geezacoleman
commented
2 days ago Thanks for finding these @MisterHW ! Not sure how those rails separated in the design, but they should be fixed now in #137. I have also added a blank backplate and raised up the receptacle/gland to avoid contact with the USB/ethernet ports on the Raspberry Pi when fitted.
MisterHW
Compact OWL as of https://github.com/geezacoleman/OpenWeedLocator/tree/68f2ee177ff4bf59b490357e216150cbcec619c8/3D%20Models/Compact%20OWL has a few issues affecting printability and the function of the o-ring face seals.
tray.stl
The bottom of the model is not flat, leading to most of the area not being exactly on the print bed
main-body.stl
The profile presents as a multi-part stl. Older slicers don't fuse everything into a single object, which causes the rails to be pinted separately, and they may or may not be bonded to the profile.
failed test print section:
frontplate.stl
The o-ring groove does not follow the main body cross-section.
partial recreation in FreeCAD
FreeCAD may not be the best choice for a redo, but I've recreated the extrusion profile and a blind end cap (which can then be cloned and specialized). This ought to give a glimpse into the workflow and level of hassle when using this particular CAD program.
OWL Compact FreeCAD example (plus 3mf, step).zip
Here's another view of the end cap with o-ring groove. It's not dimensioned absolutely correctly.
Note: 3D printed parts can be expected to exhibit some level of creep, especially since there is a heat source inside. Therefore, it's probably best to use Shore 15 EPDM foam o-ring material and a conservative amount of compression. Since the flat areas around the M5 holes butt up against
main body, it's the depth of the groove w.r.t. this plane.