FRC & different types of CNC

If it were my shop:

  • Frisbee the waterjet. Extremely maintenance and overhead intensive, offers precious few advantages over plasma+laser+router.
  • 4x8 Routers are flexible machines that are generally easy to train with.
  • Lasers are awesome, very low-risk to train on, and flexible. Just more expensive than routers.
  • CNC plasma can, and should, be used more commonly in FRC. They can rip through aluminum quickly and with ‘good enough’ tolerances to be worked around. For example, these 1/4in arms for our 2014 robot were plasma cut in <6min each.

Every bit of aluminum sheet on that robot was at least blanked-out on the plasma cutter.

It is quite time-effective to plasma-cut blanks and then dust them in on a router or mill. Cut the precision features +0.02 on the plasma table, then skim them to size on a more precise machine.

  • I would love to get a CNC press brake as part of outfitting a new shop. It’s a machine that is extremely common in industry and I’ve never, ever, seen one in a tech center or FRC workshop. Then you’ll be off to the races making sheet metal parts until you’re blue in the face.
8 Likes

Hard no from CTE lead for contamination reasons philso stated. That said, the Shop Sabre PRO 404 model that is around the 4x4 size (somewhat larger) looks really nice, and may be the one that will be selected for metal.

Yes, that’s the one. I just searched & read through several threads. I also took a closer look at the website, and with the thinner aluminum material we’re most likely to use there seem to be problems. And the machine cuts much slower than a router. I’m going to recommend skipping this one - not enough ROI & risk of wasted time.

Thanks again all. Recommendation seems confirmed: start with laser cutter the school already has & invest in a nice router. (There will be a brake as well for bending.) Next after that: CNC mill with reasonably long x axis. Plasma has potential (thanks for explaining how you use it JamesCH95), but is lower priority starting out. Small waterjet seems cool but too risky.

4 Likes

We have a CNC router, mill, plasma cutter, and a lathe.

If I were outfitting a shop specifically for FRC, I would start with a decent quality router (probably a Velox). My next purchase would be… a second CNC router. Router throughput is a bottleneck for us.

We get a decent amount of use from our CNC mill, and manufactured our 2016-18 robots on it almost entirely. It’s versatile, but it’s a huge pain for routing tubes because of the limited travel. A 4 foot bar with features on all sides requires 12 separate operations. It’s useful for “chunkier” parts like bearing blocks, but less useful for plates and bars.

I see the value in a plasma cutter, although ours is a bit hackey (It wasn’t originally designed as a CNC) and set up for heavier steel, so we don’t use it much for the robot. A well maintained, correctly tuned plasma cutter could be a major time savor for parts like what @JamesCH95 posted above.

The CNC lathe is a gimmick. The most useful thing ours has is a program that cuts e-clip grooves into the ends of axles. We’ve tried some fancy lathe parts with it, but none of it is really necessary.

Ditto what James says about the waterjet, too. Too much maintenance, too much overhead.

2 Likes

Agreed. I’ve worked with the waterjet 67 has access to at GM and can say with confidence a school / FRC team should not own one of these. The tips alone are on the order of $100 and they get broken regularly, . The maintenance cost (and labor) is substantial and it makes a complete mess. It’s likely cheaper to just pay someone to waterjet your parts for the next 20 FRC seasons. You’re better off finding a sponsor with one who has fast turn around time. The desktop versions have such a slow cutting speed you’ll be better off with a router or plasma cutter.

A CNC Lathe is probably also on the list of not very useful tools for FRC. 67 used one in '15 and '18 to make rollers, but these could easily be 3D printed (as they were in '19). I could see it being convenient for snap ring groves, albeit not a very big time saver considering how easy they are with a DRO.

I will disagree with James on the plasma cutter, I don’t want to diminish the fact that 95 does great work with a plasma cutter, but they are sponsored by Hypertherm and have had some decent support from them to make their setup top notch. There are other on here that have had less positive experiences with plasma cutters. 20 thou is perfectly fine clearance for a majority of FRC parts, but the bottom line here is that it can’t replace a router since it can’t do gearbox plates and it doesn’t do anything a router can’t (maybe cut steel quickly, but oh well). I’d rather own a second smaller router that can do more gearbox plates simultaneously and takes up significantly less room than a machine that requires I do finishing operations on another machine (which adds significant setup time).

Must have CNC tools:

  • Router - most parts in FRC are 2-2.5D, a large router with enough build area for these parts is great.
  • 3D Printer - you don’t need a Markforged, but you do need this capability and they’re cheap enough now that they can pay for themselves just by not buying COTS parts.

Must have other tools:
As others have mentioned there are some shop staples that you should have from the sounds of it your school shop is well equipped and has most of these: manual lathe (preferably with DRO), band saw, cutoff / miter saw, arbor press, drill press, belt sander, drills, etc.

Nice to have CNC tools:

  • Laser cutter - I’ve never had access to one but plywood prototypes I see come out of teams that do make me super jealous. You can do these prototypes on a router but the laser cutter makes it basically insert file receive parts as opposed to needing to use extensive CAM software.
  • 2 or 3 axis CNC Mill - This gives a lot more stiffness than a router and gives a lot more flexibility to do quick manual operations or billet parts than a router does. You don’t NEED 3 axis here, I haven’t found anything I’ve ever needed a 3 axis CNC mill to make that couldn’t be accomplished with a manual Z motion between XY paths. That being said not needing to babysit a part is pretty convenient.

Nice to have other tools:

  • Sheet metal brake - This expands design flexibility significantly since you can do some pretty cool stuff with a sheet metal brake. If you have the money a pneumatic or hydraulic press brake is fantastic, but a manual one is sufficient for thin aluminum plates. I’ve never worked with a CNC one, but I imagine the setup steps involved are not insignificant and most people in FRC that have access to one have a sponsor that does the layout and bending for them. Would love to hear from someone who has more knowledge on the CNC brakes.
13 Likes

These are mostly true facts.

While we did get it dialed in well, it was still a mediocre table at best.

If one decides to make parts with a CNC plasma cutter it is a good idea to design for that tool from the start. I’m going to stop myself there to avoid going down a huge rabbit hole!

3 Likes

This is a reasonable plan.

If it were me, I would prioritize machines something like this for FRC uses:

  1. CNC Mill
  2. 3D Printer
  3. CNC Router
  4. Lathe (CNC optional, but at least DRO preferably)
  5. Basically anything else.

TBH, for FRC uses if you have a good CNC router, a plasma cutter becomes redundant. While, a plasma cutter is probably faster than a CNC router for most workloads, it’s basically inferior in every other way (No real depth-of-cut control, limits on what kind of materials you can use on it, lots of post-processing of parts, etc.).

I’d also personally avoid waterjets because of the mess and cost of operation. IMO, a better option would be to get a sponsor with a waterjet, and then see if you can “visit” them to show the kids how it works.

My team has access to a few different different CNCs, but we primarily use these:

  • Omio X8-2200 - Used mainly for cutting out sheets of material, and some minor mill work. Initially I was pushing to get a larger CNC Mill, but due to cost and space we settled for the X8. That said, It’s been quite versatile for our uses. In a perfect world, it would have been nice to be able to have a Router large enough to stick a full 4’x8’ sheet of material in it so as to avoid cutting the sheets down first, but it hasn’t been a show-stopper for anything so far. If I were looking for a larger machine, Velox CNC or one of the kits from AvidCNC (formerly CNCRouterParts) would probably be what I’d look at (and both are made in USA, as far as I’m aware).

  • Tormach Personal CNC 770 - Used for smaller parts that require clamping (primarily cutting lightening patterns in aluminum tube). It’s on the smaller side (only 14" of X travel), but it usually gets the job done (and since we usually use it for tubing, it’s pretty easy to just reposition longer tube and keep cutting). Much higher precision than the Omio too, not to mention the more powerful spindle.

Beyond traditional CNCs, we have a number of 3D printers that we’ve been using more extensively in recent years (to the point that basically our entire turret assembly was 3D printed this past year). In particular, we mostly use a Markforged Onyx One for strength-critical parts, and a number of Prusa i3 MK3’s for less strength-critical parts.
We also use a LittleMachineShop “HiTorque” Bench Mill and “HiTorque” Bench Lathe mostly for training and smaller, simple parts; as well as an old Bridgeport knee mill we use on occasion for larger parts. All three of these machines have DRO’s

In addition to equipment the team owns, we also have access to a HAAS TM-2P CNC Mill that the school’s machine shop just got this summer (don’t ask me how), as well as a CNC Plasma Cutter (not sure the make/model but it’s roughly the same size as our Omio X8, maybe a bit larger).

We also have sponsors with a variety of machining capabilities, from sheet metal fabrication (we use this for our robots chassis) to waterjet cutting.

2 Likes

One thing that hasn’t been mentioned about a plasma is that unless you are running a water table, which really isn’t advisable in cutting aluminum, the smoke and particulate is absolutely astounding. Unless you have a heck of a ventilation system, you essentially need to use it outdoors. It cuts crazy fast, but your cut quality is not fantastic, and you won’t be making +/-.005" parts. @chrisrin I sent a reply to your PM, not sure if you got it. Budget is everything here. The types of equipment you can get on $5k, $15k, $30k, and $100k budgets varies greatly and my recommendations would be very different at each of those budget levels.

1 Like

Why CNC mill first? I would recommend a 3D printer. The spectrum lists of things to buy is what I recommend to most. While a router is limited it’s also a LOT cheaper than a CNC mill and most teams can afford a router over a mill. Also, I’d get a manual mill and lathe even before a CNC router. Despite what some people say you can succeed without CNC tools, although they do make it easier, once you have succeeded with all Manual I think you’re able to integrate CNC tools a lot better.

7 Likes

Agreed on 3D printer being at/near the top. We have a couple Markforged + a couple older-but-functional Makerbots for other materials, so we’re pretty good on that front. As far as general recommendations though, agree.

@sanddrag, thanks for the additional insight into smoke, mess, and imprecision with plasma. As far as budget, I don’t know the school’s budget precisely (& don’t know if it’s even been set)… There was a request for input from mentors about what types of machines are useful along with a statement that the school would like to expose students to a variety of things, which is why CNC plasma and small waterjet were on the list. I have shared the feedback that those things may not really have strong ROI & that CNC router & mill (along with existing laser & 3D printers) should be the priorities.

Thanks again everybody. I hope the school follows the advice I was able to provide using your collective input.

1 Like

Another must in my books is a lathe with a DRO and an old Bridgeport mill with a DRO. If your CNC’s break down or don’t work the mill will work, ones from like the 20’s to 80’s are built like tanks and even ones from the 20’s still run like a dream today provided you take care of it.

2 Likes

I put the CNC mill first because it’s probably the best machine to do work you couldn’t otherwise do without one. A router is nice, and can be used for mill work to a degree, but in general (at least on my team) it’s really used more as a time-saver than anything else. Most of the parts we make on our router could have been made more “manually” (like cutting things out on a band saw), and if you need precision “touch-up” work you can use the CNC mill for that.

Basically there’s a lot more things you make exclusively with a CNC mill than things you can make exclusively with a CNC router.

Likewise, 3D printers are also very useful, but for FRC they generally tend to be used for less critical parts (usually they’re more of a weight-saving and/or time saving option). Of all the 3D printed parts I’ve made for our robots, I can’t think of any that couldn’t have been made using either a CNC mill or other more traditional methods. We choose to use the 3D printer for these parts because it’s faster (since you can start it and walk away), and because the resulting parts are much lighter. Obviously 3D printers can be used to make various shapes that would otherwise be impossible (or at least impractical) to make using normal manufacturing methods, but for the most part you don’t see FRC teams using this capability for the parts they’re printing.

I would certainly agree that manual tools are an obvious first choice for a team due to their generally lower costs. However, the original post referred specifically to CNC equipment, so I assume they already have some manual equipment and are looking for an “upgrade”.

I disagree with the idea that a CNC mill should come before a CNC router.

A mill may help you do “work you couldn’t otherwise do without one.” but that is not the most important factor to consider when purchasing a large machine. I would claim it is the degree of impact a machine will have on a team’s build process that is most important.

A mill may be able to provide unique capabilities (I don’t even completely agree with this but lets go along with it) but uniqueness does not equate to value. The CNC router is widely regarded as one of the best investments a robotics team can make because of the game-changing capabilities it brings. It truly opens the door to being able to quickly and accurately produce a design that no other machine provides. In a single season with a CNC router (speaking from experience) a team can go from barely being able to field a competitive robot to being able to really design to the challenge and produce a great robot. It is that impact a machine has that dictates it value and is why all across CD you will see recommendations for the router over the mill.

Additionally, you cite the router as being “used more as a time-saver than anything else,” except that is exactly what this team is looking for. They find themselves getting below average shop time and want to be able to better use it. Just another reason why CNC router> CNC mill

5 Likes

That’s a fair opinion.
Guess it just depends on how you approach the question.

1 Like

While I agree that a mill can do a lot of stuff. it’s not worth it for most teams, most teams don’t even consider it because ~80-90% (just a guess) can’t afford it or don’t have the room for it. A router and two or three 3D printers are more impactful to 90% of teams while being cheaper on the whole. When you say that

3D printers are used for less critical parts

That’s false from what I’ve seen and done myself, lots of teams 3D print pullies, and those are some of the most critical parts of their robot. getting into CNC is a task of huge magnitude part of why this thread exists and throwing someone in the deep end with a CNC mill is just not the right way. and saying that you could do stuff with a CNC mill that you couldn’t do otherwise is debatable at best and at worst a waste of ~30K for a decent starter CNC mill.

1 Like

Yes, thanks for going back to the original post. Time shortage is a huge challenge for our specific team. With summers mostly off & just one meeting every week or two before January (& two or three a week from there), we need to be really selective about what we do and don’t do. We have a mostly student-led training program for using manual tools including hand tools, drill press, band saw, manual lathe, and manual mill. Sometimes we partner with a local makerspace to get more lathe/mill training time in the fall, which helps. And we have a small group of 3D printing enthusiasts that manage & train people on our printers. Given these factors & the input from you folks here, I’m quite confident that (for us) adding a CNC router + learning the skills to use it are going to be the best capability to add.

1 Like

Be advised: Adding a CNC won’t necessarily make you faster… there’s a limit to how fast it can churn out parts. We’ve had to remind our CAD crew more than once that “no, you don’t have to run EVERYTHING in the CNC” due to time limits (we run parts for us, and for other local teams).

That said, it’ll probably still be faster than just a manual mill, and I’d say probably faster than a CNC mill at the smaller parts. We try to shunt the less-critical/smaller/non-lightweighted parts over to the Bridgeport as much as we can to keep our router for the larger/batched/lightweighted parts. We’re still running the Velox pretty much all of build season…

6 Likes

Another thought: depending on the SES of your community you may find that some of your students own their own 3-D printers. I think we have 4 or 5 kids who have one at home that we can use.

1 Like

There’s definitely a limit to how many parts one can crank through a single Omio in a season. For us, that limit was almost perfectly two robots of nearly exclusively CNC’d parts. The exact volume scaled the second season due to efficiency/experience gains (learning how to run a full sheet of parts in one run was huge for us in 2020). We do intentionally try to design simpler robots so YMMV.

Depending on exactly how time-constrained the OP is they’ll want to be careful in sizing jobs for their meeting times. This is definitely a place where a longer meeting is nice-- we found ourselves with pretty much one shot to do a large operation (like a full sheet cut) on a weekday night. As you gain experience you’ll also learn what corners you can cut for speed and which ones you’ve gotta hold fast on. A lot of FRC parts are not (or should not be) high tolerance parts.

1 Like

Disagree. CNC Router is cheaper than a manual mill and makes more complex parts faster. There are very few parts most teams make that are more than 2.5D or that can’t be 2.5D if you design for the tools you have. We don’t have a mill on 3538 and it hasn’t really held us back. Omio does 98% of the non-lathe parts we need. Not sure why knowledge of a manual mill would help your team use a router better. Work flow is different because CAD / CAM is required. Speeds and feeds are different. Part fixturing is different. Drawing’s aren’t as necessary (we don’t do them for most CNC parts). Duplicate parts are just as much work as the original (unless you’re good at setting up fixtures each step of the way). That being said I think a manual mill is next on our purchase list.

I went through our workflow sheet to see how many unique parts required each tool this year (not quantity of parts, just unique parts), not everything ends up in this sheet, since we get lazy / panicked at the end of the season, but it’s mostly there:
Router parts: 49
Lathe parts: 20
3DP parts: 23
Hand fab: 5 (mostly cut tubes to length)
Beg and Borrow: 1 part, our steel flywheel we needed to go to a sponsor’s CNC. Probably could have done this in house with a different method, but it seemed easier to just go to the sponsor.

5 Likes