Role: Responsible Engineer, Scope Owner
Duration: November 2024 – February 2025
Objective: Verify that a 6061-T6 weld with 5356 filler has a minimum tensile ultimate strength of 20,000 psi after cycling to 70 Kelvin with liquid Nitrogen.
Related Projects:
Summary: CSI Rockets had never manufactured a cryogenic welded pressure vessel before. I led the development of a weld qualification tank, which was simpler to machine and lower cost compared to the full flight LOX tank. After machining and welding, the tank was hydrostatically proof tested to 1,350 psi. It was then filled with liquid Nitrogen and held for 30 minutes, before being warmed back up and hydrostatically burst tested. The weld failed at 1,925 psi due to weld porosity causing a gross leak. The demonstrated burst factor was 2.25x MEOP, above the designed burst pressure to 2x MEOP, and gave us the confidence to proceed with tank manufacturing.
Article: In my previous two years on CSI Rockets, I was responsible for the development of two bolted closure pressure vessels (2022/23 and 2023/24). In these designs, the enclosures sealed against a 6″ diameter aluminum extrusion with Buna-N o-ring seals. O-rings generally cannot be used for cryogenic applications because elastomeric materials lose their elasticity at low temperatures. When transitioning from Nitrous Oxide to LOX, one of the principal challenges was designing, validating, and testing an aluminum weld using common, inexpensive alloys for use in the LOX tank.
Two primary materials were traded for the tank: 304 Stainless Steel and 6061 Aluminum. 304 was not selected due to a poor strength-to-weight ratio in its unhardened state and frangible failure modes at burst. 6061-T6 was subsequently selected as it is commercially available, low cost, and has a great strength-to-weight ratio without any additional hardening.
6061 is weldable, although it requires a suitable filler and skilled welder to prevent cracking. I needed to investigate how the heat affected zone in the weld would compromise the T6 temper. Initial research showed that correct welding of 6061-T6 would give a minimum of a T4 temper in the HAZ, although skilled welding could give up to a 24,000 psi ultimate strength.
I designed the welded joint to be a single-v groove using a Gas Tungsten Arc Welding process. A 5.625″ ID pressure vessel did not have sufficient clearance for additional internal welds, so the groove allowed for fusion through the entire wall. NASA documentation for GTAW of Aluminum used a 4043 filler, however, I selected 5356 because our welder was significantly more familiar with that process.
The LOX tank was a complex and machining-intensive design as a flight article, and had demands to interface with our valves, pressure transducers, commercial fittings, and act as a structural member of the vehicle while minimizing weight. Each end cap of the flight tank represented approximately 20 hours of programming and machining. Therefore, I decided to first build a qualification tank with a simplified end cap design. The qualification tank is lower cost, faster to machine, and allowed us to burst test an identical weld to verify the ultimate strength.
I selected a groove angle of 70 degrees for the weld, with a 1/32″ (0.03125 in) flat face. The tank has an OD of 6″, with a wall thickness of 0.1875″.
The simplified assembly drawing is shown above. It is not intended to provide full dimensions of the end caps, instead it provides nominal dimensions for design-for-welding feedback.
We machined the qual tank end caps on the Haas ST-20Y and Haas Mini Mill in the Columbia Makerspace, then shipped them for welding at our sponsor.
I also identified an additional opportunity to test static seals for our valves and injector. The cryogenic valves use PTFE spring-loaded seals for dynamic sealing, but to save costs I designed 40% compression static PTFE face seals. PTFE is not an elastomer and it is plastically deformed during installation by design, but even when accounting for replacing the o-rings after each installation it was still a huge cost win.
There was some anecdotal information online that Viton could work at cryo, and Viton is an oxygen compatible elastomer, so that was tested as well.
