Seal Design
One of the unique challenges of the RoboSub competition is that our robots have to operate submerged at depth for sustained amounts of time. Our new vehicle has a total of thirteen o-ring seal systems alone. O-rings have been proven to be very robust and reliable, and require very little maintenance. We designed all of our o-ring glands around the parameters published under Parker-Haniffin's O-ring handbook (ORD5700).
A unique element of our new vehicle is that the main enclosure is pressurized up to 5psi, the approximate equivalent water pressure at a depth of 15 ft. Pressurizing the interior turns the load from compression by the water pressure into an internal pressure, strengthening the hull in the same way that a soda can is stronger when it's pressurized. Through most of the vehicle's mission, there will be very little force on the enclosure. In addition, if a seal fails, air will first egress the hull before water leaks in. Since o-ring failure modes are rarely 'explosive' in nature, our team will have time to surface the vehicle before any electronics suffer any damage. A pressure sensor inside the hull can make an early detection easy.
Design resources from Parker O-Ring Handbook
Most of our o-ring seals are face seals instead of radial seals, as having face seals allowed us to design parts that would have less material and weight.
Hydrophone Box Face Seal.
The lid would have to be much thicker to accommodate a radial seal.
This year, we decided to place our hydrophone FPGA and amplifiers in an external enclosure because of the frequency we needed to access it. Instead of using normal cable pass-throughs as we did in the past, we decided to explore a new COTS product this year.
Image Credit: Icotek
Our hydrophone system is very sensitive to small changes that could occur when splicing, so we opted to try Icotek's new gland system. It offers grommets of varying sizes to accommodate different cable diameters. Having a versatile system like this also allows the electronics inside to be removed without re-splicing cables.
Although we did our best to minimize the number of holes we would have, we still needed to have some fasteners pass through the carbon, such as in the swappable rear conduit panel, as epoxy would have made it permanent and defeat the purpose. To seal up these holes, we used neoprene sealing washers that use the fasteners' force to keep water from ingressing.