​‌​​‌​

AISI 310 / (1.4845) / UNS S31000 / AMS 5521 / AMS 5572

Corrosion resistance: AISI 310 has a very strong character where temperature is concerned, and retains its corrosion resistance even at high temperatures. The highest temperature at which the steel can continue to work without corrosion occurring is set at 1100 °C.​‌​​‌​

Temperature capability: Under oxidising conditions the maximum temperature for continuous service is 1100 °C, while for intermittent service 1030 °C is recommended. Its temperature capability is higher than that of AISI 316.

Weldability: In terms of weldability, the corrosion resistance of grade 310 is good. To achieve the best weldability standards, 310H with the lowest carbon content can be specified.​‌​​‌​

Machinability: AISI 310 stainless steel requires softening techniques in order to be machined.

Heat treatment: The annealing temperature is between 1040 and 1150 °C, followed by rapid cooling for maximum corrosion resistance. It is recommended that this heat treatment be repeated every 1000 hours in service above 650 °C in order to maintain softness. It cannot be hardened by heat treatment.​‌​​‌​

Applications: AISI 310 is used under continuous operating conditions in applications requiring high temperature capability. The steel can cope with oxidation and corrosive effects even at very high temperatures. Example applications: furnaces and heating elements (industrial heaters in particular); refractory linings and combustion zone components; heat exchangers and recuperators; hot air generators and gas turbine components.

AISI 310 is an ideal material for industrial applications requiring high temperature capability and oxidation resistance. Thanks to its high chromium and nickel content, this steel copes with hot environments and heavy chemical processes and is widely used in areas such as furnaces, chemical processing equipment and power generation. Under aggressive corrosion conditions such as chloride-bearing environments, however, alternatives such as AISI 316 may be preferred.​‌​​‌​

​‌​​‌​

What AISI 310 / 310S / 310H Are — the Only Difference Is Carbon

The AISI 310 family (UNS S31000 / S31008 / S31009, European W.Nr. 1.4845, EN name X8CrNi25-21) is the heat-resisting branch of the austenitic stainless family, built on a nominal 25 % chromium – 20 % nickel composition. 304 and 316 were designed for aqueous service; 310 was designed for high temperature in air. The chromium feeds scale resistance; the nickel offsets the ferrite and sigma tendency that much chromium would otherwise cause, keeping the structure austenitic even above 1000 °C.​‌​​‌​

The one-sentence definition: in air, in a sulphur-free, non-carburizing atmosphere, where thermal cycling is moderate, 310 is the cheapest iron-based material that survives long-term service near 1000 °C. Every condition in that sentence carries load. With sulphur present 310 loses quickly; in a carburizing atmosphere it is only fair; if the part heats and cools daily the scale spalls off; and if it dwells in the 600–950 °C band it loses its room-temperature toughness. That last item is the single most important fact about 310, and datasheets almost never highlight it.

310 · 310S · 310H — three different carbon decisions​‌​​‌​

Chromium, nickel, manganese, phosphorus and sulphur bands are identical in all three. The split is in carbon, and for 310H also in silicon. Carbon does two opposite jobs at once: it raises creep strength and it sets the stage for sensitization. The three grades are three answers to that dilemma.

The European side: 1.4845 is not “the same thing as 310S”

The neighbours — honest positioning

Standards by Product Form

​‌​​‌​

​‌​​‌​

This table is the family’s commercial strength: unlike 314, 310S has a named ASTM/ASME specification in almost every product form. The grade name alone is not an order — the same grade is sold in plate, bar, pipe, tube and forgings under different documents with different acceptance criteria.

ASME Code Acceptance and Maximum Code Temperatures

​‌​​‌​

The numbers in this section are CODE LIMITS, not material capability. Confusing the two is the most expensive misunderstanding about 310: the datasheet phrase “oxidation resistance to 2000 °F (1093 °C)” gets lifted and used as a design temperature. That phrase is about when scaling becomes unacceptable; the code is about allowable stress. The two diverge by more than 250 °C.​‌​​‌​

MAXIMUM SERVICE TEMPERATURE — “Continuous” and “Intermittent” Are Not the Same​‌​​‌​

These two numbers are the most frequently confused values on 310 datasheets. Two publishing traditions exist, they do not agree, and one of them is counter-intuitive and therefore constantly misread.

Product Forms Whose Scope Is Narrower Than Assumed

Whether “there is a standard” depends on the product form, and the 310 family has gaps. These are procurement’s most expensive surprises, because they are normally discovered after the part has been drawn.​‌​​‌​

Chemical Composition​‌​​‌​

ASTM and EN are given separately; tables that merge them into a single “310 chemistry” are misleading. Typical heat analysis is shown as well, because real material is produced near one end of the band, not in the middle.

Mechanical Properties​‌​​‌​

​‌​​‌​

The gap between specified minima and typical values is large here. Published standard values are the worst acceptable case; real plate comes in 20–45 % above them. Design to the minimum. Also, 310 cannot be hardened by heat treatment; its only strengthening route is cold work, and that disappears at the first anneal in service.

Elevated-temperature tensile values. The table below is typical data from short-term tensile tests — it is not creep data. Above 550 °C the life of a part is governed not by this table but by creep and rupture data. Missing that distinction is the most basic error in high-temperature design.

Physical Properties​‌​​‌​

Two numbers dominate design in 310: high thermal expansion and low thermal conductivity. Together they produce distortion and hot cracking in welding and thermal stress in service. A furnace part designed with carbon-steel habits fatigues itself on every cycle in 310.

Heat Treatment and Thermal Stability

310 cannot be hardened by heat treatment. Its only heat treatment is the solution anneal, and its purpose is not to add strength but to clean the structure: dissolve carbides and second phases back into the austenite, then cool before they can precipitate again.​‌​​‌​

SIGMA PHASE — the Single Most Important Fact About 310​‌​​‌​

Sigma (σ) is a hard, brittle iron–chromium intermetallic that nucleates at grain boundaries in high-chromium austenitic steels. 310 is almost an ideal alloy for sigma: 25 % chromium feeds it, silicon (where present) accelerates it, and 310’s service takes place inside the very band where sigma precipitates.

Its most dangerous trait is that the damage is invisible at temperature. A sigma-loaded 310 part is still ductile at 900 °C and runs normally. The failure comes when the plant stops and the part reaches room temperature: impact toughness and elongation have collapsed. A maintenance technician tries to straighten a basket with a hammer and the part shatters like glass. 310 parts that have run hot must not be impacted or forced at room temperature.​‌​​‌​

Welding​‌​​‌​

310 is weldable but not easy to weld, for one metallurgical reason: 310 weld metal is FULLY austenitic. The weld metal of 304 and 316 contains 3–10 % delta ferrite, which dissolves sulphur and phosphorus during solidification and prevents hot cracking. 310 has no such safety valve. Add high thermal expansion and low thermal conductivity and the weld becomes prone to hot cracking.

Machining

Machining 310 is like 304, only harder. The cause is the same chemistry: high nickel plus high chromium equals gummy chips, strong work hardening and poor heat conduction. Heat accumulates in the tool, chips weld to the cutting edge, the uncut surface work-hardens and the next pass enters hardened material. The most common mistake is taking light, fast passes — that means rubbing in the work-hardened layer.​‌​​‌​

Corrosion — Where It Works, Where It FAILS​‌​​‌​

​‌​​‌​

310’s corrosion story is split in two, and datasheets almost never make the split clearly: near-excellent in dry hot gas; ordinary and in places poor in aqueous service. The sentence “310 is highly alloyed, therefore more corrosion resistant” is wrong.

High-temperature oxidation — what 310 is actually for​‌​​‌​

25 % chromium builds a dense, adherent, slow-growing Cr₂O₃ layer that is markedly more stable than that of 304 or 316. One producer reports very high oxidation resistance in a test cycling between 600 and 1000 °C with 3000 h total dwell at 1000 °C. But the protection is conditional: if the layer spalls or is damaged, the metal beneath has to give up chromium, and every renewal consumes chromium. 310’s life is the life of its remaining chromium reserve — which is why thin 310 sections die disproportionately faster than thick ones.

Sulphidation — 310’s biggest weakness and the most badly written section​‌​​‌​

Here the sources openly contradict each other, and that is a real hazard. One North American alloy house writes that 310 has “good resistance to sulphidation and other forms of hot corrosion”, while German and European datasheets write, for the same alloy, “low resistance to sulphur-bearing oxidizing and reducing gases”. [CONFLICT] Both are true in a sense, because they are not talking about the same thing.

The correct distinction. (1) Oxidizing sulphur (SO₂, SO₃, excess-air combustion gas): the chromia layer survives and 310 behaves acceptably — but the ceiling drops. One producer states it plainly: continuous 1050 °C with sulphur ≤2 g/m³; above that, a maximum of 950 °C. A 100 °C penalty for the mere presence of sulphur.
(2) Reducing sulphur (H₂S, low-oxygen sulphur-bearing gas): here 310 loses, and it loses because of exactly what makes it valuable — NICKEL. With no protective oxide, sulphur reaches the metal and forms a nickel–nickel-sulphide eutectic that MELTS at about 645 °C. A liquid phase appears at the grain boundaries and the part dissolves from the inside. An alloy with 20 % nickel performs WORSE in reducing sulphidizing gas than a low-nickel alloy. Practical rule: do not use 310 in reducing atmospheres containing H₂S. This is the single most dangerous misunderstanding on 310 datasheets.​‌​​‌​

Carburization, nitriding and low-oxygen atmospheres

If carbon penetrates the protective oxide it precipitates as chromium carbide, doing two things at once: it embrittles the material and it strips chromium from the matrix, collapsing oxidation resistance. A carburized 310 part gains weight, grows, distorts and is brittle at room temperature. The limit in carburizing or nitriding atmospheres is 850–950 °C — 150–200 °C below the limit in air. European publications rate 1.4845 as low resistance [CONFLICT: for 1.4841 both “fair above 900 °C” and “poor above 900 °C” wordings circulate]. In a low-oxygen atmosphere the limit falls to ≈1000 °C, because the Cr₂O₃ layer cannot find enough oxygen to renew itself — less oxygen LOWERS oxidation resistance. For genuinely carburizing duty 310 is not the answer: the 330 class (35 % Ni) or Incoloy 800H is required. The shield against carburization is nickel, not chromium.​‌​​‌​

Aqueous corrosion — 310 is not special here

310 contains no molybdenum, and that single fact governs most of its aqueous behaviour. A PREN of ≈27 is calculated, but all of it comes from chromium and it does not represent molybdenum’s specific effect of suppressing acidification inside a pit. Measured pitting potentials: no pitting in 0.02 M and 0.5 M NaCl at 23 °C; at 50 °C, 619 mV and 383 mV respectively (vs SCE). A rise of just 27 °C moves the material from “no pitting” to “a measurable threshold”. Crevice corrosion always starts before pitting — under gaskets, on flange faces and behind support plates, 310 is at risk too.​‌​​‌​

Chloride stress corrosion cracking — 310 is NOT immune​‌​​‌​

A common myth holds that high nickel makes 310 immune to chloride SCC. It does not. Nickel genuinely helps — in austenitic stainless steels SCC resistance is worst in the 8–12 % nickel band and improves as nickel rises — but immunity in practice begins above 40–45 % nickel, and 310’s 20 % is far below that. Test results: cracking in 30–46 h in boiling 42 % MgCl₂ and 120–174 h in 33 % LiCl; a welded coupon cracked at 1006 h in 26 % NaCl; specimens survived 1344 h in 25 % NaCl (pH 1.5) and 196 h in 50 % NaOH. Read it this way: 310 lasts longer than 304 and 316 in the severe tests but it still cracks. The threshold temperature is usually quoted around 60 °C. For genuine immunity you need Incoloy 825, 904L or super duplex. On intergranular corrosion, the low carbon of 310S reduces but does not remove the risk: after long high-temperature exposure it becomes susceptible through chromium carbide precipitation, and 310 (C ≤0.25 %) is almost certainly sensitized after welding.

Honest Comparison — 310S or Something Else​‌​​‌​

Frequently Asked Questions​‌​​‌​

Is there a real difference between 310 and 310S, or is it a sales trick?

There is a real and large difference, but not where you expect it. The room-temperature mechanical minima are identical (Rm ≥515 MPa, Rp0.2 ≥205 MPa, A ≥40 %). The difference is carbon: ≤0.25 % for 310, ≤0.08 % for 310S. That changes three things. Sensitization: when a 0.25 % carbon material is welded, chromium carbide precipitation in the HAZ is nearly unavoidable; at 0.08 % the risk drops sharply. Formability: lower carbon is more ductile and welds more cleanly. Creep — and this works in the opposite direction: carbon raises creep strength, so in the creep range 310S is weaker than 310 and certainly weaker than 310H. That is why ASME requires 310H there and why 310H imposes a carbon minimum of 0.04 %. In practice most plate sold as “310” is already to 310S analysis, because modern melting delivers low carbon cheaply; but do not assume it without the certificate — for a creep-range part, document that carbon is ABOVE 0.04 %.​‌​​‌​

They say “310 is usable to 1150 °C”. Can I run my furnace at 1100 °C?

It depends on what the part does, and the answer is almost always close to no. First, what that number is: 1150 °C is a SCALING limit, and it is disputed — a second publishing tradition gives continuous 1050 °C with a 1100 °C peak for the same alloy, and the EN side gives 1050 °C for 1.4845. The real issue: at 1093 °C the typical tensile strength of 310S is ≈44 MPa, its yield strength ≈27 MPa, and its elongation 121 %. The material flows at that temperature; a basket, grid or hanger sags under its own weight. On the code side it is even clearer: the ASME VIII Div. 1 ceiling is 816 °C and the allowable design stress there is 0.75 ksi (≈5 MPa). The correct answer: a 310 part running at 1100 °C must carry no load, must be supported, must be thick-sectioned, and must be designed with a sag allowance. If it carries pressure, 310 is the wrong material to begin with. And every heating–cooling cycle spalls scale and shortens life: steady 1100 °C and 1100 °C once a day are not the same material.​‌​​‌​

The 310 part we pulled from the furnace is brittle. What happened, and can we save it?

Most likely sigma phase precipitated. If the part spent hundreds to thousands of hours in the 600–950 °C band (published bands range from 600–900 through 650–950 to 649–1010 °C), hard, brittle sigma has formed at the grain boundaries. Its most insidious trait: the part is still ductile at service temperature and runs normally; the brittleness appears only once it cools to room temperature. That is why the damage is usually discovered during maintenance, while handling or trying to straighten the part. Can it be saved? Yes. A solution anneal at 1100–1150 °C plus rapid cooling dissolves the sigma and restores toughness. Watch three things: (1) the temperature must be higher than a normal solution anneal (1040–1065 °C), because sigma is harder to dissolve; (2) cooling must be rapid, or the part passes slowly back through the damaging band and sigma returns; (3) the part may distort, so fixturing may be needed. Re-annealing critical 310 parts at planned shutdowns is an extremely economical maintenance strategy that most plants never consider.​‌​​‌​

There is sulphur in my furnace atmosphere. Can I use 310?

The answer depends on “which sulphur”, and the distinction is vital. With oxidizing sulphur (excess-air combustion, SO₂/SO₃) 310 can work, but it is penalised: one producer states plainly 1050 °C with sulphur ≤2 g/m³, a maximum of 950 °C above that — the mere presence of sulphur costs 100 °C of ceiling. With reducing sulphur (H₂S, low-oxygen sulphur-bearing gas) the answer is no, and the reason is instructive: with no protective chromia, sulphur reaches the metal and forms a low-melting nickel sulphide eutectic that melts at about 645 °C. 310’s 20 % nickel is not an advantage in reducing sulphur but a direct liability — a liquid phase forms at the grain boundaries and the part disintegrates from within. This is why the sentence “310 has good sulphidation resistance” is dangerously incomplete on some datasheets. The right approach: obtain the atmosphere analysis (sulphur species, partial pressures, oxygen potential); if reducing sulphur is present, move to low-nickel, high-chromium heat-resisting steels or to alloys developed specifically for sulphidation. Adding nickel makes this problem worse.​‌​​‌​

Common Datasheet Errors and Traps

1. Two different tables of mechanical minima in circulation. One group publishes Rm ≥515 / Rp0.2 ≥205 MPa / A ≥40 % / ≤217 HBW (consistent with ASTM A240); another publishes Rm ≥550 / Rp0.2 ≥245 MPa / A ≥45 %. The second set does not match the A240 310S row. For critical calculations go to the standard text.
2. Equating “310S” with “1.4845”. Carbon (0.08 vs 0.10 %), sulphur (0.030 vs 0.015 %), the nitrogen limit and the hardness ceiling (217 HBW vs 192 HB) all differ. The two standards are stricter in opposite directions.
3. Confusing the carbon of 310 and 310S. 310: ≤0.25 % · 310S: ≤0.08 % — more than a factor of three. Some producer sheets print ≤0.20 % for 310; the ASTM A240 text is ≤0.25 %. Specify 310S for anything that will be welded.
4. Printing 310H’s silicon for 310S. Si ≤1.50 % for 310 and 310S; ≤0.75 % for 310H. Many stockist sheets wrongly print 0.75 % for 310S — silicon accelerates sigma, so the difference matters.
5. Confusing “continuous” and “intermittent” temperatures. One tradition gives continuous 1150 °C / intermittent 1035 °C (the intermittent figure is LOWER because of scale spalling, and that is correct); the other gives continuous 1050 °C / peak 1100 °C. There is a 100 °C gap and most datasheets do not say which tradition they follow.
6. Treating the scaling limit as a design temperature. “Oxidation resistance to 1093 °C” is about scaling. The ASME VIII Div. 1 ceiling is 816 °C, where the design stress is 0.75 ksi (≈5 MPa). Three numbers, three meanings.
7. Silent conflicts in physical properties. Density 7.8 / 7.89 / 7.9 / 8.03; modulus 195 / 196 / 200 GPa; conductivity 13 / 13.8 / 15 W/m·K; resistivity 85–86 / 94 µΩ·cm. Do not average them.
8. Quoting the sigma band as a single figure. Published bands are 600–900, 650–950 and 649–1010 °C; sensitization is separately 550–800 °C. Take the conservative envelope.
9. Writing “310 resists sulphidation” without qualification. Acceptable in oxidizing sulphur, POOR in reducing sulphur (H₂S), because 20 % nickel produces a nickel-sulphide eutectic that melts at ≈645 °C. European datasheets rate the same alloy as “low resistance”. This is the most dangerous conflict on the list.
10. Using 310S as a code part in the creep range. A heat below 0.04 % carbon does not carry 310H creep stresses. Ask for dual certification.
11. Claiming “310 is easy to weld”. The weld metal is FULLY austenitic with no delta-ferrite safety valve and is prone to hot cracking. Interpass ≤150 °C, low heat input, no preheat. 304 reflexes do not work on 310.
12. Recommending post-weld stress relief. A slow cycle at 550–900 °C is exactly inside the sigma and sensitization band. The only correct treatment is a full solution anneal plus rapid cooling.
13. Using short-term tensile data instead of creep data. Above 650 °C life is governed by creep and rupture data, not the tensile table — and 310S does not share the 310H curve.
14. Leaving expansion out of the calculation. A 3 m 310S beam heated to 1000 °C grows by about 55 mm. Restrained expansion kills 310 parts before corrosion does.​‌​​‌​

​‌​​‌​