On the CWRUBotix MATE ROV team, our enclosure and actuator testing used to rely on limited access to bodies of water, where the depth was lower than that at competition. 

Our testing limitations often caused enclosure failures to go unnoticed, damaging expensive electronics and hurting us at competition. It also slowed our pace of development.

I proposed and built a hydrostatic testing chamber to solve these problems.

Process

I began by defining the functional requirements, such as the ability to pressure test every enclosure (except the main E-bay) to a factor of safety of 4x, as well as non-functional requirements such as documentation and safety. 

I then sketched a concept, got into rough CAD and BOM, and went through several design iterations to implement feedback from professors and design reviews. 

Some of the changes included developing a waterproof checklist to ensure safe operation, relief valve sizing, and mechanical design improvements.

The final design uses a hydrostatic pump with check valve and gauge, pressure relief valve, and modular endcaps that support our standard enclosure penetrations. 

I designed every manufactured part to be waterjet cut from a single sheet of spare 1/2" aluminum that we had on hand, aside from the aluminum extrusion frame. This achieved very high material yield, reduced machining time, and allowed for very efficient secondary operations. 

I used a combination of manual and CNC machining for features that could not be waterjet such as O-ring grooves and blind tapped holes.

I ended up turning the COTS gasket fittings into hard-stop O-ring boss fittings by bonding a part I machined to them; this was necessary since the gasket would cause leaks if the fitting was not tightened perfectly, or if it was subjected to any side-loading.  

Using teflon tape and large, greased O-rings eliminated leaks, allowing pressure to be maintained over several days. 

Results