Introduction
Liquid hydrogen (LH₂) and liquid oxygen (LOX) are the highest-performing chemical propellant combination available to launch vehicle designers, delivering specific impulse values in the 450–460 second range in vacuum. The BE-3U engine powering Blue Origin’s New Glenn GS-2 upper stage, the RL-10 family powering Centaur and DCSS, and the J-2X heritage all depend on this combination. The Space Shuttle Main Engine and SLS RS-25 push the propellant combination to its performance limits.
These propellants also impose some of the most demanding structural requirements known to engineering. LOX tanks operate at 90 K (–183 °C); LH₂ tanks operate at 20 K (–253 °C), only 20 degrees above absolute zero. Both service environments impose large cryogenic thermal strains, cyclic pressurization, and — in the case of the LH₂ tank in particular — intimate contact with the most chemically aggressive embrittling agent in the materials engineer’s vocabulary: atomic hydrogen.
This post examines the mechanisms of hydrogen embrittlement (HE) and their interaction with fatigue crack initiation and growth for the two candidate tank materials — aluminum alloys and austenitic stainless steels — across both propellant environments. The two tanks on a typical LH₂/LOX vehicle are physically adjacent, thermally coupled, and often share structural elements, yet their materials challenges are meaningfully different.
Temperature Context
Before examining material behavior, it is worth anchoring the temperature environments:
| Propellant | Boiling Point (1 atm) | Tank Operating Temp | ΔT from Ambient (300 K) |
|---|---|---|---|
| Liquid Oxygen (LOX) | 90 K (–183 °C) | ~90–100 K | ~210 K |
| Liquid Hydrogen (LH₂) | 20 K (–253 °C) | ~20–22 K | ~280 K |
This 70 K difference between the two tanks is structurally significant: an LH₂ tank imposes ~33% more thermal strain per fill cycle than a LOX tank. For aluminum alloy 2219 (α ≈ 22 × 10⁻⁶ /K), a 280 K cool-down produces a linear strain of approximately 6,200 microstrain — well into the yield zone for thin weld seams and orthogrid intersections.
The LH₂ temperature is also below the condensation point of atmospheric oxygen (90 K) and nitrogen (77 K). If insulation fails, air constituents can condense and accumulate on the outer tank surface — a fire and explosion hazard, but also a potential source of differential thermal mass that worsens thermal shock during rapid fill.
Hydrogen Embrittlement: Mechanisms
Three principal mechanisms drive hydrogen-assisted degradation in structural alloys:
Hydrogen-Enhanced Decohesion (HEDE). Atomic hydrogen accumulates at grain boundaries, phase interfaces, or ahead of crack tips, reducing the cohesive strength of atomic bonds. The result is intergranular fracture at stresses well below the macroscopic yield strength. HEDE dominates in high-strength steels and in over-aged 7xxx aluminum alloys under sustained tensile loading.
Hydrogen-Enhanced Localized Plasticity (HELP). Dissolved hydrogen increases dislocation mobility in the crack-tip plastic zone, concentrating slip on fewer planes and causing locally ductile but macroscopically brittle behavior. HELP is observed in FCC metals including austenitic stainless steels and aluminum alloys. Under fatigue loading, HELP accelerates crack propagation by enhancing crack-tip blunting and re-sharpening per cycle.
Adsorption-Induced Dislocation Emission (AIDE). Hydrogen adsorbed on the crack-tip surface reduces the surface energy required to nucleate and emit dislocations. AIDE facilitates crack advance under lower applied stress intensity than in hydrogen-free environments and is considered an important mechanism in aluminum and austenitic steels.
All three mechanisms have a common consequence for fatigue: they reduce the effective ΔK threshold for crack propagation and increase the Paris law coefficient C, resulting in faster fatigue crack growth rates (FCGR) for a given applied stress intensity range.
A critical parameter governing all these mechanisms is hydrogen diffusivity, which is a strong function of temperature and crystal structure. In BCC ferritic/martensitic steels, room-temperature diffusivity is approximately 10⁻⁸ m²/s — fast enough to supply crack tips with hydrogen during fatigue cycling. In FCC aluminum and austenitic steels, diffusivity at room temperature is 10⁻¹⁰ to 10⁻¹⁶ m²/s, orders of magnitude lower. At cryogenic temperatures, diffusivity falls further still, which is one reason these materials are considered for LH₂/LOX service.
The LH₂ Tank: Maximum Hydrogen Exposure
Why the LH₂ Tank Is the More Demanding Case
The LH₂ tank is unique among propellant tanks in that it stores the embrittling agent itself. Hydrogen exposure comes from multiple sources:
Dissolved and gaseous hydrogen in the propellant. Liquid hydrogen is pure H₂. At the liquid-wall interface, hydrogen molecules adsorb onto the tank interior surface and a small fraction dissociate to atomic hydrogen. The rate of this surface dissociation is a function of surface oxide state, temperature, and catalytic impurities.
Ullage gas. During draining, the ullage above the liquid surface is pressurized gaseous hydrogen. Gaseous H₂ — particularly at the warmer ullage temperatures of 50–80 K — is more aggressive than liquid hydrogen because molecular dissociation rates are somewhat higher and gas-phase transport to the metal surface is more efficient than through liquid.
Manufacturing hydrogen. Gas tungsten arc welding (GTAW) and even friction stir welding (FSW) can introduce hydrogen into the heat-affected zone. GTAW is particularly susceptible: moisture on filler wire, base metal oxide contamination, and shielding gas impurities are all hydrogen sources. The result is hydrogen porosity — discrete gas pores in the weld bead and HAZ that serve as fatigue crack initiation sites with known equivalent pre-crack sizes.
Permeation from the exterior. During ground hold, moisture on the outer insulation surface can slowly permeate toward the metal wall. This is generally not significant for metallic tanks with intact insulation, but becomes relevant for aging vehicles or damaged insulation.
Aluminum Alloys in LH₂ Service
The Cryogenic Diffusivity Advantage
At 20 K, hydrogen atom diffusivity in aluminum falls to values so low that transport-controlled HE mechanisms are effectively frozen. An aluminum atom at 20 K has thermal energy far below the activation energy for hydrogen diffusion along grain boundaries or through the matrix. This is the primary engineering justification for aluminum LH₂ tanks: the same material that is the propellant container is kinetically protected from its most dangerous degradation mechanism at operating temperature.
This reasoning led to the use of aluminum alloy 2219-T87 for the Space Shuttle External Tank LOX and LH₂ tanks, and subsequently Al-Li 2195-T8 for the Super Lightweight Tank (SLWT) configuration. The SLS Core Stage reverted to 2219 FSW for its better weldability and cost. NASA’s recent material selection review for LH₂ aircraft cryotanks confirms that Al 2195 demonstrates resistance to hydrogen embrittlement, while noting that further study is warranted for welded structures in extended service.
Alloy-Specific Susceptibility
Not all aluminum alloys behave equivalently. The key variables are precipitate coherency, grain boundary chemistry, and temper:
Al 2219-T87 (Al-Cu-Mn). The benchmark aerospace LH₂/LOX alloy. θ′ (Al₂Cu) precipitates in the T87 temper are distributed within grains; grain boundary precipitate-free zones (PFZs) are narrow. SCC resistance is good. Hydrogen embrittlement susceptibility under cryogenic conditions is considered acceptably low for the flight cycle counts typical of expendable vehicles. Recent work on FSW 2219 joints shows hydrogen charging reduces yield and ultimate strength modestly via HELP, with base metal more susceptible than the weld nugget due to higher θ′ precipitate density serving as hydrogen trapping sites.
Al-Li 2195-T8 (Al-Cu-Li-Mg-Ag-Zr). Higher strength and lower density than 2219 due to lithium addition. Demonstrated resistance to hydrogen embrittlement in cryogenic conditions for the SLWT program. However, Al-Li alloys exhibit anisotropic fracture toughness and reduced through-thickness ductility that requires careful attention in dome forming. Cryogenic toughness can be lower than room-temperature toughness for some aging conditions, requiring proof testing strategy adjustment.
Al 7075-T6 and 7050-T7451. Used extensively in airframe structure but not in propellant tankage. These alloys are susceptible to hydrogen-assisted SCC and fatigue crack growth at ambient temperatures, particularly in sustained loading near yield. The T73 and T7351 over-aged tempers significantly reduce SCC susceptibility by coarsening grain boundary precipitates, but residual susceptibility remains. These alloys are generally avoided for propellant tank wetted surfaces.
Fatigue Life Drivers in Aluminum LH₂ Tanks
The dominant fatigue concerns in aluminum LH₂ tanks, ranked by severity in reusable service:
1. Low-cycle thermal fatigue at structural discontinuities. Each fill/drain cycle imposes ΔT ≈ 280 K. For an orthogrid barrel with grid intersections, the local stress concentration factor Kt at the fillet radius elevates the strain amplitude to levels that accumulate low-cycle fatigue (LCF) damage on a Coffin-Manson basis. For reusable vehicles targeting hundreds of flights, this is the primary life-limiting mechanism in the tank barrel.
2. Weld HAZ porosity as crack initiation sites. Even with FSW, which eliminates fusion welding hydrogen pickup, pre-existing defects in the parent material (shrinkage porosity, intermetallic particles) serve as equivalent initial flaw sizes (EIFS). The NASA SP-8088 damage-tolerant framework requires tracking the largest tolerable initial flaw through the mission life.
3. Cryogenic reduction in crack closure. Fatigue crack growth tests on aluminum alloys at cryogenic temperatures show that crack closure loads — which normally retard crack growth in ambient conditions — decrease at low temperatures because plasticity-induced closure is reduced. This effectively increases ΔKeff for a given applied ΔK, raising the FCGR in the near-threshold regime.
4. Gaseous hydrogen ullage effects. The warm-side ullage at 50–80 K is the regime where hydrogen diffusivity in aluminum, while still low, is higher than at 20 K. In this zone, hydrogen-assisted fatigue crack growth accelerations of 1.5–3× relative to inert environments have been documented in laboratory studies, though sustained flight conditions make this mechanism less severe than in static pressurization.
Austenitic Stainless Steel in LH₂ Service
Austenitic stainless steels (304L, 316L, Nitronic series) have been used for LH₂ tankage in pressure-stabilized thin-wall designs — most famously in the Centaur upper stage, which has flown since the 1960s. The Centaur tank’s stainless steel construction is pressure-stabilized: below a minimum internal pressure, the tank walls would buckle. This design achieves very low dry mass but requires continuous pressurization in all ground operations and flight phases.
The Metastable Austenite Problem at Intermediate Temperatures
The HE susceptibility of austenitic stainless steels is temperature-dependent and non-monotonic — it peaks in an intermediate cryogenic range rather than at the coldest operating temperature. Understanding this behavior is essential to safe LH₂ tank design.
Types 304 and 316 stainless steel are metastable austenite at room temperature. When deformed — by pressurization, mechanical loading, or the thermal contraction stresses of cryogenic cool-down — the austenitic phase (FCC, γ) partially transforms to deformation-induced α′-martensite (BCC). BCC martensite is orders of magnitude more susceptible to hydrogen embrittlement than FCC austenite. The result:
- At temperatures where significant plastic deformation occurs during fill (150–250 K), martensite fraction is substantial, and hydrogen embrittlement susceptibility peaks.
- At very low temperatures (20–77 K), the reduced mobility of hydrogen atoms kinetically suppresses HE even though martensite continues to form. Studies on 304L show that while hydrogen pre-charging reduces ductility at 20 K, the effect does not worsen dramatically below 77 K, and strength is substantially elevated by cryogenic solid-solution hardening.
- At room temperature, non-deformed 304L has acceptable hydrogen resistance; sensitivity rises with plastic strain and martensite content.
Research on 316plus stainless steel confirms that hydrogen embrittlement is substantial at all temperatures tested, reaching maximum severity at 77 and 20 K — with reductions in area of 40–50% in hydrogen-charged specimens compared to uncharged controls. Importantly, strain-induced martensite (SIM) fraction alone does not control embrittlement; the governing factor appears to be hydrogen trapped within narrow deformation zones at phase boundaries.
The practical consequence for LH₂ tank operations is that the thermal transient during fill and drain — when the tank wall passes through 150–250 K with developing thermal stresses — is the window of maximum risk. For the short fill-and-fly profiles of expendable vehicles, this is generally managed by proof testing and inspection. For high-cycle reusable systems, this transient becomes a cumulative fatigue life driver that must be explicitly modeled.
Stable Austenite Grades
The solution is to use fully stable austenitic grades that resist martensite formation:
| Grade | Ni Content | Martensite Stability | Cryogenic HE Resistance |
|---|---|---|---|
| 304 / 304L | 8–10% | Metastable | Moderate (degrades with strain) |
| 316 / 316L | 10–14% | Moderately stable | Better than 304 |
| 310S | 19–22% | Fully stable | Good |
| Nitronic 40 (21-6-9) | ~6% Ni + 9% Mn, N-stabilized | Stable | Good |
| A286 | ~26% Ni + precipitation hardened | Stable | Good (tested to 4.2 K) |
Fatigue in Stainless Steel LH₂ Tanks
Stainless steel LH₂ tanks (Centaur heritage) operate under continuous pressurization. Fatigue life drivers include:
Pressure cycling fatigue. Each flight cycle pressurizes and depressurizes the tank. For pressure-stabilized designs, any reduction below minimum pressure constitutes a potentially damaging load reversal. Fatigue crack growth at cryogenic temperatures in 304L is accelerated by hydrogen pre-charging at room temperature and at 190 K, though the effect at 20 K is less clear.
Hydrogen environment embrittlement (HEE) of fatigue strength. Studies on SUS304L show that HEE reduces fatigue crack initiation resistance at ambient temperature and at 190 K, manifesting as a reduction in the number of cycles to crack initiation for a given stress amplitude. At 20 K the HEE effect on fatigue is not clearly observed because diffusion is suppressed.
Weld residual stresses. Austenitic stainless steel welds retain tensile residual stresses approaching yield strength in the HAZ. These add to the pressurization mean stress, reducing fatigue life relative to stress-relieved material. Proof pressure testing to 1.25–1.5× MEOP serves as the primary flaw screening tool.
No endurance limit. Austenitic stainless steels, like aluminum alloys, lack a classical endurance limit. S-N curves slope continuously through 10⁸–10⁹ cycles. For reusable systems, there is no safe operating stress — only a cycle-dependent service life.
The LOX Tank: Lower Temperature, Different Risk Profile
Hydrogen Sources in a LOX Tank
A LOX tank does not store hydrogen, but hydrogen intrusion pathways remain:
Common bulkhead and adjacent plumbing. In most bipropellant vehicles, the LOX tank shares a common bulkhead with the LH₂ tank or is connected by crossfeed plumbing. Hydrogen can diffuse through thin sections over vehicle lifetime.
Manufacturing hydrogen. GTAW of aluminum LOX tanks introduces the same weld-zone hydrogen as in LH₂ tanks. Porosity and HAZ embrittlement from welding are the primary manufacturing-origin defects.
Electrochemical reactions on exterior surfaces. Ground handling in humid environments can generate hydrogen at cathodic sites on the exterior, particularly at scratched or abraded surfaces of 7xxx aluminum alloys.
Adjacent pressurization systems. Helium pressurant stored in COPVs within or adjacent to the LOX tank does not introduce hydrogen, but the COPV liner — typically aluminum — can be susceptible to hydrogen from the same manufacturing pathways.
The LOX-Specific Compatibility Constraint
Before hydrogen embrittlement, the LOX tank designer must satisfy a constraint that doesn’t apply to the LH₂ tank: compatibility with liquid oxygen itself. LOX is a powerful oxidizer; many materials that are structurally sound will ignite or react violently in its presence under impact, friction, or electrical spark.
Aluminum alloys pass LOX compatibility testing (ASTM G86, promoted ignition per NASA/MSFC-SPEC-1040) when clean and free of organic contamination. Particulate contamination — particles of any combustible material entrained in the LOX stream and impacting an aluminum surface under pressure — is the primary ignition risk. Tank cleanliness specifications are correspondingly stringent.
Austenitic stainless steels are also LOX-compatible in clean conditions. However, the density penalty of stainless steel (8.0 g/cm³ vs. 2.7 g/cm³ for aluminum) makes stainless LOX tanks weight-prohibitive for most launch vehicle applications. This is why New Glenn GS-2, like essentially all modern upper stages, uses aluminum alloy orthogrid construction for its LOX tank.
Aluminum LOX Tanks: Hydrogen Embrittlement Assessment
At 90 K operating temperature, hydrogen diffusivity in aluminum is low enough that transport-controlled HE mechanisms are kinetically suppressed. The Al₂O₃ native oxide further restricts hydrogen ingress. For these reasons, HE is not considered a primary structural degradation mechanism in aluminum LOX tanks under steady-state conditions.
The risk windows are:
Thermal transient (300 K → 90 K during fill). The 210 K cool-down activates thermal stresses at weld seams and stiffener intersections. If hydrogen is present from manufacturing, it can assist fatigue crack initiation during the thermal cycling phase, when temperatures are still warm enough for some diffusion.
Ambient-temperature storage with residual hydrogen. Between flights, the tank is at ambient temperature. Manufacturing-derived hydrogen in weld HAZ regions has maximum mobility at these temperatures. Sustained tensile residual stresses at weld toes provide the necessary stress field for HE-assisted SCC in susceptible alloys (7xxx). For 2xxx series alloys in T87 or T8 tempers, SCC resistance is substantially better.
Cryogenic fatigue crack growth. Reduced crack closure at 90 K increases ΔKeff relative to room-temperature conditions for a given applied ΔK. This is accounted for in damage-tolerant design by using cryogenic-temperature FCGR data (Paris law C and m at 90 K) rather than room-temperature values.
Stainless Steel LOX Tanks: Assessment
Stainless steel is occasionally used for LOX tankage in ground storage applications (dewars, test facility tanks) where weight is not critical. In flight hardware, stainless steel LOX tanks appear in pressure-stabilized upper stage configurations (Centaur LOX tank) and in SpaceX’s Starship, which uses 304L stainless for both LOX and liquid methane tanks — prioritizing repairability, cost, and high-temperature capability over mass efficiency.
For stainless steel LOX tanks: HE susceptibility at 90 K is very low for metastable grades because hydrogen mobility at LOX temperature is minimal. The 150–250 K transient window during fill remains a concern for metastable grades under deformation. Stable grades (310S, Nitronic 40) are preferred where hydrogen exposure and reusability are both design drivers.
Comparative Fatigue Life Framework: LH₂ vs. LOX Tanks
Load Cases
| Load Case | LH₂ Tank | LOX Tank |
|---|---|---|
| Thermal ΔT per fill cycle | ~280 K | ~210 K |
| Tank pressure (typical) | 0.2–0.4 MPa | 0.3–0.6 MPa |
| Hydrogen exposure | Direct (propellant) | Indirect (manufacturing, diffusion) |
| Fatigue environment | Cryogenic H₂ gas + liquid | Cryogenic, oxidizing |
| LCF risk (reusable) | High (thermal ΔT largest) | Moderate |
Damage-Tolerant Life Prediction Elements
A complete fatigue life assessment for either tank requires:
1. Equivalent Initial Flaw Size (EIFS) distribution. Established from NDE inspection results. Quantitative fractography on fatigue fracture surfaces allows back-calculation of the EIFS as an equivalent crack depth consistent with LEFM predictions.
2. Paris law FCGR data at service temperature. Fatigue crack growth rate data at 20 K (LH₂) and 90 K (LOX) in the appropriate environment is essential. These data differ meaningfully from room-temperature values due to reduced closure, elevated yield strength, and environmental interaction. NASA-HDBK-5010 provides guidance on test methodology.
3. Hydrogen environment correction factor. For stainless steel tanks, an HEE acceleration factor on Paris law coefficient C should be applied in the 150–250 K transient regime. For aluminum tanks, an acceleration factor for the gaseous hydrogen ullage zone (50–80 K) may be warranted for high-cycle reusable applications.
4. Thermal fatigue Coffin-Manson analysis. For reusable vehicles, LCF damage accumulation at orthogrid intersections, dome gore-panel weld intersections, and boss attachment rings must be tracked using ΔT per cycle, local Kt, and the material’s cyclic stress-strain curve at cryogenic temperature.
5. Proof pressure screening. A proof test at 1.25–1.5× MEOP establishes an upper bound on surviving initial flaw size via fracture mechanics. This is an important advantage of 2219: its toughness increases at cryogenic temperature, so room-temperature proof testing is conservative.
6. Residual stress state at welds. For FSW joints in aluminum, residual stresses are typically compressive or near-zero — significantly better than GTAW. For GTAW stainless steel, residual stresses approaching yield must be included in the crack driving force calculation.
Material Property Summary
| Property | Al 2219-T87 | Al-Li 2195-T8 | 304L SS | 316L SS | Nitronic 40 |
|---|---|---|---|---|---|
| Density (g/cm³) | 2.84 | 2.71 | 7.9 | 8.0 | 7.8 |
| RT Yield Strength (MPa) | 345 | 460 | 170 | 205 | 380 |
| 20 K Yield Strength (MPa) | ~500 | ~620 | ~1000 | ~1050 | ~1200 |
| Fracture Toughness KIC, RT (MPa√m) | 36–44 | 38–50 | >100 | >100 | >80 |
| Cryogenic Toughness trend | Increases | Depends on aging | Increases | Increases | Increases |
| HE susceptibility (20 K, LH₂) | Very low | Very low | Low (SIM suppressed) | Low | Very low |
| HE susceptibility (150–250 K transient) | Low | Low | Moderate–High (SIM) | Moderate | Low |
| No endurance limit | Yes | Yes | Yes | Yes | Yes |
| Weld process | FSW preferred | FSW preferred | GTAW | GTAW | GTAW |
| LOX compatible | Yes | Yes | Yes | Yes | Yes |
Historical Case Studies
Space Shuttle External Tank. The ET LH₂ tank (Al 2219-T87, later Al-Li 2195) accumulated 135 flights without a structural fatigue failure in the tank barrel. TPS foam adhesion and COPV integrity were the failure modes that led to the Columbia and Challenger accidents respectively — not propellant tank structural fatigue. This validates the damage-tolerant design approach applied to aluminum cryogenic tankage.
Centaur Upper Stage. Pressure-stabilized 304L stainless steel LOX and LH₂ tanks operational since 1963 with an exemplary flight record. The thin-wall design requires disciplined pressurization management throughout ground operations — loss of pressure causes collapse. Stainless steel enables welding repair on the ground and long shelf-life storage.
SpaceX Falcon 9 COPV. The 2015 CRS-7 and 2016 Amos-6 failures both involved COPVs inside the LOX tank. Solid oxygen accumulating between the aluminum liner and carbon fiber overwrap — caused by thermal differential shrinkage — provided ignition energy. This illustrates that the fatigue and compatibility risk in LOX tanks is not always in the tank wall itself, but in secondary pressure vessels and joints within the tank.
SLS Core Stage Reversion. NASA reverted from Al-Li 2195 to Al 2219 FSW for the SLS Core Stage, citing weldability, cost, and the fact that 2195’s advantages over 2219 are more modest at cryogenic temperatures than at room temperature.
Practical Design Guidance
Match alloy temper to stress state. Over-aged tempers (T73, T7351 for 7xxx; T87, T8 for 2xxx) provide the best balance of SCC resistance, fatigue life, and hydrogen compatibility. Peak-aged tempers (T6) maximize strength but increase susceptibility to grain boundary attack.
Use FSW where possible. Friction stir welding eliminates fusion welding hydrogen pickup, reduces HAZ softening, and produces compressive or near-zero residual stresses compared to GTAW. FSW is now the industry standard joining method for aluminum barrel sections.
Treat the 150–250 K transient as a design load case. For both aluminum and stainless steel, this temperature range is where thermal strain, residual stress, and hydrogen-assisted degradation mechanisms overlap. Fill rates and cool-down sequences should minimize dwell time and thermal gradient in this zone.
Conduct proof pressure testing at room temperature for aluminum. Aluminum’s cryogenic toughness exceeds room-temperature toughness, making room-temperature proof tests conservative with respect to the cryogenic service condition.
Choose stable austenite for high-cycle stainless steel applications. If reusability demands more than a few dozen thermal cycles, metastable grades (304, 316) should be replaced with fully stable alternatives (Nitronic 40, 310S) to avoid cumulative martensite-assisted HE damage.
Characterize EIFS from weld inspection data. Translating NDE probability-of-detection (POD) curves into EIFS distributions — using fracture mechanics back-calculation from fatigue test data — provides a rigorous probabilistic basis for service life prediction.
Closing Thoughts
The juxtaposition of an LH₂ tank and a LOX tank on a single vehicle creates one of the most demanding material environments in engineering: one tank holds a cryogenic fuel that is also the primary embrittling agent, operating at a temperature 20 degrees above absolute zero; the adjacent tank holds a cryogenic oxidizer at slightly warmer temperatures but with its own ignition compatibility requirements. That these propellant systems have flown reliably for sixty years — from Centaur in 1963 to New Glenn’s GS-2 today — is a testament to the rigor of damage-tolerant design, quantitative fractography, and careful material selection.
The key engineering insight is that both aluminum alloys and austenitic stainless steels are fundamentally protected from the worst hydrogen embrittlement mechanisms at their cryogenic operating temperatures by the thermodynamic suppression of hydrogen diffusion. The vulnerabilities lie in the thermal transients on the way to operating temperature, in manufacturing-introduced hydrogen at weld sites, and in the cumulative interaction of thermal fatigue and hydrogen-assisted crack growth over many cycles in reusable vehicle designs. Engineering those vulnerabilities out — through alloy selection, FSW joining, proof testing, and damage-tolerant life management — is the practice of cryogenic structural integrity.