I was messing with 3d printed threads. I’ve already made some threaded objects that work great. But when I scaled up my recent project, the whole thing went to shit. I was postulating something like a bolt that’s as thick as a campstove; because it was really a campstove gadget and not a bolt at all.
Admittedly, that’s super huge. A bolt like that would only appear on weird shit like a bridge superstructure or an aircraft carrier or something. My assumption of scalability was naïve. My 3d printer told me all about my failures as a designer.
3d printers like to lay down layer after layer. They don’t like “overhang” where a layer sticks out into space without a supporting layer beneath. It seems the larger a bolt’s diameter gets the more the thread’s overhang goes from hassle to deal breaker. I’d experimented but had created naught but a series of print failures. Being the nerd that I am, I interpreted the whole thing as a fascinating intellectual puzzle.
Eventually I had a revelation.
There’s no reason threads have to be symmetrical.
Is your mind blown? It isn’t? Are you mad? It ought to be!
Using a way of thinking I’d never before imagined, I created threads with gradual slope on the downhill side (thus supporting the print and avoiding sag) and a flat surface on the top (which is good in a mating surface). I projected the shape as a helical inclined plane, just like a “normal” thread.
This asymmetric black magic would make a metal lathe burst out in hives. But it makes sense for a 3d printer. 3d printed objects are a whole different animal than lathe machined metal. That means 3d printed stuff doesn’t and shouldn’t play by the rules of the machined (or injection molded) parts we’ve always experienced. Every nut and bolt I’ve ever seen in my life, most of which are metal, has had symmetric thread. So what?
I’m not the first guy to think of this. They’re called “buttressed threads”. They exist in screw jacks or some other obscure shit. I’ve never encountered them. But they solve my 3d overhang issue. I drafted a trapezoidal cross section in a FreeCAD sketch, performed an additive helical rotation, and I think it’ll work. I was feeling pretty studly.
But it was late. A sane guy would be in bed hours ago. I’m deliberately and only partially successfully trying to adjust to being a retired semi-normal human being. Part of that is keeping my inner mad scientist in check. There will be time to tinker more on another day. I should get a good night’s sleep and take a break from playing with weird math. It was already 1:30 am!
I shut my computer down and started walking to my house. This is when I indulged in another of my idiosyncrasies. It was just shy of 2 am when I wandered into the forest.
More in my next post.
