- Team Resources
After much consideration and prototyping, the team has decided to go with the 1200 Pelican boxes to house all of our electronics and power supply. We will have two of them on our ROV and the frame is being built around these two large boxes.
Our original plan was to create a cylindrical box to house all of out electronics and the 1200 Pelican box for the power supply. The cylinder was to be made out of 4 in diameter polycarbonate from Port Plastics. To seal the box, end caps machined out of Aluminum with O rings were going on both sides. The best things about this method is that it used O rings in the best shape, a circle. It can also be best fitted to the electronics inside, this limits down on the amount of air inside the robot so we won't need weights to make the bot neutrally buoyant. The polycarbonate was too expensive so we bought 4 in diameter PVC, made an end cap and it was successfully water proofed. This Box was put on the shelf due to it's lack of flat surfaces needed for the connectors for the electronics. Our next attempt was to create a rectangular box with a transparent polycarbonate lid sealed with a gasket. We discussed several ways on making the box: Make it out of pieces of aluminum and bolt it together with epoxy to fully seal the seams, make it out of aluminum pieces and weld them together, or buy a huge block on aluminum and machine it completely out like what we did with the ROAR project. We tested all three designs. The bolted epoxy leaked just a tiny amount but it was tested half way submerged; it would leak a lot more at depth. The welded design was attempted but our welder had to stop a little early and we never fully finished welding it. The machined design was tested with a much smaller version of the box made to test different types of underwater connectors. It took several hours just to create small 5 inch by 3 inch by 1.5 inch box. This would take a very large amount of material and time in the machine shop.
The seal gasket was thought to be better than O rings because it would seal over a bigger surface area. However, this point was also it's downfall. The gasket, since it covers a larger surface area, did not seal as well and o rings might because the compression force was distributed evenly across the whole gasket. So instead of have a very good small seal we had a very large bad seal. Ideas to mill out a canyon so the gasket would mainly seal on the two sides of the canyon. Like the reverse of an O ring, sealing not in a groove but rather the area around the groove. As of now, this has not been tested.
The biggest enemy in this business is time. While testing all these designs the thing that ultimately led to their demise was the lack of time and the lack on manpower allowed for each box. But we learned a lot!