Identify reasons for Viton bands / O-rings breaking

After #451 (closed) and !522 (merged), we need to continue to gather evidence to understand why O-rings sometimes break repeatedly on the insertion step.

@gneizen: I've been wondering if we could modify the tool somewhat to make the edges less sharp? I know of someone who's snapped more than 5 o-rings while trying to assemble their microscope, and they're the recommended FKM 75 Shore O-Rings from Simply Bearings.

@williamwadsworth: We tried to address some of this issue in !522 (merged). The current master build (https://build.openflexure.org/openflexure-microscope/master/) includes those revisions, including some changes to the description of band insertion which might help. There is also a draft video of band insertion which was posted on LinkedIn https://www.linkedin.com/posts/openflexure_openflexure-openhardware-microscopy-activity-7406319740551512066-06-S

However, we do not yet understand why bands break, which makes this really hard to address clearly. In working on !522 (merged) I tried inserting bands in all of the 'wrong' ways to see if there was a particular thing that was causing breakages. I was not able to break a band, except once where I deliberately hooked on one side and trapped the band so that there was only a very short section of band to stretch up to the other hook. This was hard to do, so I don't think it is likely to happen accidentally. I note that I was working on the body after !512 (merged) (which is also in the current master). This has small nubs on top on the inner actuator columns to set the correct range of motion. They might also be helping a little by making the hooks stand a little away from the roof inside the actuator. This could make it less of a problem if the nut tool is not used to lock the stage. I was also using the main_body_open_actuators.stl from that merge request, so that I could see a bit more what was going on when the band broke - except that they did not break. Writing all of this down, I have a suspicion that I might not have cut the leg ties, as I was just testing insertion not following the full build process. If the leg ties were left intact, then they would provide some of the same locking that the nut tool does. It would put strain on the mechanism, so is not something to recommend, but it might be why I could not break bands. (I tried again later with a body that had free actuation and still bands did not break using various 'wrong' insertion techniques)

I am pretty sure that there is something about technique, because we can see repeated breakages in some cases, but breakages are in general rare. For breakages from kits, we have the same stock of bands, and the same printing processes, which eliminates those possible causes. I have even taken parts that someone was breaking bands in and inserted bands cleanly and immediately, but I have not had the chance recently. However I have tried all the obvious different techniques that I have seen and not managed to have reliable breakages to investigate.