OOI RSN Housings Pass Proof Pressure Tests

Wednesday, December 12, 2012
OOI RSN Housings Pressure Test image
Proof Pressure Test Figure 3

 Figure 3. Randy Fabro and Larry Nielson lowering a housing into the UW School of Oceanography Pressure Test Vessel.

--Photo by Geoff Cram

Proof Pressure Test Figure 6

Figure 6. Pause for thought. (Sign in the lobby of the building that houses the pressure test facility.)

--Photo by Geoff Cram

Systems designed to operate below the surface of the ocean must survive the relentless corrosiveness of saltwater and often great pressure as well. Protecting sensitive electrical components in sturdy, corrosion-resistant housings that have been tested on land is one of the key requirements for ensuring a successful undersea installation. Geoff Cram, OOI RSN Principal Mechanical Engineer, describes below pressure testing three such housings at a School of Oceanography facility on the UW Seattle campus in November 2012. The items tested were aircraft-grade titanium housings for the three types of secondary nodes: the Medium-Power Junction Box, the Low-Voltage Node, and the Low-Power Junction Box.  These nodes serve as bridges between the OOI RSN instruments and the Primary Infrastructure, providing power and communications to the instruments, and transmitting their data back to shore.

At 7:38 PM on Thursday, November 8, 2012, the last of three OOI RSN First Article Secondary Nodes pressure housings had successfully completed its proof pressure test. This is an important milestone: the test validates the housing’s structural design, which means that we’re confident it is strong enough to withstand the intense water pressure at 3500 m below the surface of the ocean, i.e., 5200 pounds per square inch (psi), the equivalent of 354 atmospheres.
The Design Envelope
Engineers generally approach a new design effort systematically, starting with defining a set of facts that we call the design envelope. We began our housing design exercise by asking questions about contents, size and weight constraints, material preferences, working life, and total number of housings. As is usually the situation, we accumulated a range of answers that we reduced to a workable solution with the other stakeholders who were, in this case, our RSN electrical and metallurgical engineering colleagues, lead field engineer, and project scientists.
Finite Element Analysis
After consensus had been reached on the basics of these particular housings, we designed each one using a Computer Aided Design (CAD) package called SolidWorks. This allowed us to create accurate models of all parts to make sure they would fit together precisely and to generate part drawings. But before we could hand those drawings off to a machine shop for fabrication, or even before we were ready to buy the raw material, we needed to find out if our designs were strong enough. To do that, we exported the CAD models to another computer package for stress analysis, also known as Finite Element Analysis, or FEA for short. Guided by the engineer, the FEA package divides each model into thousands of very small elements, each having the properties of the intended housing material, and applies virtual loads to those elements. After several hours of churning away, the program spits out a result. As is often the case, we found places in the models where the stresses were greater than the material can withstand, so it was back to SolidWorks to adjust the geometry, followed by another pass in the FEA package. Eventually, we ended up with acceptable stresses and no more material than necessary for all three housing designs.

But we weren’t ready for full-scale production yet. These housings will be made from Grade 5 titanium, a very strong and corrosion-resistant alloy that costs a lot of money and can take months to acquire. Machining is also expensive. Because we will be deploying 14 housings in two different sizes and of three types in 2013, we needed to be sure they would be up to the job before we cut any metal. After all, these housings are the heart of our program – if they don’t work, almost nothing else will. To establish beyond the shadow of a doubt that our designs are structurally sound, we needed to conduct a proof pressure test for each housing type.
The Proof Pressure Tests
A proof pressure test is an easy process to describe, somewhat harder to execute: assemble a housing (we call these units first articles to distinguish them from the production housings that will subsequently be built), and subject it to a higher pressure than it would experience when deployed.  Then take it to a brightly lit room, and open it up to look for leaks or damage; either event would mean that the test (and possibly the design) had failed.  We only do this to first articles; our production housings will be tested at the lower, design pressure of 5200 psi.

Our proof pressure tests involved two brief intervals at 125% of design pressure followed by a one hour “soak” at this value, which is 6500 psi.  We’re fortunate that the UW School of Oceanography has an excellent pressure test facility that is only a quarter mile from our assembly lab in Ben Hall.  Adding spice to the tests, however, is the facility rule that, if the test pressure will exceed 6000 psi, the building must be evacuated by all personnel except those conducting the test.  This leads to curious thoughts on the part of first-time participants and means a long day for all involved, since such tests are conducted after 6 PM.

All three types of our housings passed with flying colors. No retests were needed, and we breathed a collective sigh of relief at the end of the third run. Now it’s full speed ahead for production – deployment is only seven months away!

Special thanks to Larry Nielson, RSN Lead Field Engineer; Patrick Waite, RSN Senior Mechanical Engineer; and Randy Fabro, School of Oceanography Engineering Tech.
--Article written by Geoff Cram, PE, OOI RSN Principal Mechanical Engineer

(Full set of photos may be found to the left on this web page, under the search bar.)