I’m making mechatronic Dr Octopus arms (they’re gonna be pretty sweet), and I’m in the process of prototyping the segments. They’ll be roughly 2.5 cm in height, with a 2.5 cm gap from segment to segment, and they need to be attached securely, with smooth rotation on xyz to ≈ 10°. Planning on housing the servos in the base, and running cables down the arms to actuate them. Roughly 24 segments total, with separate control cables for the first 12 and second 12 (so the first half can bend in one direction, and the second half another), along with a drive cable down the center so it can rotate on the z axis. This also means I’ll need some empty space running down the center of the segments to run the drive cable and the control cables for the second half, so they don’t impede/aren’t impacted by the bending of the first half.

As far as I can figure, spherical bearings are my only real option here. As a test, I bought 4 steel spherical bearings from Amazon to connect the segments. And man, I love these things. I’d never handled them before, and they are smooth as butter. Unfortunately, they were like $7/per, and I need around 100 of them. So, slightly more than I’m hoping to spend haha. Aliexpress has them for ≈ $1.50/per, but the shipping is obscenely expensive, and would take like 2 months.

So I’m trying to figure out another option. I’m a member of a local maker’s space, so I have access to a bunch of tools (metal shop with lathe, CNC, FDM 3D printers), but still don’t think I have a great solution.

The tools for making these the right way are obscenely expensive. I’m thinking I could machine the race in two halves and weld/clamp them around the ball, which would be a ton of work, and wouldn’t be as smooth, but would probably be sufficient.

I could 3D print them, but I can’t figure out a process that would be as smooth as I want. They’d need to be filled/sanded, probably coated with a dry lubricant. But the filling/sanding process would introduce tolerance issues, and I can’t afford to have slop with how the cables will work.

Anyone have any insights here to help me out? Thanks!

  • kersplooshMA
    link
    English
    3
    edit-2
    11 months ago

    Take a look at Igus plastic bearings. The plastic material will save weight versus steel, which you’re going to need. The price varies by size and type but should be less than what you paid on Amazon.

    Your project sounds awesome! You should post pictures.

    Edit: Sorry, was typing faster than I was thinking. Even if the Igus bearings are $4 each, that’s still $400 worth of parts. That seems high for a hobby project. I agree with @[email protected], 3D printing and sanding is probably the way to go for a low-cost solution. Do you only have access to a FDM printer, or is resin printing an option?

    • @HaphazardFinesseOP
      link
      English
      111 months ago

      I was just typing a response on the Igus cost haha. The size I need are only $2.50/per, still pricey, but more reasonable. I’d be more inclined to try them out if it didn’t take 30-60 days to get an order; I’d like to do a test fitting before I commit to $250 of parts, and 60-120 days is too long to wait for the full batch.

      I may be able to do them in resin. My work has a whole bunch of resin printers, but this would be a lot of parts to do on machines that are supposed to be doing actual production work haha. That and the resin we use is accuracy focused, not particularly strong, they’d wear out fairly quickly.

      The maker’s space recently acquired some resin printers, so might be worth it to wait until they get those operational.

      Thanks for the insight!