Hardness in a duplex-to-carbon-steel dissimilar weld
A fabricator is joining a 12 mm thick 2205 duplex stainless steel nozzle to a 25 mm thick ASTM A516 Grade 70 carbon steel shell. The procedure uses 2209 duplex filler, 50 C preheat, 150 C maximum interpass, and 0.8 kJ/mm heat input. No post-weld heat treatment is planned because the joint contains duplex stainless steel. After welding, phased-array ultrasonic testing is acceptable, but a hardness traverse shows 380 HV10 in the coarse-grained heat-affected zone on the carbon steel side; the procedure limit is 350 HV10. Liquid penetrant testing shows no surface cracks. The vessel will operate between -20 C and 120 C and the client is concerned about hydrogen-assisted cracking. The client proposes a 620 C post-weld heat treatment to soften the hard zone and asks for a technical position before the joint is accepted.
1. Assess the significance of 380 HV10 in the carbon steel heat-affected zone and evaluate whether a 620 C post-weld heat treatment is an acceptable fix given the 2205 duplex nozzle.
Read the worked answer
A 380 HV10 reading exceeds the 350 HV10 procedure limit and indicates a hard, untempered coarse-grained heat-affected zone. That zone has elevated susceptibility to hydrogen-assisted cracking, especially if moisture, grease or hydrogen-containing service is present. The absence of surface indications does not clear the joint, because cracking can initiate below the surface.
A 620 C post-weld heat treatment is not a safe fix here. The joint contains 2205 duplex stainless steel, and holding it around 600–900 C risks sigma-phase formation, embrittlement, and loss of toughness and corrosion resistance. The carbon steel may soften, but the duplex side can be degraded, and differential thermal expansion adds residual stress. The technical position should be that PWHT is not acceptable for this joint as designed. The hard zone should be controlled by the welding procedure instead: butter the carbon steel face with a nickel-based or 309L electrode, then complete the duplex-to-butter weld with 2209; increase preheat and heat input; and use temper-bead or controlled deposition to temper the coarse-grained zone. Hardness should be rechecked on a procedure qualification coupon.
2. Propose a revised welding procedure and inspection plan that would meet the hardness limit without post-weld heat treatment.
Read the worked answer
Revise the joint preparation so the carbon steel is buttered first with a compatible austenitic or nickel-based filler, such as ENiCrFe-3 or 309L, depending on the design code and dilution analysis. Buttering isolates the hard carbon steel HAZ from the duplex weld and allows the final weld to be made with 2209 without excessive dilution. Raise preheat to about 100–150 C for the carbon steel side, increase heat input to roughly 1.0–1.5 kJ/mm within the duplex limit, and control interpass temperature to avoid both hard zones and sigma formation. Use stringer beads and temper-bead technique where access allows.
Inspection should include a procedure qualification with hardness traverses across the carbon steel HAZ, fusion line and duplex side, plus impact testing at -20 C and corrosion testing of the duplex side if specified. Production inspection should combine phased-array ultrasonic testing with hardness checks on a coupon or on a representative production weld, and ferrite content checks on the duplex side. If hardness still exceeds 350 HV10, the butter layer or heat-input window is not adequate and the procedure should be requalified rather than accepted with PWHT.
3. The client insists on 620 C PWHT for code compliance. How would you respond, and what alternative would you offer?
Read the worked answer
I would respond in writing that 620 C PWHT is technically unsuitable for the 2205 duplex component because it can form sigma phase, reduce toughness and corrosion resistance, and may not reliably soften the carbon steel HAZ anyway. Code compliance should not be treated as a single heat-treatment checkbox; the applicable code may allow an alternative joint design or material qualification route. I would provide the hardness data, the duplex sigma-phase risk, and a procedure qualification plan showing how the hard zone will be controlled without PWHT.
The alternative is to change the joint detail: butter the carbon steel and use a nickel-based or austenitic transition, or use a full nickel-based filler for the dissimilar joint, then qualify the procedure with hardness, impact and corrosion testing. If the client still requires PWHT, the joint should be redesigned so the duplex nozzle is not part of the PWHT envelope, for example by using a flanged connection or a length of pipe that can be isolated. Accepting PWHT without qualification would be an engineering compromise that transfers risk to the operating plant.