What duplex is, and why the grade name matters
Duplex steels hold roughly equal parts ferrite and austenite — usually 45 to 60 % ferrite. The ferrite brings higher strength and resistance to stress corrosion cracking; the austenite brings toughness and general corrosion resistance. Because the yield strength is well above a standard CrNi austenitic grade, designers can specify thinner sections, and the pitting, crevice and stress corrosion resistance is better than CrNi or CrNiMo austenitics. The main grade is X2CrNiMoN22-5-3, better known as 2205, used as rolled, forged and cast product throughout oil and gas.
PRE, and the line between duplex and super duplex
The pitting resistance equivalent is calculated from the alloy composition as PRE = %Cr + 3.3 × %Mo + 16 × %N, by weight, and it is what formally separates the families. Below 40 is duplex — X2CrNiMoN22-5-3 comes out at 35, and lean duplex X2CrNiN23-4 at 26. At 40 and above it is super duplex: X2CrNiMoN25-7-4 reaches 42. At 48 and above it is hyper duplex, such as X2CrNiMoCoN28-8-5-1. The practical consequence is that a weld deposit with a lower PRE than its parent becomes the first thing to corrode, so matching the family is not optional.
Duplex is not welded like austenitic steel
ESAB states this directly, because the assumption that it can be is the most common and most expensive mistake on duplex work. The parent material arrives with its ferrite-austenite balance set by the mill's solution anneal. Weld metal gets no such treatment: it solidifies fully ferritic and forms austenite only as it cools. That makes the cooling rate the controlling variable. Cool too fast — which is exactly what austenitic practice encourages — and the deposit retains excessively high ferrite, with insufficient corrosion resistance and toughness. Cool too slowly and phase precipitation and embrittlement follow.
The window has two walls
This is what makes duplex demanding: there is a minimum heat input as well as a maximum, and procedures written for austenitic stainless usually violate the lower one. ESAB gives a favourable cooling time of t12/8 = 8 to 10 seconds and recommends stringer beads with no weaving to keep heat input down at the top end. The numbers below are the handbook's, and they differ by family — using the lean duplex figures on 2205 will leave the weld too ferritic, and using the duplex figures on super duplex risks precipitation.
Nitrogen is a consumable here, not just a gas
Nitrogen counts three times as heavily as molybdenum in the PRE calculation, and it is a strong austenite former, so losing it from the weld pool costs both corrosion resistance and the austenite balance the procedure depends on. ESAB recommends pure nitrogen as the backing gas for duplex. This is routinely under-specified on site — argon is what is on the trolley — and the loss does not show up until corrosion testing or, worse, in service.
Preheat: usually not, with two exceptions
Preheating duplex is usually unnecessary, and modern grades with nitrogen content at or above 0.15 %, along with lean and super duplex, generally need none at all provided the heat input is correct. ESAB names two cases where preheating to 100 °C may be advisable, and both are low-heat-input situations: thicknesses above 12 mm welded with very low heat input, and short tack welds. The logic is consistent throughout duplex welding — the risk being managed is cooling too fast, not too slow.
Joint preparation is wider than you would expect
Duplex weld metal is more sluggish than austenitic, so wider included angles of 70 to 80 degrees have been introduced for butt welds to let the pool flow into the joint and avoid lack of fusion. The joint and the adjacent area must be thoroughly cleaned, and only tools and wire brushes dedicated to stainless should be used. One detail that is easy to miss: the arc must not be struck anywhere that will not subsequently be welded over, because a stray strike creates a corrosion initiation point on a material chosen for its corrosion resistance.
Autogenous welding is out
Welding duplex without filler material should generally be avoided, because of the risk of cracking and of insufficient microstructure formation — an autogenous weld has no access to the extra nickel that the filler supplies to drive austenite formation. ESAB stresses ensuring a sufficient supply of filler material, particularly with manual TIG where it is easiest to run short. This is the same reason duplex consumables are over-alloyed in nickel relative to the parent in the first place.
After welding
Post-weld heat treatment is usually not required. In exceptional cases, solution annealing at 1020 to 1050 °C with a five-minute hold followed by water quenching can be carried out — but that is a full re-solution treatment, not a stress relief, and it is rarely practical on fabricated assemblies. What is always required is thorough cleaning: slag and oxides must be completely removed before passivation, and ESAB specifies manual brushing in preference to mechanical brushing to avoid forming grooves that themselves become crevice corrosion sites.
