​‌​​‌​

AISI 9310 / UNS G93106 / AMS 6260 / AMS 6265

​‌​​‌​

AISI 9310 is a nickel-chromium-molybdenum alloyed steel processed by carburising (case hardening). Within the alloy steel group it is a grade developed specifically for aircraft engine gears; its UNS designation is G93106. It is also known as E9310.

Its low carbon content (0.10%) makes the material suitable for carburising: the surface is enriched with carbon and hardened while the core stays tough. This structure gives high wear and contact fatigue resistance on the gear tooth flank while preserving impact strength at the tooth root.​‌​​‌​

The 3.25% nickel in its composition supports core toughness, the 1.20% chromium supports hardenability and surface hardness, and the 0.12% molybdenum supports grain boundary strength. The material is usually produced by vacuum arc remelting (Vac-Arc), which lowers inclusion content and so extends fatigue life.

Core properties measured after carburising range from 1069-1289 MPa tensile strength and 331-375 HV core hardness depending on the heat treatment cycle. It is used in aircraft engine gears, transmission components and power transfer parts subject to high cycle fatigue. It is supplied as round bar.​‌​​‌​

​‌​​‌​

What AISI 9310 Is — and Why It Is the Default Aerospace Gear Steel

AISI/SAE 9310 (UNS G93106 / W.Nr. 1.6657 / EN 14NiCrMo13-4) is a low-carbon nickel–chromium–molybdenum carburizing (case-hardening) steel: nominally 0.10 C – 3.25 Ni – 1.20 Cr – 0.12 Mo. It is not sold as a strength steel. It is sold as the composition of two different materials that exist after carburizing — a hard case at 60–62 HRC carrying compressive residual stress, and beneath it a ductile core that still delivers better than 50 % reduction of area.​‌​​‌​

The single sentence that separates 9310 from its rivals is this: this much nickel (3.00–3.50 %) at this little carbon keeps the core simultaneously hardenable and tough. In 8620 nickel is 0.40–0.70 %; core hardenability runs out in heavy section. In 4320 it is 1.65–2.00 %; better, but it does not reach 9310’s core toughness. 9310 is not a chemistry number, it is a toughness choice — and the price of that choice, paid through low carbon and low molybdenum, is weak wear resistance, poor temper resistance and zero corrosion resistance.

Buyers confuse three things most often, and all three cost money on the order line: (1) AMS 6260 and AMS 6265 are not the same steel — the chemistry is nearly identical, the melt route is not; (2) 1.6657 / 14NiCrMo13-4 is NOT a drop-in equivalent of 9310 — the molybdenum and chromium bands differ; (3) 9310 is not an ASME pressure-equipment material — it lives in the AMS/aerospace world, not the code world. Each is opened up below.​‌​​‌​

The one decision rule a buyer should take from this table​‌​​‌​

No temperature problem and parts failing in fatigue: 9310. A temperature problem (loss-of-lube case, continuous oil-in above 150 °C, high pitch-line velocity): Pyrowear 53 or M50NiL. A distortion problem at moderate load: a nitriding steel. A cost problem with thin section: 8620 or 16MnCr5. Mixing these four up is the most common material-selection error in the sector — because they are all called “case-hardening steels” and they solve four different problems.

Standards by Product Form​‌​​‌​

9310’s standards map works completely differently from a corrosion alloy’s. In a Hastelloy there is one ASTM number per product form. In 9310, the ASTM side defines the chemistry and the bar, while the decisive AMS number defines the MELT ROUTE and the cleanliness. If there is no AMS number on your order line, you have not actually ordered an aerospace material.​‌​​‌​

ASME Code Acceptance and Temperature Ceilings — the Honest Answer​‌​​‌​

​‌​​‌​

The most honest thing to publish here is that the expected table is EMPTY. 9310 is not listed as a pressure-retaining material in ASME Section VIII, Section I or the B31 piping codes. The “maximum code temperature 427 °C” lines you read on a Hastelloy or a stainless have no counterpart in 9310. That is not a deficiency; it is looking for the material in the wrong world: 9310 is not a vessel steel but a gear steel, and its governing regime is not ASME but AMS plus AS9100 plus the customer’s (OEM’s) own specifications.

Product Forms With NO Standard — the Commercially Most Valuable Section

This is the section your sales engineers should memorise. 9310’s specification coverage is much narrower than people assume: AMS 6265 covers bar, forgings, forging stock and mechanical tubing. Everything else is a matter of agreement.​‌​​‌​

Chemical Composition​‌​​‌​

There are two different steels in this section, and confusing them is the most expensive 9310 mistake. The US side (AMS/SAE 9310) and the European side (EN 10084 14NiCrMo13-4 / 1.6657) are in the same family but they are not the same chemistry. Read the two tables below side by side.

AMS 6265 versus AMS 6260 — Why It Matters on an Aircraft Gear

This is the section of a 9310 page that saves the most money. The two specifications’ chemistries sit inside one another, which is why most buyers assume they are interchangeable. They are not, and the reason is a single word: INCLUSIONS.​‌​​‌​

The Carburizing Route — 9310’s Real Manufacturing Process​‌​​‌​

Ordering the material is the easy part of 9310. The part’s performance is decided in the six or seven step heat-treatment chain that follows the bar, and every link in that chain changes fatigue life measurably.

Double quenching — when, and why

Carburizing runs at 900–930 °C. That is a higher austenitising temperature than the low-carbon core needs, and a part held there for hours grows its austenite grain. Quenching straight from that temperature (direct quench) is the cheapest and lowest-distortion route, but it leaves coarse martensite packets inherited from coarse austenite.
Double quenching fixes that: the part is slow cooled from carburizing and then reheated and quenched twice — or a single reheat is accepted. The three routes Carpenter itself publishes show exactly this trade-off:
(A) 927 °C / 8 h → direct to oil → 149 °C temper: core 1289 MPa tensile, 1117 MPa yield, 15 % elongation, 51 % reduction of area, 375 HBW.
(B) 927 °C / 8 h → slow cool → oil from 774 °C → 149 °C temper: 1069 MPa tensile, 896 MPa yield, 15.5 % elongation, 52 % RA, 331 HBW.
(C) 927 °C / 8 h → slow cool → oil from 829 °C → 149 °C temper: 1207 MPa tensile, 1069 MPa yield, 16 % elongation, 53 % RA, 363 HBW.
What that says: route (B) costs about 17 % of core strength but refines the grain; route (C) is the balanced point between the two and gives the highest ductility (16 % / 53 %) at a sensible strength. That is why (C)-type routes are common on aerospace gears. Route selection is not a cost decision but a trade between core toughness and distortion.​‌​​‌​

Retained austenite — friend or enemy

Both. In carburized 9310 the case carbon rises to 0.8–1.0 %; together with the nickel this drives M_s and M_f down, and after quenching an appreciable amount of retained austenite remains in the case.
The harm: austenite is soft and lowers case hardness; if it transforms to martensite under stress in service it expands and moves dimensions; if it transforms during grinding it leaves tensile residual stress at the surface.
The benefit: a limited amount of retained austenite can improve surface-fatigue (pitting) behaviour by spreading plastic flow in the contact zone and by blunting a crack tip as it transforms. That is why bearing and gear practice targets not “zero retained austenite” but “controlled retained austenite”.
The control levers: lowering the carburizing carbon potential (not over-carburizing), the cryogenic cycle, and tempering. The warning worth publishing: the target retained-austenite percentage varies with the OEM specification, and no single universal acceptance limit for 9310 could be verified in this study — do not publish a percentage; ask for the customer’s specification.​‌​​‌​

Low-pressure (vacuum) carburizing and gas quenching

Alongside the classic gas carburize plus oil quench route, a second route is spreading in aerospace: low-pressure carburizing (LPC, acetylene) plus high-pressure gas quenching. Advantages: no intergranular oxidation — because there is no oxygen in the atmosphere; cleaner surfaces, less grinding stock, more predictable distortion. The drawback: gas quenching is slower than oil, so hitting core hardness in heavy section gets harder — which is exactly where 9310’s high hardenability earns its keep, taking LPC far more comfortably than 8620 does. Numerical LPC cycle parameters are specific to furnace and part geometry; do not publish a generic recipe.​‌​​‌​

Mechanical Properties

​‌​​‌​

In 9310 a “mechanical property” is never a single number. At least three different things get measured and datasheets confuse them continually: (1) core tensile properties (usually on a pseudo-carburized specimen — one put through the carburizing cycle but given no carbon), (2) case hardness, and (3) fatigue strength measured on an actual gear.​‌​​‌​

Bending and Surface Fatigue — 9310’s Real Selling Argument

A gear steel is not sold on tensile strength; it is sold on fatigue strength. The table below is comparative bending-fatigue data measured by NASA Glenn on the same rig, the same geometry and the same method — and it is the most honest single dataset about 9310, because it reports the scatter alongside the mean.​‌​​‌​

Physical Properties

Physically, 9310 is an ordinary low-alloy steel — and that is good news: its expansion, conductivity and modulus match the housing steel and produce no assembly surprises. The values below are converted imperial data; the source publishes them for a single temperature — no temperature-dependent curve was found.​‌​​‌​

Welding​‌​​‌​

9310 is a weldable steel — but WHEN it is welded is what matters. Carbon is 0.07–0.13 %, which puts it on the good weldability side among low-alloy steels. The problem is not carbon but high hardenability from the nickel and chromium: untempered martensite forms easily in the heat-affected zone.

What actually goes wrong

1. A “small repair weld” on a carburized part. The most common and most expensive mistake. The weld locally tempers the case in the heat-affected zone; hardness drops, compressive residual stress vanishes, and that spot becomes the origin of a fatigue crack. An apparently sound repair produces an early field failure.
2. Skipping stress relief after welding. 9310 is a hardenable steel; weld residual stresses turn into distortion in the subsequent carburizing, and the grinding stock no longer covers it.
3. The weld zone behaving differently in carburizing. If the filler differs in chemistry from the base metal, case depth and case hardness come out different there. Do not use dissimilar filler on a surface that will be carburized.
4. Surface contamination. Oil, cutting fluid, paint and especially phosphate or lead-bearing machining residues cause trouble in both welding and carburizing. Pre-carburizing cleanliness is not cosmetic: a dirty surface means a patchy case depth.​‌​​‌​

Machining

9310 is NOT a free-machining steel, and the reason is not its hardness but its ductility. A high-nickel, low-carbon, ductile matrix produces gummy chips, poor chip breaking and built-up edge (BUE). The manufacturer’s own data rates machinability at 40 % of B1112 (annealed, 179–229 HBW) — that is markedly harder than 8620 and close to 4340.​‌​​‌​

Corrosion and Surface Protection — WHERE IT FAILS​‌​​‌​

​‌​​‌​

The honest heading for this section is: 9310 has NO corrosion resistance. At 1.0–1.4 % chromium it is nowhere near forming a passive film (the stainless threshold is around 10.5 % Cr). 9310 is a low-alloy steel and it rusts like one. That is not a defect but a limit the design accepts — it just goes unwritten on sales pages and therefore produces field surprises.

Frequently Asked Questions​‌​​‌​

Our supplier offered AMS 6260 where we asked for AMS 6265. They say the chemistry is the same. Can we accept it?

The chemistry claim is true; the conclusion is false. The two specifications’ composition bands sit inside one another — a spectrometer cannot tell them apart. What tells them apart is what the spectrometer cannot see: the inclusion population and the dissolved gas.
AMS 6265 permits only VAR (vacuum consumable electrode remelt). AMS 6260 says aircraft quality and leaves air melting open. AMS 6267 also permits ESR — so the three specifications are three different cleanliness levels.
Why that matters, in one sentence: in a carburized gear tooth the fatigue crack usually starts not at the surface but at a subsurface oxide or aluminate inclusion. The case is in compression, which makes surface initiation hard; so the weak link is the largest inclusion at the depth where compression has run out but stress is still high. Fatigue life is set not by the average inclusion but by the worst one — and the size of the worst one is set by the melt route.
Practical outcome: if the part is sized by fatigue (rotating, flight-critical, life-limited), do not accept AMS 6260. If the part is sized by static strength (ground equipment, low-criticality drive, fixturing), AMS 6260 is both acceptable and markedly cheaper. The decision criterion is not the price difference but which damage mechanism sized the part. And in every case, call out the AMS 2300/2301/2304 inspection level separately: it is as decisive as the melt route.​‌​​‌​

Our European supplier is shipping 1.6657 / 14NiCrMo13-4 and calling it equivalent to 9310. Is it?

It is not “equivalent”, it is “closest” — and the differences sit exactly where gear performance is decided.
The only major element that genuinely overlaps is nickel: 3.00–3.50 % in both. Everything else diverges. Carbon: 0.11–0.17 % in 1.6657 against 0.07–0.13 % in 9310 — almost 50 % more carbon at the top of the band, meaning a harder and less ductile core. Molybdenum: 0.20–0.30 % against 0.08–0.15 % — roughly double; the hardenability curve and the temper resistance come out different. Chromium: 0.80–1.10 % against 1.00–1.40 % — lower; case carbide behaviour will not match. Sulphur ceiling: ≤0.035 % against AMS 6265’s ≤0.015 % — more than double; and a sulphide inclusion is a direct fatigue initiator in a carburized gear.
On top of that comes the melt-route difference: AMS 6265 mandates VAR; EN 10084 mandates no melt route at all. So the gap between the two is not only chemistry but a cleanliness regime.
The practical answer: for general machine gearing, an industrial gearbox or any non-aerospace application, 1.6657 is a perfectly good carburizing steel and is used without hesitation. Using 1.6657 in place of AMS 6265 on an aerospace gear is not a substitution but a DEVIATION, and it cannot be done without the OEM’s written approval. If your supplier says “equivalent”, ask in writing which specification they are equivalent to — if no answer comes, the answer has already come.​‌​​‌​

Our 9310 gear is pitting. Should we upgrade the material, or is something else wrong?

It is almost never the material. Pitting (surface fatigue) is solved by looking at four things before material quality, and none of the four is specific to 9310.
1. Oil film thickness (the λ ratio). Pitting happens where there is metal-to-metal contact. Oil viscosity, temperature and surface roughness together set λ. If λ < 1, changing material will not help; you will buy the same damage in a more expensive steel.
2. Surface finish and grinding quality. Grinding marks create local pressure peaks in the contact. Superfinishing or honing can extend pitting life more than a material change would.
3. Residual stress. The NASA measurement is unambiguous: shot peening raised pitting life on carburized 9310 gears by a factor of 1.6, and the gain comes not from the surface but from the 40 % increase in compressive stress at 178 µm depth — because that is exactly where a pitting crack starts. Before upgrading material on an unpeened gear, try peening. And make sure peening comes AFTER grinding; in the wrong order the gain is zero.
4. Case depth and core support. If the case is too thin, the maximum shear stress falls below the case and the damage arrives not as pitting but as subcase fatigue / case crushing. That is a completely different problem whose fix is a deeper case or a harder core, not cleaner steel. Look at the pit morphology: small, shallow, scattered pits are classic pitting; large areas where slabs of case have lifted off are a subcase problem.
When is a material upgrade right? When you have a temperature problem (Pyrowear 53, M50NiL), or when all four of the above have been fixed and life still falls short. Change the process first, the material second.​‌​​‌​

Our drawing calls out both 9310 and 8620. Which is right, and are they interchangeable?

They are not interchangeable, and which one is right is decided by the part’s SECTION and its load type.
Although the two chemistries look like family, the nickel content is decisive: 0.40–0.70 % in 8620 against 3.00–3.50 % in 9310 — five to seven times. Nickel does two things: it raises hardenability (the core can form martensite even in thick section) and it raises core toughness.
The practical dividing line is section. In thin section (small module, small pinion) 8620’s core hardens fully in oil anyway; 9310’s hardenability advantage never materialises and 8620 is both cheaper and much easier to machine (machinability around 60–65 % against 40 % for 9310). In thick section 8620’s core cannot harden through; the support beneath the case weakens and the risk of subcase crushing appears. Here 9310 makes a real, measurable difference.
The second dividing line is impact. Core toughness is the last defence against a tooth breaking off at the root. In a drive with shock loading, reversed loading or an overload case, 9310’s high-nickel core is a concrete safety margin.
If a drawing names two materials, it is almost always a revision leftover. The fix is not to pick a material but to get the drawing corrected — because the heat-treat specification, the case-depth target and the inspection level all change with the material, and those lines are probably inconsistent too. Two different steels mean two different carburizing recipes; running one furnace cycle on both takes both outside target.​‌​​‌​

Common datasheet errors — check before you order

1. UNS confusion: G93100 or G93106? For aerospace-quality 9310 (E9310) the correct UNS is G93106. Many pages print G93100. Both circulate in the literature; put the AMS number on the order line, not the UNS — that is what governs.
2. The equation “1.6657 = 9310”. WRONG. In 14NiCrMo13-4 carbon is 0.11–0.17 % (9310: 0.07–0.13), molybdenum 0.20–0.30 % (9310: 0.08–0.15), chromium 0.80–1.10 % (9310: 1.00–1.40), sulphur ceiling ≤0.035 % (AMS 6265: ≤0.015). It is the closest grade, not an equivalent.
3. ASTM A331 listed as if still current. A331 was withdrawn in 2004. Do not cite it on a new order.
4. Treating AMS 6260 and AMS 6265 as equivalent. The composition band is the same; the melt route is not. 6265 is VAR only, 6267 is ESR or VAR, and 6260 permits air melting. Three specifications, three prices, three fatigue scatters.
5. Presenting core tensile strength as the material’s strength. 1289 MPa is a pseudo-carburized core value. The surface of a carburized gear never sees a tensile test; the case is 60–62 HRC and is not described by a tensile figure.
6. Publishing a single “hardness” value. 9310 has at least three: annealed 179–229 HBW, core 331–363 HBW (≈33–40 HRC), case 60–62 HRC. Any table that does not say which one it means is unusable.
7. Quoting a “maximum service temperature” with no source. 9310’s ceiling is set by the tempering temperature (121–177 °C). The 175 °C given by one industry source is plausible but single-sourced, and the “x % hardness loss per 10 °C” style degradation rates could not be independently verified.
8. A factor-of-ten spread in the copper ceiling. One distributor page prints Cu ≤0.035 %, an aerospace mill page prints Cu ≤0.35 %. Both cannot be right — confirm from the current AMS 6265 revision.
9. A specification called “MIL-S-93030”. The correct number is most likely MIL-S-83030; “93030” appears to have propagated as a typo. Do not copy a military specification number from a single web page.
10. Publishing “9310 is weldable” without qualification. The correct statement is: it is weldable BEFORE carburizing. A weld on a carburized gear tooth locally tempers the case and creates a fatigue initiator exactly there.
11. Inventing a plate/sheet specification. AMS 6265 does not cover plate — it covers bar, forgings, forging stock and mechanical tubing. “9310 plate” is an agreement product.
12. Assigning an ASME P number. 9310 is not listed as an ASME pressure-boundary material; no P-No. assignment could be verified in this study. Do not publish a P number.
13. Confusing mechanical tubing with pressure pipe. AMS 6265 covers mechanical tubing. It is not a pressure service pipe specification and cannot be used as one.
14. Getting the peening/grinding sequence wrong. Peening comes AFTER grinding. Peening applied in the wrong order has no measurable benefit — the grinding removes the compressive layer that was just created.
15. Presenting the physical properties as if they were temperature-dependent. The published density, modulus, conductivity and expansion values are for a single temperature; no temperature-dependent curves for 9310 were found in this study.​‌​​‌​

​‌​​‌​