I ordered the first printed prototypes last week. They’re being made at three times actual size, because the silicone insert has a lip running around it that’s a quarter of a millimetre wide and the printer’s tolerance is bigger than the lip. At real size I’d be measuring the machine instead of the design. Stretch is a ratio and ratios don’t care about scale, so a triple-size model answers the question I actually have.
Then the printer emailed me. Silicone prints best lying flat with nothing hanging over open air, and my part has features on both faces. Whichever way up it goes, that lip ends up printing over nothing.
Three ways round it. Print it with supports, flatten one face, or cut the part in half and glue it back together. Supports would land on the lip, and the lip is the one feature this whole design gets judged on. Flattening a face deletes it. So the part gets cut.
The cut runs down the middle, which is the one place on it where nothing is being measured, and each half keeps a whole lip. Because the insert is a ring, both halves turn out to be the same shape, so it’s one file printed twice with one of them turned over by hand.
What I don’t know is whether the glued joint survives. Snapping the insert over the channel pulls the two halves apart in the exact direction a glue line is weakest. If one fails at the seam I won’t be able to tell whether that’s the glue or the design.
The material isn’t quite what I thought either. It’s a photopolymer made to behave like silicone rather than actual silicone, and it’s softer than what I’m aiming for. Softer deforms into the channel more easily, so if it passes, that might be the material being generous rather than the design being right.
Parts should land soon. First thing I do is measure them before anything gets fitted.