We are Norman Advanced Robotics, a group of mostly high school students and we are currently working on an underwater ROV for a professor at the local university (OU). The ROV is intended for surveying organisms on the lakebed of Lake Tanganyika at a maximum depth of ~60m. Our epoxy-coated PLA design worked for a little bit and then failed, with water entering the hollow part and pushing the epoxy out from the inside.
We plan on imitating this BlueRobotics part but the 2D drawings do not indicate the diameter of the part that the 238 O-ring touches.
If someone could give us proper dimensions and a correctly sized O-ring for a flange for a 4.5" ID tube, it would be greatly appreciated.
If experiencing significant pressures (which 60 m certainly is) a 3D printed end cap, flange, or o-ring gland is unlikely to work, even if reinforced with epoxy. Water is extremely insidious, especially under pressure, and seals and potting must be must be designed meticulously and carefully constructed to ensure a reliable seal. The gaps in any 3D printed part will prevent proper sealing. I strongly recommend using a solid machined or laser cut end cap and flange.
The way some of the cables are sealed could also pose a problem. The cable jacket needs to be fully intact from end to end, if any wires are exposed if there are cuts in the cable water can leak through the individual conductors.
After using the Apple Rubber O-Ring Gland Calculator, we found that the provided dimensions for the 4" BlueRobotics flange and tube with a -238 size O-ring (the one listed as the radial O-ring for that size flange) resulted in a %Stretch value outside of recommended:
Increasing the Groove Diameter resulted in worse stretching and decreasing it resulted in the %Compression value going below recommended. Does this mean that we can stretch O-rings slightly more than recommended?
We have decided on machining an aluminum tube to use for the flange, you are definitely correct about the gaps in 3D printed parts causing big problems.
As for the exposed cable issue, I see that this is how the T100 thruster cable is terminated at the motor.