​‌​​‌​

Maraging 350 / AMS 6515

​‌​​‌​

Maraging 350 is a special high-alloy steel with low carbon and a high nickel and cobalt content. It is the highest strength member of the maraging family within the nickel alloy group, one step above the 250 and 300 grades.

What separates maraging steels from other high strength steels is that they take their hardness from intermetallic precipitates rather than from carbon. The carbon content is held below 0.03%, and strength comes from precipitation hardening in the nickel-molybdenum-titanium system.​‌​​‌​

The practical consequence is this: the material is easily machined in the soft martensitic condition and is then aged at a low temperature of around 480-510 °C. That low temperature reduces post-machining distortion to almost nothing, which is the decisive advantage for complex parts held to tight tolerances. Weldability is also very good for a steel in this strength class.

It is used in aerospace and defence for rocket motor cases, landing gear parts and critical fasteners; in tooling for injection and die casting moulds working under high load; and in motorsport for shafts and drivetrain components.​‌​​‌​

​‌​​‌​

What Maraging 350 Is — and Why It Is the Extreme End of the Family

​‌​​‌​

​‌​​‌​

​‌​​‌​

Maraging 350 (18Ni-350 / C350 / Vascomax C350 / ATI C-350) is the highest-strength and lowest-toughness step of the 18Ni maraging family. The 350 in the name denotes the nominal 350 ksi (2413 MPa) tensile class in the aged condition; the measured 0.2% yield strength lies in the 2275–2365 MPa band. It hardens not through carbon but through intermetallic precipitation; the low carbon that defines the family (C ≤0.03%), quench-free hardening and freedom from distortion all apply here too. But this grade is a toughness bargain, and we are obliged to say so plainly.

The distinguishing property in one sentence: Maraging 350 exists for duties where the absolute ceiling of tensile strength is what matters, where crack tolerance does not govern the design, and where the environment is dry. Put in the wrong place, it does not forgive what the other grades in the family forgive.​‌​​‌​

Where it sits in the family: cobalt and titanium jump here​‌​​‌​

The design logic of the 18Ni family is visible in one line: cobalt and titanium rise together, molybdenum stays nearly constant, nickel and carbon never change. Titanium is the direct fuel for the Ni₃Ti precipitate; cobalt forms no precipitate of its own but lowers the solubility of molybdenum in the matrix and thereby forces Mo to precipitate as Ni₃Mo and Fe₂Mo. In the 350 grade both elements jump — and the bill is paid directly out of toughness.

THE TOUGHNESS PENALTY — the most important section on this page

Showing the strength table when selling Maraging 350 is easy; the engineering responsibility is to show the toughness table at the same size. The values below come from a single defence report in which all four grades were measured in the same laboratory, in the same programme, so they are directly comparable. (They are K_Q values, i.e. provisional fracture-toughness results whose validity conditions were not separately confirmed — read them as a ratio between grades, not as absolute design values.)​‌​​‌​

Why there is no carbon, and how the mechanism is pushed in the 350​‌​​‌​

A conventional alloy steel takes its hardness from carbon trapped in the martensite lattice by quenching, and pays for it in brittleness, quench cracking and hydrogen cracking. In maraging steel carbon is deliberately held at ≤0.03%; the structure that air cools from high temperature is not a hard carbon martensite but a soft, ductile iron-nickel martensite at 30–35 HRC. All of the strength is added afterwards, during ageing, by nanoscale Ni₃Mo, Ni₃Ti and Fe₂Mo intermetallics precipitating on the high dislocation density of the martensite itself.

In the 350 grade this mechanism is pushed to its limit. Raising titanium to 1.30–1.60% markedly increases the volume fraction of Ni₃Ti precipitates; raising cobalt to 12% leaves even less molybdenum dissolved in the matrix. The result is a very dense, very fine precipitate distribution in which dislocation motion is almost completely locked. What is gained is a yield strength of order 2300 MPa; what is lost is the ability to absorb energy by plastic deformation at a crack tip — and that is exactly what the collapse in fracture toughness and Charpy energy measures. High titanium also increases the tendency to form coarse Ti-rich particles at grain boundaries, which act as crack initiators — that is why double-vacuum melting is written into the specification for this grade.​‌​​‌​

Overageing and reverted austenite. If the ageing temperature or time is exceeded, the precipitates coarsen and austenite re-forms (reverted austenite) in nickel-rich regions. Academic work on 18Ni-350 has measured this behaviour in detail: during overageing at 640 °C the volume fraction of austenite rises with time; yield and tensile strength fall while tensile ductility rises. Impact toughness behaves in two stages: in the early stages of overageing a small amount of reverted austenite appears beneficial, but with prolonged overageing the coarsened titanium-nickel intermetallics initiate cracks and produce severe embrittlement. The practical consequence: at this grade, temperature control of the ageing furnace is a quality record.

Standards by Product Form​‌​​‌​

The standards landscape for Maraging 350 is NARROWER than for the other grades in the family, and that is a commercial fact the sales side needs to know. The 250 grade has a dedicated AMS number for sheet and plate (AMS 6520); so does the 300 grade (AMS 6521). For the 350 grade, none was found.​‌​​‌​

ASME Code Acceptance and Pressure-Equipment Status​‌​​‌​

The short answer: Maraging 350 is not an ASME pressure-vessel code material — and at this grade that is not merely a missing document, it is also the engineering-correct outcome. Code design rests on assumptions of ductility and damage tolerance; a fracture toughness of order 39 MPa√m is far below the behaviour the code expects of pressure-boundary materials.

Product Forms With NO Standard — the Commercially Valuable Section

This is the section your sales engineers should memorise. The list of forms that can honestly be sold “to a standard” for Maraging 350 is short: bar, forgings, rings, tubing (AMS 6515) and plate, sheet, strip, extrusions (via the MIL-S-46850 route only). Everything else is sold to a mill specification and a customer drawing.​‌​​‌​

Chemical Composition​‌​​‌​

The nominal mill composition is 18.5Ni – 12.0Co – 4.8Mo – 1.40Ti – 0.10Al, balance iron. The AMS 6515 title carries these nominal values directly. The band values are compiled from supplier specification tables.

Three lines that actually matter on a certificate. (1) Titanium. The difference between 1.30% and 1.60% is not small at this grade: the top of the band means higher strength, the bottom means higher toughness. If toughness is critical, narrow the band and write it into the order. (2) Carbon and nitrogen. Working together with high titanium, these two form coarse TiC/TiN particles at grain boundaries, and in the 350 those are crack initiation sites. (3) Melt route. AMS 6515 requires double vacuum (VIM + VAR). That is not negotiable at this grade — the link between inclusion cleanliness and measured toughness is far tighter here than in the other grades of the family.

Mechanical Properties​‌​​‌​

Three separate sets of numbers circulate in this section and must not be mixed: specification minima, typical mill values and independent laboratory measurements. And the sources diverge badly on elongation — all of them are given below exactly as published.

The commercial conclusion to draw from this table. In Maraging 350 the gap between minimum and typical is narrow and the margin is small: minimum yield 2275 MPa against a typical 2317 MPa. The alloy runs at the edge of its own specification. The same is even more marked in elongation — a 2.8% minimum against a 6.2% typical. A small deviation in the ageing cycle can put a part outside specification. At this grade, manage heat treatment with furnace records, thermocouple calibration and witness coupons.

Physical Properties​‌​​‌​

The physical properties of the 350 grade differ measurably from the 250, and the difference is real — 12% cobalt raises both density and elastic modulus. The values below come from the same mill datasheet family and are therefore directly comparable with the 250.

Heat Treatment and Thermal Stability

A common misconception needs correcting here: Maraging 350 is not aged LONGER than the 250 — it may be aged HOTTER. Mill datasheets publish two alternative cycles, and the higher-temperature one is the SHORTER of the two. In the same defence programme the 250 grade was aged at 900 °F for 3 hours and the 350 grade at 950 °F for 3 hours — same time, different temperature.​‌​​‌​

Welding​‌​​‌​

The family’s greatest advantage — needing no preheat — applies to the 350 grade too, but here we must be honest: being weldable is not the same as being reliable in the welded condition. With carbon at ≤0.03% no hard carbon martensite forms in the HAZ, the risk of hydrogen cracking is practically absent, and mill datasheets state “good weldability without preheating or post heating” and “good repair weldability”. But when the parent metal’s fracture toughness is already ~39 MPa√m, every loss of toughness in the HAZ has proportionally far heavier consequences.

Machining

The rule is the same and matters even more here: do the roughing and every possible finishing operation in the solution-annealed (soft, 30–35 HRC) condition, then age. After ageing the material is at 55–60 HRC, the hardest state in the family to machine. The parameters below are the mill datasheet’s annealed-condition starting points, given for the family generally.​‌​​‌​

Nitriding​‌​​‌​

Maraging steels can be nitrided, and this is used in tooling applications that need wear resistance. The metallurgically convenient point: nitriding temperatures overlap the alloy’s ageing band, so nitriding and ageing can be combined in a single cycle. Honest limits: (1) no verified case hardness or case depth value specific to the 350 grade was found that could be published here — the available academic work is mainly on 18Ni-250 and 18Ni-300. (2) Adding a hard, brittle nitride case to a material whose core toughness is already low can raise notch sensitivity; do not skip that assessment on fatigue-loaded parts. Case data must be obtained from the treatment house on a part-by-part basis.

Corrosion — Where It Is Good, WHERE IT FAILS​‌​​‌​

This section must be published at least as prominently as the strength section. Maraging 350 is the family member MOST susceptible to stress-corrosion cracking, and that is not a laboratory curiosity but a measured failure mode.​‌​​‌​

The basic fact: it is not stainless

Maraging 350 is not a stainless steel. Chromium is a residual, not an alloying addition; there is no passive oxide film. Its corrosion resistance is comparable to that of an ordinary low-alloy martensitic steel. Bare parts will rust in humid, marine and industrial atmospheres. For any service outside a dry, controlled environment, plan on conventional protection: plating, phosphating, paint, dry-film lubricant or an oil or grease film. Suppliers selling the product as “maraging 350 stainless steel” — some of them carrying that phrase in their page addresses — are metallurgically wrong.​‌​​‌​

Where it is relatively good: general corrosion rate and polishability

Despite carrying no chromium, mill documents report better pitting and corrosion resistance than common tool steels and an excellent polish. General rusting advances uniformly; localised penetrating attack is not typical — commercially meaningful for tooling surfaces.​‌​​‌​

WHERE IT FAILS — stress-corrosion cracking

The sentence “maraging steels resist stress corrosion and hydrogen embrittlement” is not merely misleading when applied to the 350 grade — it is dangerous. An independent NASA study tested all four grades side by side and the result is unambiguous: the lowest-strength grade (200) was the least susceptible and the highest-strength grade (350) the MOST susceptible.​‌​​‌​

Hydrogen embrittlement​‌​​‌​

Maraging 350 is susceptible to hydrogen embrittlement, and — as in every steel at a 2300 MPa yield — that susceptibility rises sharply with strength. The practical consequences bear directly on your manufacturing route: acid pickling, electroplating (cadmium and zinc especially), electropolishing and cathodic protection all charge hydrogen into the part. For plated parts a post-plating hydrogen bake-out must be treated as mandatory, planned at a temperature that does not disturb the aged condition (the classical 190–200 °C range sits far below the 482–510 °C ageing temperature). No verified numerical hydrogen threshold was found for the 350 grade in this research — the plating and bake-out specification should be written jointly with the customer, and mechanical plating or a plating-free solution should be preferred where possible.

Where it must not be used​‌​​‌​

(1) Any acid or chemical-process service — there is no chromium. (2) Under tensile stress in permanent salt-spray, marine immersion or high-humidity service — the measured SCC behaviour forbids it. (3) H₂S (sour) service — it is not listed under NACE MR0175 / ISO 15156. (4) In damage-tolerant structures where crack tolerance governs the design — 39 MPa√m is too low for that job. (5) In welded primary structural members — the HAZ toughness gap creates unacceptable risk at this grade. (6) Anywhere that must be non-magnetic. (7) In continuous service above ~450 °C.

Frequently Asked Questions​‌​​‌​

Do we really need 350, or will 300 or 250 do?

This is the most important question on this page, and the honest answer is usually “you do not need 350”.
The strength difference is real but smaller than you might think: the 300 grade gives ~2068 MPa yield and the 350 grade ~2275–2365 MPa yield — a difference of about 12–14%. Against the 250 the difference is 35%.
What you pay for it is disproportionately large. Measured in the same laboratory: fracture toughness 115–124 MPa√m for the 250, ~77 for the 300, ~39 for the 350; Charpy energy 41 J → 24 J → 12 J. In other words, going from 300 to 350 you give up half the toughness for 12% more strength.
And on the corrosion side the difference is not a step but a change of class. In the NASA study the 250 did not fail in salt water up to 90% of yield; the 300 failed at 75%; and the 350 failed at 55%, in under two days.
Decision rule: if your design is purely stress-limited, the part is solid and machined, the environment is dry and protected, crack tolerance does not govern, and weight genuinely is critical — the 350 is the right choice. If any one of those does not hold, Maraging 250 or the 300 grade is better engineering and cheaper. A bigger number is not automatically an improvement.​‌​​‌​

Is ageing longer for 350 than for 250? How should we write the cycle?

Contrary to the common belief, it is not longer — it is HOTTER.
Mill datasheets give two alternatives for the 350: 6 hours at 482–496 °C (900–925 °F) (→ 55–60 HRC) or 3 hours at 510 °C (950 °F) (→ 56–60 HRC). The second cycle is hotter and shorter. In the same defence programme the 250 grade was aged at 900 °F for 3 hours and the 350 at 950 °F for 3 hours — same time, different temperature.
Why: the very high titanium of the 350 starts Ni₃Ti precipitation earlier and faster; holding too long coarsens the precipitates and moves the material towards the reverted austenite window. A shorter hold at a higher temperature completes the reaction while limiting coarsening.
How to write the cycle: (1) Never omit the solution anneal — 815–830 °C, 1 hour per inch, air cool. (2) Read from the customer specification which of the two ageing cycles applies; they are not equivalent, and even where the hardness band is similar the precipitate distribution and toughness differ. (3) Use witness coupons. At this grade the margin between minimum and typical is narrow (minimum yield 2275 MPa, typical 2317 MPa) and a small furnace deviation can put the part out of specification. (4) Do not re-age an already aged part; that is a route into overageing. Resetting requires a fresh solution anneal followed by a fresh age.​‌​​‌​

Is 350 stainless, and how should corrosion protection be planned?

No — and at this grade the corrosion plan is a design requirement, not a preference.
In Maraging 350 chromium is not an alloying addition; there is no passive oxide film. Its corrosion resistance is comparable to an ordinary low-alloy martensitic steel, not to a 300- or 400-series stainless. That some suppliers carry “maraging stainless steel” in their page titles and even in their page addresses is metallurgically wrong.
The real issue is not general rusting but stress-corrosion cracking. Independent measurement: the 350 grade failed in salt water at only 55% of its yield strength, in under two days. Another finding of the same study is more disturbing still: air at 98% relative humidity proved more aggressive than salt-water immersion. You do not need to be at the coast — a humid warehouse is enough.
Practical plan: (1) Do not leave the part under sustained tensile stress; account for assembly preloads and residual stresses. (2) Plate it — but electroplating carries a hydrogen-embrittlement risk; make the post-plating bake-out mandatory, or prefer routes that do not charge hydrogen, such as mechanical plating, paint or dry-film lubricant. (3) Write a handling and storage specification: humidity control, oil film, VCI packaging. (4) Define an inspection interval. (5) If the environment really is humid or salt-laden, consider changing material — AerMet 100 is markedly better in that duty.​‌​​‌​

How should we compare Maraging 350 with AerMet 100 and 300M?

These three sit on the same shelf but do not do the same job. The dividing line is not strength — it is toughness and stress-corrosion behaviour.
Maraging 350 is by a clear margin the strongest: yield ~2275–2365 MPa, tensile ~2350–2430 MPa, 55–60 HRC. In return: K_Q ~39 MPa√m, Charpy ~12 J, and cracking in salt water at 55% of yield. Its advantages: quench-free hardening, freedom from distortion, welding without preheat, simple heat treatment, easy machining in the annealed condition.
AerMet 100 (0.23C – 13.4Co – 11.1Ni – 3.1Cr – 1.2Mo) is a carbon-bearing secondary-hardening steel: yield 1724 MPa, tensile 1965 MPa, elongation 14%, reduction of area 65%, K_Ic 126 MPa√m, K_ISCC 88 MPa√m. It delivers about 75% of the 350’s yield with more than three times the fracture toughness. The price: quench plus deep freeze (−73 °C) plus 5 hours at 482 °C, so it distorts; and its carbide content makes it harder to machine. Where damage tolerance and a salt environment are involved, AerMet 100 is clearly the right answer.
300M (0.40C – 1.6Si – 0.8Cr – 1.8Ni – 0.4Mo – 0.07V) is the classic quenched-and-tempered landing-gear steel: the AMS 6417 minima are 1862 MPa tensile / 1517 MPa yield, 8% elongation, 30% reduction of area, ≥52 HRC. Well below the 350 on strength; being oil quenched it is the most distortion-prone and the most hydrogen-sensitive option. In return it is far cheaper. 300M is a cost-performance choice.
Decision rule in one sentence: for the absolute tensile ceiling and distortion-free manufacture, Maraging 350; for damage tolerance in a salt environment, AerMet 100; where budget dominates and the part is a solid forging, 300M; for a balanced structural material, Maraging 250.​‌​​‌​

Common Datasheet Errors and Traps

1. UNS number confusion — the biggest single trap on this alloy. Three numbers circulate: K93160, K93540 and K93120. K93120 belongs to Maraging 300 and using it for the 350 is simply wrong — yet some aerospace supplier pages are headed “Maraging 350 AMS 6515 UNS K93120”. Sources also diverge between K93160 and K93540. On orders and certificates, rely on AMS 6515 plus the chemistry, not on the UNS number.
2. Yield strength given as ~2400 MPa — WRONG, that is the tensile strength. The published values are yield 2275–2365 MPa and tensile 2350–2430 MPa. The name “350” refers to the 350 ksi = 2413 MPa tensile class. Even a widely used encyclopaedia source lists “yield ~2413 MPa” for the 350; that figure is the nominal tensile, not a measured yield.
3. AMS 6515 presented as a sheet/plate specification — WRONG. Its title is “Bars, Forgings, Tubing, and Rings”. There is no verified AMS sheet or plate number for the 350 (AMS 6520 exists for the 250, AMS 6521 for the 300). For plate, the only verified route is MIL-S-46850.
4. A 350 grade attributed to ASTM A538 — WRONG. A538 contains only Grade A (200), Grade B (250) and Grade C (300); there is no 350 grade. And A538 is in any case an old, no longer active document.
5. ASTM A579 presented as current — WRONG. A579/A579M was withdrawn in May 2024 with no replacement. In addition, which grade number in the 71–75 range corresponds to the 350 could not be independently confirmed.
6. A W.Nr. or EN name is invented. The 250 has 1.6359 / X2NiCoMo18-8-5; no confirmed European equivalent was found for the 350. Do not put a number you cannot verify on a certificate.
7. Sold as “Maraging 350 stainless steel” — metallurgically WRONG. Chromium is at residual level and there is no passive film. Some suppliers carry “maraging-350-stainless-steel” even in their product-page addresses.
8. “Maraging steels resist stress corrosion and hydrogen embrittlement” repeated for the 350. The measured reality: the 350 fails in salt water at only 55% of yield in under two days and is the most susceptible member of the family. Using that sentence without qualification on a 350 page is not defensible engineering.
9. A single elongation figure published. Published values scatter across 2.8% · 4.0% · 6.2% · 8.0% (differences of gauge length, orientation, section and cycle). Quoting one number without stating the condition is misleading.
10. “350 is aged longer than 250” — WRONG. The mill cycles are 482–496 °C for 6 hours or 510 °C for 3 hours; the second is hotter and shorter. In the same defence programme the 250 was aged at 900 °F/3 h and the 350 at 950 °F/3 h.
11. The 250 grade’s physical properties copied across to the 350. Density, modulus and expansion coefficient differ (8.08 vs 8.0 g/cm³ · 200 vs 186 GPa · 11.3 vs 10.1 × 10⁻⁶/°C). Moreover conductivity, specific heat, resistivity and melting range could not be verified for the 350 — do not fill those rows with the 250 values.
12. The MIL-S-46850 toughness minimum applied wrongly. One revision of the specification requires a minimum fracture toughness of 50 ksi√in for the 300 class. The measured 350 value (35.2 ksi√in) is below that — one grade’s minimum cannot be applied to another. Read the minimum that applies to the 350 from your own specification revision.
13. Export control skipped. The gas-centrifuge control text explicitly names “maraging steel capable of an ultimate tensile strength of 1.95 GPa or more”. The aged tensile strength of Maraging 350 (~2.35–2.43 GPa) is well above that threshold — it is the grade in this family that is most unambiguously caught by export control. Thresholds vary between regimes, and some lists apply lower thresholds to sheet, plate and tube forms (the exact figure could not be independently verified in this research). Always have end-use statements, export licences and destination-country documentation checked for international shipments.​‌​​‌​

​‌​​‌​