S690QL root HAZ hardness above the 350 HV10 limit
A manufacturer is welding a 30 mm S690QL quenched-and-tempered lifting-beam butt joint with SAW. The qualified WPS specifies 3.2 mm matching wire, agglomerated basic flux, preheat 150°C, interpass 200°C maximum, and heat input 1.8 kJ/mm. The root is deposited with 350 A, 28 V at 35 cm/min; fill passes use 500 A, 30 V at 50 cm/min. After welding, hardness traverses show 380–395 HV10 in the coarse-grained HAZ about 1.5 mm below the root surface. The specification limit is 350 HV10. Tensile tests meet the 690 MPa requirement and UT shows no cracks. The root side is not accessible for a tempering bead. The client asks whether the joint can be accepted, whether a tempering pass on the cap side would solve the problem, or whether the WPS must be re-qualified. Production is already 60% complete.
1. Compare the root and fill arc energies and explain why the maximum hardness occurs at the root HAZ rather than in the fill HAZ. Is 1.8 kJ/mm the likely culprit?
Read the worked answer
Using a SAW efficiency of about 0.9, the root arc energy is 0.9 × 28 × 350 / (35/60) ≈ 1.5 kJ/mm; the fill passes are 0.9 × 30 × 500 / (50/60) ≈ 1.6 kJ/mm. Both are close to, but below, the nominal 1.8 kJ/mm. The root therefore cools faster than the fill for two reasons: it has a smaller weld volume and it is close to the backing or root gap, which acts as a heat sink. The shorter t8/5 in the root coarse-grained HAZ produces a harder transformation product.
A heat input of 1.8 kJ/mm is not excessive for 30 mm S690QL and is not the obvious cause. The hardness pattern points to local fast cooling and possibly inadequate preheat at the root, not to over-welding. Raising heat input further could soften the HAZ but risks losing tensile strength. Check the actual root t8/5 and whether 150°C preheat was maintained through root deposition, especially after tacking and fit-up.
2. Assess the proposal to add a tempering pass on the cap side. What corrective options would you recommend, and what are their risks?
Read the worked answer
A cap-side tempering pass will not temper the root HAZ 30 mm away. The heat source is too distant and the tempering effect is limited to a few millimetres below the cap. Accepting the joint on that basis would leave the hard root HAZ in service.
The practical options are: (1) increase and verify preheat, for example 175–200°C, and maintain it until the root is covered; (2) raise root heat input within the qualified range, but check that tensile strength and toughness still meet S690QL requirements; (3) temper the root from the root side if access can be created, which is not possible here; or (4) re-design the joint to allow access. Increasing preheat too far can soften the HAZ and reduce strength, so the change must be qualified, not applied ad hoc. Hydrogen control must also be confirmed: basic flux baked and handled correctly, clean wire, and no moisture or contamination. The existing 60% production should be hardness-tested at the root before acceptance.
3. Define the re-qualification tests and acceptance criteria you would require before accepting the revised procedure.
Read the worked answer
Re-qualify the WPS by a new PQR under ISO 15614-1 using the revised preheat and root heat input, with thermocouples or t8/5 recording at the root. The PQR should include transverse tensile tests meeting the 690 MPa class requirement, bend tests, impact tests at the specified design temperature, a macrograph, and hardness traverses under ISO 9015-1 at the root, fill and cap. The acceptance criterion is a maximum 350 HV10 at the agreed locations, with no isolated hard spots above the limit.
Also require production hardness testing of the first joints and a check that the revised thermal cycle has not reduced tensile strength below specification. NDT should confirm freedom from cracking. If the revised procedure cannot meet both hardness and strength, the joint geometry or steel selection should be reviewed rather than accepting a hard root HAZ.