CAD Dojo
06: Watch case
Mon 11 October 2021
Background
I've long had an interest in mechanical clocks, and recently designed a novel pendulum for a floor-standing clock. Lately I'm more into watches and think I might like to have a go at making my own watch, so I'm getting started by modelling a clone of a watch case that I had lying around in my watch parts box.
Goal
Create a CAD model of a watch case.
Also, 3d print it and see how close it is to working. Will it fit a movement? A crystal? Can I wear it and use it as an actual watch?
Design
If I were to design a watch case for 3d printing, I wouldn't design it like this. This is quite a delicate and subtle design, and has inconvenient overhangs that are difficult to print for little to no benefit. But this is the watch case I have to work with, so it's the one I'm cloning.
I copied the dimensions from the reference part with Vernier calipers.
I started by modelling a cross-section of the case and then revolving it. Initially I did not include the lugs for the strap, but when I came to model these I discovered that the surfaces were formed by cutting excess material away from a revolution, leaving the lugs behind, so I went back and added more material to my revolution so that I could form the lugs.
The constraints in this sketch are quite messy, I should get better at laying out the dimensions so as not to obscure the sketch.
This sketch does almost all of the work. We're creating the surfaces that the crystal fits against, the surfaces that the movement fits against, and almost all of the final visible surfaces.
I didn't bother trying to make the threads to screw on the case back, because they're extremely fine and I wouldn't be able to measure them correctly, let alone 3d print them. If I do make a watch, I'd probably attach the case back with screws.
I created the top lugs in one feature and then mirrored it to create the bottom lugs. There's o particular reason not to model them all in one feature, I just thought it might save a few clicks to do it this way.
With lugs in place we need to make the holes to attach the strap. If I were to make this part with any technique other than 3d printing or casting, I would probably need to drill the holes from the outside, so they'd need to go all the way through. But in CAD we can make inaccessible blind holes all day long with impunity.
The next part is the little hole for the winder stem to go through. The watch I'm copying has a relief around this for the crown to sit inside. Most of these sketches are referenced to the flat face caused by this relief. The relief face is created by making a sketch on the Y/Z plane and manually editing its "Attachment Position Z" property to bring it up to where I want, and then pocketing out a circle to make my flat face.
I am always confused about how the attachment positions work in FreeCAD. It seems to work inside the coordinate system of the sketch, rather than the global coordinate system, which is quite counterintuitive because it seems like the coordinate system of the sketch can't be determined until after you've attached it to something. I mean, isn't the sketch always definitionally at Z=0 relative to its own coordinate system?
Finally, I filleted some of the edges on the lugs. I can't quite tell what kind of geometry the edges of the lugs are supposed to be on the real watch. It's possible that it's just dead square corners that have been polished down, but filleting the outer edges seems reasonable.
I sliced it with support material, and in PrusaSlicer's "0.10mm DETAIL" preset, which I haven't used before, but it seemed applicable.
After a 46-minute print in Prusament Galaxy Black PLA, I assembled this:
Evaluation
I didn't actually expect all the parts to fit on the first try, so I'm pleasantly surprised that they did. The movement is a much tighter fit than would be ideal. I was particularly surprised to find that the case back snaps in against the layer lines, even though I didn't bother modelling the thread.
I didn't try fitting a strap because I don't have a suitable one on hand, but the holes look like they printed perfectly well so I don't see why a strap wouldn't fit.
I think I broke the movement by forcing it in, because now the watch won't run for more than a few seconds before stopping, which means I'm not likely to want to wear it. Never mind. And while I can wind the watch, I am no longer able to set the time. When pulled out to set the time, the crown just spins around without moving the hands.
If I were to make another watch case, I would provide more clearance for the movement and the winder stem, and I would make my own case back, with provision to attach it with small screws.