach us to operate it?

The internet has a lot of great resources, from documentation, to video tutorials. For example, here's the go-to document for learning to use a lathe for years and years: South Bend's How To Run a Lathe. This plus documentation specific to your machine should get you going.

[75vs1885]

What ever you get make sure that it is really useful, having a CNC is really nice, If we had it running earlier then I wouldn't have had some much trouble with the drivetrain( I messed up on it like 3 times using a horizontal mill :l ). Also, don't go over board and spend a ton of money, there are plenty of good brands which are cheaper. ie: Grizzly