Deep Sea Tolerant Bristlemouth

[wpardis]

What are the barriers to pressure-rating Bristlemouth (BM) for deep-sea deployment? Which features of the Dev Kit dictate the 100m depth limitation? What formal pressure testing has been conducted on the interconnects, and in what configuration?

I have been discussing these questions with a former colleague @ehayden, a research engineer at Woods Hole Oceanographic Institution’s (WHOI) Deep Submergence Laboratory (DSL). DSL could be a great place to help consider some of these challenges having nearly a century of experience engineering in the deep sea.

Why do this?

Two opportunities where this could have impact and expand the breadth of Bristlemouth platforms and devices while enabling novel oceanographic research:

Multidisciplinary Instrumentation in Support of Oceanography’s (MISO) Synchronous Digital Subscriber Line (SDSL)

SDSL effectively enables a subsea network switch at the end of a standard 3 conductor sea cable (~10 km) to a ships onboard network. It provides scientists with networked telecommunications on oceangoing CTD carousels—common platforms used for oceanographic research and instrumentation to evaluate the water column. By enabling this network switch with a BM interface, we could incentivize technology developers to adopt the standard. Simultaneously as device adoption grows, scientists would gain access to an expanding set of ready-to-use tools for oceanographic research.

Example Figure: A CTD carousel carrying the SDSL system operated off the R/V Thompson. The operation used real-time imagery and data from a custom methane instrument to successfully locate seeps in Puget Sound.

National Deep Submergence Facility (NDSF)

NDSF is a public facility where scientists can access deep-sea research capabilities. Woods Hole Oceanographic Institution (WHOI) operates three NDSF vehicle assets (HOV Alvin, AUV Sentry, & ROV Jason). By enabling these assets with BM technology, we could give scientists access to an ecosystem of ready-to-use BM devices and ensure interoperability between vehicles. As a facility designed to lead and support innovative deep-sea technology, NDSF would be an ideal starting point for developing and demonstrating this standard in the deep sea.

It is important to note that although these are WHOI operated these are public facility granting access to scientist around the world. Which would make this development lasting and inclusive. Would love to have this conversation with the BM community.

[estackpole]

@wpardis I love this topic! I helped to develop the physical Bristlemouth connector so its deep sea capabilities and robustness are something I’ve thought a ton about and am excited to discuss.

The official rating is 150m, because we have not done thorough a testing following established testing procedures beyond that depth. That being said, I am quite a nerd about deep sea design and happen to have a Full Ocean Depth (11km) test chamber that I built in my garage. Anecdotally, I have been able to take open-face Bristlemouth sockets with 1atm backing to a pressure of over 800 bar (~ 12000 psi) or 8km of equivalent depth without failure or signs of damage. I’ve also done a test all the way to 1100 bar (FOD) which did have a failure, but the reason was inconclusive. I think more flange reinforcement on the socket may be needed to keep it from deflecting. These tests were done this with a given connector a few times early on to see if and how it might fail so the test were by no means exhaustive. They did not include cyclic testing or testing in varying thermal environments or with varying mechanical stresses.

The design of the socket includes reinforcement of the flange that holds the socket against the surface it seals to using features in the metal core for the part. This reinforcement was included in the design in anticipation of where extreme pressure may cause the part to fail, and future iterations of the design may include an even thicker flange feature to assure the sockets can withstand FOD in an open face configuration. In a closed-face configuration (with the plug attached) I suspect the mated pair may already be able to withstand FOD, but I didn’t get around to testing that before lending my chamber to some colleagues.

The Bristlemouth plug (the part that mates with the socket) was also designed specifically with extreme high pressure and robustness in mind. The sealing interface uses a face-seal o-ring so that when the fastener is tightened there is virtually no extrusion gap along the interface line, which is a place o-rings are likely to fail at high pressures.

Having an individual plug for each conductor, though a bit more work than having numerous conductors in one plug, makes the system much more robust because it keeps the design simple (which reduces both field and manufacturing failure points) and allows for much bigger connecting parts that are also more robust and easier to inspect. The isolation of the two connectors from each other also means that both connectors would need to fail simultaneously for an electrical short to happen, and the distance between those points creates much more resistance through the saltwater path than if they were right next to each other. Once again, everything about this design is focused on robustness.

Going back to deep-water considerations, the connector is just one part of the entire Bristlemouth system. Depending on your waterproofing technique for the electronics being used, there may also be pressure limitations there. The current Bristlemouth Mote we use at Sofar has not been specifically designed for ambient pressure. It has a few electrolytic capacitors as well as an exposed crystal oscillator onboard which are components known to fail at high pressures. What pressures those would fail at is TBD, but it certainly seems possible that a new Mote could be designed to more likely withstand high pressure. If the electronics are likely to be contained in a 1atm housing, all should be fine.

I once got to visit WHOI as the (then brand new) large Full Ocean Depth chamber was being used, and it blew my mind. I THINK THAT STUFF IS SO COOL!!! It would be amazing to start testing and qualifying Bristlemouth for these types of applications, and of course, please keep us in the loop if plans to test the technology out at high pressures comes together. I’m sure there are ways we could help support that testing!

[wpardis]

@estackpole I appreciate your enthusiasm—it truly shines—and all your hard work. Thank you for your thoughtful response and consideration. I am now quite optimistic about the technical feasibility of this idea. To address your considerations, let me be more specific about the proposal:

I suggest a development pipeline that progresses through increasing levels of risk and complexity. We would start with the low technical risk integration on the SDSL system (“networked cage on a cable”), then advance to Jason ROV (“networked robot on a cable”), followed by SENTRY AUV (“Autonomous Robot”), and finally tackle the platform with the strictest requirements—the ALVIN HOV with people on board.

Please see the diagram below showing a proposed SDSL BM integration. This diagram is derived from Figure 1 in the SDSL documentation.

Regarding your comments on mote pressure tolerance, all the embedded electronics would be housed in a hermetically sealed pressure vessel at 1 atm, so nothing innovative there. The core challenge of this integration would be implementing the BM sockets as penetrators on the SDSL end cap. Conveniently, the SDSL system runs on a 24Vdc deep-sea battery—matching the bus voltage of BM.

We need to fabricate a “Mote to Ethernet Adapter.” @zack_j do you know if anyone has already done this or at least considered it?

The biggest hole I see here is a classic chicken/egg dilemma. If we build an interface, where is the deep-sea BM tool? Wait… I have an idea - @NickRaymomd could we fit PiCam into a MISO GoPro housing and BM enable it?

So much fun! and yes WHOI pressure test facility is a awesome resource and one of the many reason why WHOI-DSL collaboration seems like a great fit. My guess is this is the vessel you were talking about…

Image taken from: SOLARIS, WHOI pressure test facility - ALIVIN certifying SOLARIS a deep sea reactive oxygen instrument.

@estackpole Would you be willing to go into a little more detail on the mechanical implementation of the socket you used for your pressure testing? Is it feasible to machine that into these SDSL end caps? How would you suggest going about that?

[luk]

Hi @wpardis

About Mote to Ethernet Adapter: if I understand correctly BM works on top of Ethernet then in theory I think Mote to Ethernet adapter should not even be needed. In theory it should be enough to have 10Base-T1L to “regular Ethernet” adapter, like i.e. EVAL-ADIN1100 board which @spiderkeys was using here Bristlemouth Native Linux Support - A work in progress! or even the one I’ve made 10Base-T1L RPi header with power over data line extension. In practice however I think some version of native Linux support for BM would be needed in both cases.

On the other hand because both are basically Ethernet I think it should be possible to use both BM and non-BM devices.