971 CNC Router

[cxcad]

Looking through 971's photo album (I'm a super fan, lol), I saw that they have made their own CNC router. And some questions popped up.

• Why would a team do such a thing when so many CNC routers are out there? Cost of a custom one?
• What type of resources like books, manuals, designs were used to build such a router? I know there's tons of resources on building ones that can route through wood but very few on aluminum specific ones
• How much skill does it take to design such a router? It would be cool to have the students work on it as an off season project.

....Any plans to share the design (wishful thinking)....

[asid61]

For your third question:
I think 971 did theirs all custom will ball screws and the like IIRC, but for the typical teams there's a lot of options out there. The Shapeoko 3 is like $1,000 and can do aluminum (albeit slower than an industrial machine) and is just put-together. Other kits out there are anywhere from $500 to $10,000, and the skill required to build it varies.
However, most I've seen take almost no tools as they're built for hobbyists. The average FRC team should have few problems.

[Jacob Bendicksen]

I know 2471 has built their own if you wanted to reach out to them. They also just built their own powder-coating rig - I'm impressed.

[cadandcookies]

I can't speak for 971, but my college organization, GOFIRST, is considering building our own. Our reasons may be similar, but I haven't talked to 971. Our reasons are a combination of wanting something that will fit into our space and wanting something designed specifically for aluminum. Outside that, we also think the exercise of building a router will be an interesting and educational exercise.

[Travis Schuh]

Originally Posted by cxcad

Looking through 971's photo album (I'm a super fan, lol), I saw that they have made their own CNC router. And some questions popped up.

• Why would a team do such a thing when so many CNC routers are out there? Cost of a custom one?
• What type of resources like books, manuals, designs were used to build such a router? I know there's tons of resources on building ones that can route through wood but very few on aluminum specific ones
• How much skill does it take to design such a router? It would be cool to have the students work on it as an off season project.

....Any plans to share the design (wishful thinking)....

We decided to build a router after the 2014 competition season, and got it cutting parts this past August. We made the 3rd robot (the one we used at Chezy Champs) almost entirely on the router, and we were really pleased with the performance.

• When we started looking at routers, we decided pretty early on that the commercial routers on the market that we found within our budget all had design compromises that we were not a fan of. Our goal was to be able to have enough rigidity and precision to be able to cut bearing press fits in aluminum. This lead us to build our own using Chinese ball screws and linear bearings. This took a huge amount of time (I would estimate a 971 season or so of work), but I think the time invested payed off.
• We did some research of what other routers (and gantry mills) out there looked like, combined with a lot of design iterations and FEA. We analysed a simplified model of multiple router architectures using FEA to characterize what their deflections were. We went through a few different ideas before we found one we liked. Then we spent a bit of time working on getting the stiffens of that design up. We also got a lot of feedback from 973 and 1323 about their router experiences, including 973's custom build.
• Designing a router has a lot of parallels to designing a FRC bot (at least in our experience), and we were able to leverage a lot of this experience for designing the router. We did spent a lot of time on the design and fabrication of the router. We started meeting every weekend early in the 2014 summer, and continued right up until kickoff. Fabrication started around September and assembly around the end of November. We unfortunately barely ran out of time to get it running for the season, but were able to get it up during the 2015 summer in time to make the 3rd robot. The design and fabrication effort was done mostly by mentors because of the nature of the scope and because the kids had their own off season projects that they were working on that fit better with our team learning objectives. I hesitate to recommend designing and making the router we did as an off-season project for the kids, but I don't know anything about your team.

[cxcad]

Originally Posted by Travis Schuh

We decided to build a router after the 2014 competition season, and got it cutting parts this past August. We made the 3rd robot (the one we used at Chezy Champs) almost entirely on the router, and we were really pleased with the performance.

• When we started looking at routers, we decided pretty early on that the commercial routers on the market that we found within our budget all had design compromises that we were not a fan of. Our goal was to be able to have enough rigidity and precision to be able to cut bearing press fits in aluminum. This lead us to build our own using Chinese ball screws and linear bearings. This took a huge amount of time (I would estimate a 971 season or so of work), but I think the time invested payed off.
• We did some research of what other routers (and gantry mills) out there looked like, combined with a lot of design iterations and FEA. We analysed a simplified model of multiple router architectures using FEA to characterize what their deflections were. We went through a few different ideas before we found one we liked. Then we spent a bit of time working on getting the stiffens of that design up. We also got a lot of feedback from 973 and 1323 about their router experiences, including 973's custom build.
• Designing a router has a lot of parallels to designing a FRC bot (at least in our experience), and we were able to leverage a lot of this experience for designing the router. We did spent a lot of time on the design and fabrication of the router. We started meeting every weekend early in the 2014 summer, and continued right up until kickoff. Fabrication started around September and assembly around the end of November. We unfortunately barely ran out of time to get it running for the season, but were able to get it up during the 2015 summer in time to make the 3rd robot. The design and fabrication effort was done mostly by mentors because of the nature of the scope and because the kids had their own off season projects that they were working on that fit better with our team learning objectives. I hesitate to recommend designing and making the router we did as an off-season project for the kids, but I don't know anything about your team.

I had no idea about the scope of your project. The kids on the team are definitely not there yet. Why the router, when the team has so many sponsors? I guess lead time. Did 973 take a similar development process?

[kevincrispie]

Originally Posted by cxcad

I had no idea about the scope of your project. The kids on the team are definitely not there yet. Why the router, when the team has so many sponsors? I guess lead time. Did 973 take a similar development process?

Lead time is definitely a large part of it. For getting parts back from sponsors, often the best we can do for large quantities of parts is 2 weeks. With the router, that time is shortened substantially. It also allows us to raise the quality of our prototyping and allow for some more iteration during build season. Sponsors are great, but we don't get an unlimited amount of parts from them. Having a router allows us to make some parts, try them out, and then change things up if necessary. In the past, we've had to get a design as final as possible before sending parts out to be made, since we had limited ability to make substantial mods once the parts where shipped. We plan on continuing to work closely with sponsors to get some parts made, but we will start to make some in house. We also have a press brake that allows us to bend the sheetmetal parts once they come off the router.

I'm not as familiar with 973's development process, but we were certainly able to learn from the problems they ran into. They were a big help.

[Travis Schuh]

Originally Posted by cxcad

I had no idea about the scope of your project. The kids on the team are definitely not there yet. Why the router, when the team has so many sponsors? I guess lead time. Did 973 take a similar development process?

The router does a good job of complimenting our manufacturing sponsors skill sets. We still plan to have similar levels of sponsor involvement during the season, but we will be able to prototype better and iterate better after our initial manufacturing run. We also have trouble finding sponsors to make tube-stock parts, and the router will let us do these easily in house. Typical sponsor lead times for us are 2 to 3 weeks, so having something in house will be a huge help.

973 built a much simpler router based on some linear actuator modules that they got surpluses somewhere. You can ask Adam for more info.

[cxcad]

Originally Posted by Travis Schuh

The router does a good job of complimenting our manufacturing sponsors skill sets. We still plan to have similar levels of sponsor involvement during the season, but we will be able to prototype better and iterate better after our initial manufacturing run. We also have trouble finding sponsors to make tube-stock parts, and the router will let us do these easily in house. Typical sponsor lead times for us are 2 to 3 weeks, so having something in house will be a huge help.

973 built a much simpler router based on some linear actuator modules that they got surpluses somewhere. You can ask Adam for more info.

I'll talk to Adam. What advantages does your router have over the simpler one?

[Travis Schuh]

Originally Posted by cxcad

I'll talk to Adam. What advantages does your router have over the simpler one?

Having a machine with more rigidity and low backlash lets you make more precise cuts and aggressive cuts, so we can make better parts faster. Our preliminary testing shows that the machine is capable of holding tolerances needed for making FRC gearboxes directly on the machine without having to do any post machining like reaming the holes.

I don't know enough about other routers to say how this compares to something you could buy. We like nice things, and for us it was worth it to trade off time and effort to get a result that better fit our goals with the budget we had.

[josesantos]

Any chance of releasing the design notes or the CAD model? I'd particularly like to learn more about how you analyzed the different configurations and modified the design to get your desired rigidity.

[MattRain]

Any links to these pictures? Would like to see what we are talking about.

[Knufire]

Originally Posted by MattRain

Any links to these pictures? Would like to see what we are talking about.

https://picasaweb.google.com/1177698...87711612872482
https://picasaweb.google.com/1177698...87709375928370
https://picasaweb.google.com/1177698...87706482902946

[Travis Schuh]

Originally Posted by josesantos

Any chance of releasing the design notes or the CAD model? I'd particularly like to learn more about how you analyzed the different configurations and modified the design to get your desired rigidity.

Releasing something is definitely an option, I would have to look through the state of our documentation. Probably won't happen until after the season, as we have a lot of projects that we are working on right now.

[AustinSchuh]

To add to the fun, one of our mentors put together a video of the build process.

https://www.youtube.com/watch?v=m8s6-6CvS1E