3D printable shaft-passer: parametric shaft-passer in OpenSCAD | Hacker News Reader

3D printable shaft-passer: parametric shaft-passer in OpenSCAD

cubehero.com

22 pointsiamwil15 comments

In case some of you were wondering, this is related to a previous post:

https://news.ycombinator.com/item?id=6817451

It's a device that allows cables to pass through it. As it happened, we were talking about in on HN, and a few days later, I found this 3D printed version.

As I said on the previous thread, I'd really like to see actual evidence that the thing existed, worked, and was used. Every serious discussion I can find goes back to the mention in Feynman's autobiography, in which he presented the thing as a friend-of-a-friend tale, and it's a bit improbable to accept on that alone--it doesn't look impossible, but it seems like it would be very difficult to keep the gear turning smoothly without binding.

Without convincing evidence, it's hard to see the shaft passer as anything more than an engineer's prank. I'd love to be proven wrong, though.

What if you had a spoked wheel (bicycle wheel) where every spoke had a little spring loaded collar at the location of the shaft. As the shaft hits the collar, it is pushed out of the way while the other spokes, secure in their collars maintain enough tension to keep the wheel attached.

spoke closed '===', the collar is '/' and the shaft '(asterisk)'.

Spoke direction of travel is s4 -> s0.

  s0c------===------

  s1c------===------

            /   
  s2o------/*-------

  s3c------===------

  s4c------===------

  etc.

When the collar closes, its spoke will have pulled apart a little bit under the tension and widened the gap. You'd need some kind of motorized grabber-puller collars to make this work.

The beauty of OpenSCAD never ceases to amaze me .. minkowski as a built-in?

http://en.wikibooks.org/wiki/OpenSCAD_User_Manual/Transforma...

Wow, its just so .. elegant ..

Maybe, but it of limited use for most CAD problems - "standard" visual sketch-and-extrude systems are MUCH easier to use for 99% of things you'd want to model.

Exceptions are highly repetitive and mathematical objects like gears, CNC art, fasteners and so on.

OpenSCAD is pretty good for programmers transitioning into 3D modeling. The minkowski can get slow for high triangle counts, so it's not used too often. But it's as a result of relying on the underlying 3D graphics library, CGAL.

3D printed: http://www.thingiverse.com/thing:193022

Cool, but I'd be surprised if this actually works well - the shaft would surely catch on the shoulder bit, and the force pushing the shaft through is perfectly in line with the axis of the wheel, so there is very little moment turning the wheel.

These can be fixed by:

1. Smoothing out the outside a lot. 2. Rotating the "guards" around the wheel by 45 degrees about the axis of the wheel, so the shaft enters at the top or the bottom of the wheel rather than in the middle.

I can make a drawing if that isn't clear.

Plus I haven't come across any low-friction 3D printed materials yet but there's not much you can do about that.

Can you please describe what this is exactly and how it works?

Sure, I found this blog post that describes it pretty well:

http://somethingsurprising.blogspot.com/2011/11/feynmans-sha...

But basically, it was something that was posted recently on HN, about how Richard Feynman (famous physicist) happened upon a mechanism designed in WWII to avoid mines being detected by trip wires between boats, by allowing the trip-wire to pass through the shaft-passer

Wow - that's actually really cool.

Am I in the wrong for briefly wondering why 3d-printed dildos was on HN?

Yes, you are wrong. 3D printed anything is still pretty amazing, and now we have people who are trying to disrupt the adult toy industry. A worthy hack IMO.

And an excellent moniker for it, Jade D'Drag.

Reply on news.ycombinator.com