Duplex stainless steel

Ferrite carries the strength and the chloride cracking resistance. Austenite carries the toughness and the ductility. Get them in roughly equal proportions and you keep most of both, which is the whole idea. IMOA puts a typical 2205 at 40–50% ferrite.

That balance is a heat-treatment outcome, not something the chemistry guarantees on its own. The first-generation duplex grades of the 1930s to the 1960s performed well as parent metal and badly as welded fabrications: the heat-affected zone came out with too much ferrite, so it lost toughness and corrosion resistance against the base metal. What fixed it was deliberate nitrogen alloying from the late 1970s, which stabilises austenite and brings the as-welded heat-affected zone back towards the parent properties. Every modern grade is nitrogen-alloyed for that reason.

So a duplex part is only duplex if it was processed properly. Weld it with the wrong heat input, or stress-relieve it in the wrong temperature band, and you have changed the phase balance and thrown away part of what you bought. On a pump this matters most on a fabricated item, a repair weld or a modified baseplate connection, and it is a question for the fabricator rather than the pump supplier.

IMOA groups the modern grades by corrosion resistance, and the spread is wide enough that two duplex steels can behave nothing like each other in the same water.

IMOA prints the common austenitics alongside for comparison: 316L (S31603, EN 1.4404) at a PREN of 24–25, 304L at 18–19. That crosses two families, so take it as the reason a chloride duty moves off 316, not as a ranking that survives into any other medium.

PREN is the Pitting Resistance Equivalent Number, an empirical ranking of a grade's pitting resistance in chlorides from its chromium, molybdenum, tungsten and nitrogen content. It is genuinely useful for ranking grades inside one family and it is routinely asked to do more than that. IMOA's own guidance is blunt about the limits: the relationships are not linear, the way chromium and molybdenum reinforce each other is ignored, and the number assumes ideally processed material, so it says nothing about a part that was badly heat treated. Treat a PREN as a sorting tool, not a rating. The threshold that defines super duplex is set on it, and the formula is worked through on that page.

Most of the duplex in this trade is a shaft, an impeller or an option line on a materials table, rather than a pump you order from a shelf. Some examples from ranges we supply, taken from the manufacturers' own literature.

Take the Grundfos inline multistage range. On the CRN 32, 45 and 64 the booklet gives the shaft as EN 10088 1.4462, the 2205 duplex, where the equivalent CR of the same size uses EN 10088 1.4057. Higher up the range the CR joins it. The booklet footnotes a duplex shaft on the CR 125 to 255, and the CRN table gives EN 10088 1.4462 from 95 up. So whether you get a duplex shaft depends on which variant and which frame you are quoted, which is worth checking rather than assuming. The CRN sets out the rest of the build. Alfa Laval builds the DuraCirc circumferential piston pump with Duplex 1.4460 shafts as standard, which is 329, a first-generation grade that IMOA rates at a PREN of 30–31. It is there for strength and shaft stiffness in a hygienic pump, not because the product is aggressive.

On the wastewater side the pattern is the same. Hidrostal's own brochure for the Compact range lists the materials as cast iron, ductile iron, Hi-Chrome, stainless steel and Duplex, and the same choice runs through the larger submersibles. Vogelsang will make the pump chamber components of a rotary lobe pump in duplex on request, which is a different commercial proposition from a stock item.

The trade-off is the one every non-standard material carries. It is not shelf stock, so the price and the lead time both move, and on a submersible you are paying for a casting rather than a bar-stock part. Worth it where chlorides are eating the pump. Wasted where they are not.

Duplex is not a high-temperature material, and the reason is the ferrite. Held anywhere between about 300°C and 525°C it forms alpha prime, which hardens the ferrite and embrittles it. This is the 475°C embrittlement effect and it is why pressure vessel design codes cap the temperature at which they will allow a maximum design stress at all. IMOA tabulates them: ASME stops at 315°C across the grades listed, and the German TÜV code is more conservative again, at 280°C for unwelded 2205, 250°C once it is welded, and 250°C for 2507 in tube form. Sigma phase is a separate reaction higher up, 700–950°C in 2205, and that one is a fabrication and annealing problem rather than a service one.

The corrosion resistance has limits too, and they are worth stating because "duplex" gets treated as a synonym for chloride-proof. Duplex grades resist chloride stress corrosion cracking far better than 304 and 316, which is the reason many of their chemical-industry uses are austenitic replacements. They are not immune. IMOA names the conditions where they still crack: the boiling 42% magnesium chloride test, drop evaporation where the metal surface runs hot, and pressurised aqueous chloride systems above the temperature you can reach at ambient pressure.

And a duplex part still lives or dies by its passive film like any other stainless. Weld scale, heat tint and iron picked up from a grinding wheel all break it locally, and the part pits where the film is broken rather than where the specification is weak. Passivation is what puts that right, and skipping it on a fabricated duplex item wastes the material.

Duplex earns its money where chlorides are the problem. Seawater and brackish water, brine, chlorinated effluent, and any duty where a 316 wet end has come back pitted or cracked. That is why it turns up in marine and seawater service more than anywhere else, and why the higher-chloride end of that work goes to super duplex rather than 2205.

It is the wrong answer for a reducing acid. Duplex is a stainless steel, so everything it does rests on a chromium passive film, and hydrochloric acid attacks that at practically any concentration. Alloying more chromium into it does not help, because chromium is not the mechanism that survives hydrochloric. That duty belongs to a nickel-molybdenum alloy such as Hastelloy, or to a thermoplastic or lined wet end if the temperature and pressure allow one. The full order of preference is on corrosive duty.

So the question that decides it is which of the two problems you have. If the medium is a chloride-bearing water and the failure is pitting or cracking, duplex is the family to look at, and the chloride concentration and the temperature decide which grade. If the medium is an acid attacking the film itself, no duplex grade fixes it and you are in a different part of the materials ladder. Send the medium, its concentration and the temperature it runs at, and say which of the two has actually failed on you, because that one fact moves the answer further than the duty point does.