​‌​​‌​

Haynes 25 / (L-605) / UNS R30605 / AMS 5537 / AMS 5759

​‌​​‌​

Haynes 25 — also widely known as Alloy L-605 — is a cobalt-based superalloy. Unlike the nickel-based alloys in the nickel alloy group, its principal element is cobalt; its UNS designation is R30605.

Its composition contains approximately 20% chromium, 15% nickel and 10% tungsten. It takes its strength from solid solution strengthening rather than from precipitation hardening. That difference matters: the structure stays homogeneous, no ageing heat treatment is required, and the tendency to crack during forming is low.​‌​​‌​

It offers high yield and tensile strength up to approximately 980 °C, and its creep resistance is satisfactory in service up to 815 °C. Its fatigue resistance is what makes it the choice for rotating systems such as turbines, and the ease with which it can be welded and formed is an advantage on the manufacturing side.

It is used in aerospace for turbine components and combustion chamber parts, in power generation and nuclear technology for high temperature parts, and in the chemical process industry for corrosion resistant components. Grades covered by ASTM F90 are also used in the manufacture of medical implants.​‌​​‌​

​‌​​‌​

What Haynes 25 (L-605) Is — and Why It Is Not a “Nickel Alloy”

Haynes 25 (UNS R30605 / W.Nr. 2.4964 / DIN-EN designation CoCr20W15Ni) is a wrought, single-phase face-centred-cubic (γ) cobalt-chromium-tungsten-nickel solid-solution alloy: nominally 51 Co – 20 Cr – 15 W – 10 Ni. It is sold most often as L-605, and also as Stellite 25, Udimet L-605, Alacrite and Nickelvac L-605. Most distributor catalogues file it under “nickel alloys”; it is a cobalt alloy, and the nickel is there only to keep the FCC lattice stable (9-11 %). That single fact explains both everything the alloy is outstanding at and everything it fails at.​‌​​‌​

The one sentence that separates it from everything else: L-605 cannot be precipitation hardened. It gains strength only from solid solution and cold work; it resists sulphidation and metal galling better than nickel-base alloys can; and it falls behind in oxidation above 980 °C and loses room-temperature ductility after long intermediate-temperature exposure. It is also not an ASME pressure-vessel code material — the wall most often hit at the sales desk, and it has a section of its own below.

Four quite different cobalt alloys are routinely confused with one another by buyers and by datasheets alike. None substitutes for another, and two of them are not structural alloys at all.​‌​​‌​

Honest positioning against nickel-base high-temperature alloys​‌​​‌​

There are three areas where L-605 genuinely wins. First, sulphidation: a cobalt matrix does not form the equivalent of the low-melting Ni-S eutectic that damages nickel-base alloys in sulphur-bearing combustion products, which is a concrete advantage where materials such as alloy 601 and alloy 800H struggle. Second, resistance to metal galling: on dry sliding faces, in bearing housings and on seal contacts, the manufacturer lists this among the alloy’s three principal features. Third, cold formability and spring temper: the alloy takes up to 40 % total cold reduction and reaches beyond 1500 N/mm² in spring temper — and very few spring materials can also serve in the 900 °C band. Where it loses is just as clear: in long-term oxidation it sits behind nickel-base alloys of the alloy X and 230 class; in aqueous corrosion it is not even comparable to C-276 or C-22, because it has no molybdenum and no published isocorrosion chart; and where precipitation-hardened strength is needed, alloy 718 or Waspaloy give far higher yield. Do not buy L-605 unless at least one of its three real strengths is genuinely required.

Standards by Product Form​‌​​‌​

There is a surprise here and the purchasing desk should know it in advance: R30605 is an aerospace (AMS) and a surgical implant (ASTM F) material. It has NO place in the ASTM B-series non-ferrous specifications, in the ASME SB specifications, or in any pipe, fitting or flange specification. The table below separates what is verified from what is not.​‌​​‌​

ASME Code Acceptance — and Why This Section Is a Warning, Not a Table​‌​​‌​

This is the most commercially useful fact on the page and reading it once is enough: in this research no allowable-stress entry in ASME Section II Part D, no SB specification number and no valid code case could be found for R30605.

Understanding why makes the conversation easier. ASME Section II Part D is fed largely by ASTM B-series non-ferrous specifications. R30605 appears in none of them, because the alloy was developed in the 1950s for the gas turbine hot section and its standardisation ran down the aerospace (AMS) route. The implant route (ASTM F90, ISO 5832-5) was added later. The pressure-vessel route was never opened, and that is a design decision rather than an oversight: the thermal stability of L-605 at intermediate temperatures is not ideal for the tens of thousands of hours of steady service that code design assumes.

Product Forms With NO Standard — the Commercially Valuable Section​‌​​‌​

This is the section your sales engineers should memorise. Published product specifications exist for R30605 in only four main forms: plate/sheet/strip/foil, bar/forgings/rings, welding consumables, and implant product. Everything else is sold to a company specification.

Chemical Composition

AMS 5537, AMS 5759 and ASTM F90 carry effectively the SAME chemistry table — a real convenience compared with the nickel-alloy world. In weight %: Co balance · Cr 19.00-21.00 · W 14.00-16.00 · Ni 9.00-11.00 · Mn 1.00-2.00 · C 0.05-0.15 · Si ≤0.40 · Fe ≤3.00 · P ≤0.040 · S ≤0.030.​‌​​‌​

Note that carbon has a MINIMUM. C ≥0.05 % is not a typographical error: grain-boundary carbides are part of this alloy’s creep strength, and taking the carbon out would weaken it. That is the exact opposite of the design logic behind deliberately decarburised nickel alloys such as B-3. One mill publishes that it aims for carbon in the 0.08-0.11 % band (single source; that is an aim chemistry, not a specification). Chromium (19-21 %) supplies the protective Cr₂O₃ scale and therefore the entire oxidation resistance; 20 % works to 980 °C and no further, which is precisely the problem the lanthanum addition in alloy 188 was made to solve. Tungsten (14-16 %) is the main solid-solution strengthener — and simultaneously the source of the thermal-stability problem, because tungsten-rich secondary phases are what precipitate at intermediate temperatures. Strength and instability come from the same element; you cannot have one without the other. Nickel (9-11 %) keeps the FCC lattice stable at and below room temperature.

Mechanical Properties

​‌​​‌​

There are two separate systems here and mixing up the rows is the commonest single error: on the aerospace route AMS 5537 (sheet/plate) and AMS 5759 (bar/forgings) give different minima; on the implant route ASTM F90 gives two separate sets, one annealed and one cold-worked.​‌​​‌​

Physical Properties​‌​​‌​

Warning — the sources diverge badly in this section. The manufacturer’s own brochure values are taken as the baseline below and every outlier is marked. For density, melting range and elastic modulus there are at least three mutually inconsistent numbers each in circulation.

Heat Treatment and Thermal Stability

​‌​​‌​

First the most important sentence: L-605 cannot be precipitation hardened. There is a solution anneal and there is cold work. Everything in between — “ageing”, “two-step heat treatment”, “γ′ precipitation” — does not apply to this alloy. A heat-treatment procedure written out of 718 or Waspaloy habit does nothing but harm here.​‌​​‌​

Welding​‌​​‌​

What actually goes wrong​‌​​‌​

1. COPPER CONTACT — specific to cobalt alloys and almost never published. The manufacturer’s fabrication guide is explicit: in cobalt alloys copper contamination causes liquid metal embrittlement cracking, and surface contact with copper or copper-bearing materials in the weld region must be avoided. In practice that means: do NOT use a copper backing bar, do NOT use a copper chill block, and keep copper-tipped clamps away from the weld zone. A crew coming from nickel-alloy work does not carry this habit, and the result is cracking at the fusion line.
2. Low-melting contaminants. The manufacturer names lead, sulphur and phosphorus. Tapping compound, marker pen, cutting oil and handprints must be completely removed before welding. The sulpho-chlorinated oil that is specifically recommended for machining this alloy is exactly why cleaning before welding is mandatory.
3. Interpass temperature drifting. 93 °C is not a comfort figure. In a heavy multi-pass joint the part heats itself and 93 °C quietly becomes 250 °C. Measure with a contact thermometer before every pass; in cobalt alloys this is enforced more strictly than in nickel alloys. The same logic bans the wide weave bead: it puts in more heat and holds the pool longer inside the secondary-phase band, so stringer beads plus a controlled travel speed are a rule here, not a preference.
4. Weld metal mechanical properties — published, and good. The manufacturer’s own data at room temperature: GTAW transverse Rp0.2 499 MPa / Rm 925 MPa / elongation 36.5 %; SMAW all-weld-metal Rp0.2 611 MPa / Rm 972 MPa / elongation 31.5 %. In other words a properly made weld delivers base-metal strength and keeps its ductility — a genuine strength of this alloy, and one almost no distributor page bothers to print.

Machining​‌​​‌​

A caveat that must be stated openly: no manufacturer-published speed and feed table specific to L-605 could be found in this research. What follows are individual figures published on product pages and family-level guidance sentences. Publish them as “starting parameters for solution-annealed L-605”, not as a verified producer table.

Corrosion and Oxidation — Where It Is Good, and Where It FAILS

Do not skip this section, because the most dangerous misinformation about L-605 lives here. The alloy has 20 % chromium and no molybdenum (≤1 % as a residual). Those two facts determine its entire aqueous corrosion performance.​‌​​‌​

High-temperature oxidation — with the real numbers

The manufacturer’s own burner-rig data (980 °C, 1000 hours) reads as follows: metal loss for alloy 25 was 198 µm (7.8 mils); in the same test alloy 230 lost 71 µm (2.8 mils) and alloy 188 lost 28 µm (1.1 mils). In other words at 980 °C L-605 loses roughly 2.8 times as much metal as 230 and about seven times as much as 188. At 1095 °C for 500 hours alloy 25 exceeded 635 µm (25 mils) of loss — by the manufacturer’s own assessment, inadequate.​‌​​‌​

That is why the manufacturer’s continuous-service ceiling is 980 °C (1800 °F). Higher short-duration temperatures are possible; continuous service is not. Against this, the figures in circulation are: one publisher 1090 °C, another 1095 °C, and a third which on the same page says “continuous 1093 °C, intermittent 871 °C”. That last one contradicts the manufacturer’s own data directly and is odd as engineering too — an intermittent ceiling 200 °C below the continuous one could only be explained by severe scale spallation, which in turn supports the 980 °C figure. The number to publish is 980 °C; give the others as a footnote on capability, not as a limit.

Sulphidation — the alloy’s real advantage​‌​​‌​

Here L-605 is ahead of the nickel-base alloys, and that follows directly from the cobalt matrix. The manufacturer lists sulphidation resistance among the alloy’s principal features. In environments where sulphur-bearing fuels burn — heavy fuel oil, some process gases, coal-derived combustion products — nickel-base alloys degrade quickly through the low-melting nickel-sulphur eutectic. A cobalt matrix is not exposed to that mechanism in the same way. The commercial consequence: if your customer’s problem is not oxidation but sulphidation, L-605 survives where alloy 601 and alloy 800H struggle. But if oxidation is dominant in the same environment, alloy 188 answers both mechanisms at once.

Aqueous corrosion — WHERE IT FAILS​‌​​‌​

The clearest sentence comes from a producer: L-605 is “NOT DESIGNED for resistance to corrosive aqueous media”. That the same page states a few lines above that it is “highly resistant to hydrochloric acid, nitric acid and wet chlorine” shows that the page contradicts itself — and that contradiction has been copied widely across the industry. The correct reading is this: L-605 is a high-temperature and wear alloy; aqueous corrosion resistance is a secondary property, and no published isocorrosion chart, critical pitting temperature (CPT) or critical crevice temperature (CCT) value could be found in this research.

The honest comparison is this. C-276 and C-22 carry 13-16 % molybdenum, and their pitting and crevice resistance comes from that molybdenum; B-3 with 28 % molybdenum is the specialist in reducing acids. L-605 has none of this. What cobalt offers is sulphidation resistance when hot and non-galling behaviour in metal-to-metal contact; in aqueous corrosion it offers no more than a 20 % chromium alloy should be expected to. If a process engineer has specified L-605 as a corrosion alloy, they probably meant MP35N or a Co-Cr-Mo implant grade — and that question is asked before delivery, not after.

Frequently Asked Questions​‌​​‌​

The customer specification says L-605, ASME Section VIII Div. 1, 600 °C design temperature. Can we supply?

Not as a pressure-retaining material — and the reason is scope, not temperature. Separating those two matters, because the entire conversation with the customer depends on which one you have hit.
Metallurgically L-605 is not in difficulty at 600 °C: the manufacturer publishes a typical tensile strength of 823 MPa at 649 °C and a 1000-hour rupture strength of 231 MPa at 649 °C. The material works perfectly well at that temperature.
The problem is that R30605 is not an ASME code material. In this research no allowable-stress entry in Section II Part D, no SB specification number and no valid code case could be found. With no allowable stress there is no design — not at 600 °C, and not at any temperature. The same applies to Section VIII Div. 2, Section I, B31.1 and B31.3. No ASME Section IX P number has been assigned either, so even procedure qualification has to run through the unlisted-material route.
There are three realistic answers. One: if the part is not a pressure boundary — an inner liner, a heat shield, a flame holder, a bearing housing — L-605 can be used and does not conflict with the code; it simply does not enter the pressure calculation. Two: if it is a pressure boundary, move to a code-listed alloy — alloy 625, alloy X, alloy 800H. Three: if the application is aerospace, the route is MMPDS 6.4.1 design allowables and AMS 5537 / AMS 5759 acceptance instead of ASME — and there L-605 is a fully equipped material. What you must not do is quote “maximum operating temperature 980 °C” off a distributor page and let it turn into a code design temperature. That is a capability figure and no code accepts it.​‌​​‌​

The customer wants “L-605 seamless pipe to an ASTM standard”. What do we offer?

The honest answer is that no such standard exists — and saying so up front is what saves you. Published product specifications exist for R30605 in four forms only: plate/sheet/strip/foil (AMS 5537), bar/forging stock/forgings/rings (AMS 5759), welding wire and electrodes (AMS 5796 / 5797), and implant product (ASTM F90, ASTM F1091 for surgical fixation wire, ISO 5832-5 internationally). For pipe, tube, fittings, flanges, bolting and castings there is no product specification covering R30605.
This is where the trap starts. Distributor pages routinely cite ASTM B626 for L-605 tube, ASTM B637 for rings and ASTM B564 / B446 for forged blocks. These are all NICKEL alloy specifications and none of them covers cobalt R30605. B626 is in any case a welded tube specification, not a seamless one. You will also see pages that write “AMS 5759 — bar, forging, ring and tube”; there is no tube in the title of AMS 5759.
So what do you do? Keep selling the product, but write three lines onto the order confirmation: (1) chemistry is certified to AMS 5759; (2) mechanical acceptance is stated explicitly — in most cases the customer will accept the AMS 5759 bar minima (≥862 MPa tensile, ≥310 MPa yield, ≥30 % elongation); (3) dimensional and surface tolerances are to the producer standard or the customer drawing. A supplier who writes those three lines does not have an acceptance argument six months later. This is exactly how the medical stent tube world already works: the tube is made under ASTM F90 chemistry plus the producer’s procedure, not under a tube product specification.​‌​​‌​

Our combustor liner is L-605. After 10,000 hours at 870 °C it cracked during strip-down. Was the material faulty?

Most likely not — most likely the material behaved exactly as expected and the specification was the wrong choice. This is L-605’s best-documented limit and the manufacturer’s own brochure says so.
What happens is this: held for long periods at intermediate temperatures, L-605 precipitates secondary phases and loses room-temperature ductility. The part is sound at service temperature; it is brittle once it cools. The crack therefore often appears not during service but during cooling, or during strip-down and handling — exactly as you describe.
Which phase is responsible is where the sources diverge, and we are not hiding it. The manufacturer’s brochure points to the Co₂W Laves phase. One mill gives the window as 760-925 °C. Against that, a peer-reviewed ageing study — 3 months at 600 °C, 6 and 12 months at 800 °C, 3 and 6 months at 1000 °C — never observed the Co₂W Laves phase at all; it found α-Co₃W (L1₂-ordered) at 800 °C and W₃Co₃C (M₆C) carbide at 800 and 1000 °C. The engineering consequence is the same under either reading, so do not let the phase argument distract you.
A second mechanism may also be in play: 870 °C is below the manufacturer’s 980 °C oxidation ceiling, so oxidation alone does not explain it — but if you burn a sulphur-bearing fuel and the fuel changed, the picture changes with it.
What to do? If the part really will spend tens of thousands of hours in the 600-1000 °C band, L-605 is not the right alloy for that duty — the manufacturer states in its own brochure that newer alloys are superior where thermal stability is critical. If sulphidation is also dominant, look at alloy 188; if the problem is purely oxidation and stability, look at 230 or alloy X. Choose L-605 when short to medium duration service, high short-term strength, galling resistance or cold formability is what you genuinely need.​‌​​‌​

Our quotation is far above the one we gave six months ago. What happened to cobalt?

The problem is not the alloy, it is the cobalt raw material, and it is a supply-and-policy story. L-605 is about 51 % cobalt by weight — so its cost is very nearly the cobalt price. Add 15 % tungsten, which is not a cheap element either. The price of this alloy is built on those two elements and it does not follow the same wave as the nickel alloys.
Here is what happened. Roughly three quarters of world cobalt production comes from the Democratic Republic of the Congo. The DRC imposed a blanket export ban on 21 February 2025, and then moved in October 2025 to a quota system capping quarterly exports per producer. The result shows in the price table: cobalt went from US$24,424 per tonne in January 2025 to US$41,880 in October 2025 — about a 71 % increase. On the standard-grade cobalt hydroxide side a rise of roughly 167 % between January 2025 and June 2026 has been reported. Indonesian capacity has not been able to close the gap, and the DRC quotas are expected to persist through 2026. Practical consequence: do not give long quotation validity, explain the difference between old stock cost and a new heat, and discuss the nickel-base alternative early.​‌​​‌​

Common datasheet errors — check these before you place an order

1. “ERCoCr-A” or “RCoCr-A” quoted as L-605 filler — WRONG. That classification belongs to the Stellite 6 type HARDFACING alloy. AWS A5.14 is the wrong address too; A5.14 is the nickel-base bare filler specification. L-605 filler has no AWS classification; it is ordered as AMS 5796 (bare wire) and AMS 5797 (covered electrodes). At least one consumable datasheet lists the L-605 chemistry in its table while printing A5.13 RCoCrA in its heading.
2. ASTM B-series numbers cited for R30605 — WRONG. Distributor pages show ASTM B626 for seamless tube, ASTM B637 for rings and ASTM B564 / B446 for forgings. All of these are nickel alloy specifications. B626 is a welded tube specification in any case.
3. “AMS 5759 — bar, forging, ring and tube” — WRONG. The title of the specification is bars, forging stock, forgings and rings. There is no tube.
4. “Mo 1.20-1.40 %” appearing in the chemistry — WRONG. There is no molybdenum in the specification chemistry tables; the producer’s nominal table carries only a Mo ≤1 % residual ceiling. One distributor turned this into both a range and a figure above that ceiling.
6. Density circulating as four different values: 9.07 (the manufacturer’s own brochure) · 9.13 (most common) · 9.20 · 9.27 g/cm³. The 2.2 % spread lands on the tonnage price. Melting range likewise circulates as 1330-1410 °C (manufacturer), 1410-1438 °C and 1300-1330 °C — most likely solidus/liquidus confusion.
7. Oxidation ceiling: 980 or 1095 °C? The manufacturer’s continuous service ceiling is 980 °C and it is backed by its own burner-rig data (over 635 µm loss in 500 hours at 1095 °C). The 1090 / 1093 / 1095 °C figures in circulation are capability limits. One page writes “continuous 1093 °C, intermittent 871 °C”, contradicting both the manufacturer and itself.
8. Rupture strength given with NO TIME BASE. “269 MPa at 649 °C” is unusable: is that 100 hours or 1000? In the manufacturer’s table 649 °C gives 328 MPa at 100 h and 231 MPa at 1000 h. Never put a rupture value with no duration into a design.
strength comes from cold work only. An ageing procedure written out of 718 habit does nothing here.
12. An ASME P number or a maximum code temperature quoted — WRONG. For R30605 no Section II Part D entry, no SB number, no P-No. and no valid code case could be found in this research. Do not publish a code temperature.
13. Alloy mix-ups. Haynes 188 (R30188) contains lanthanum and is roughly seven times better in oxidation — it is not L-605. Stellite 6B is a wear alloy, not a structural one. MP35N (R30035 / ASTM F562) is Co-35Ni-20Cr-10Mo with no tungsten. Stellite 25, however, really is L-605 — that is the one legitimate synonym.
14. W.Nr. and BS number conflicts. 2.4964 is the established number; one source also writes 2.4967. For the British standard one source says BS HR 40 and another BS HR 5. Both are single-source; do not publish without verifying against the standard itself.
15. “Highly resistant to HCl” and “not designed for corrosive aqueous media” printed on the SAME PAGE. The second is correct. No published isocorrosion chart, CPT or CCT value could be found for L-605, and the alloy has no molybdenum. Do the aqueous corrosion work with C-276 or C-22.​‌​​‌​

​‌​​‌​