Flux-cored consumable substitution on impact-tested plate girders
A shop is fabricating 24 m plate girders in S355J2+N, with 25 mm flanges and 12 mm webs. The qualified WPS uses a gas-shielded rutile flux-cored wire classified ISO 17632-A T 46 2 P C 1 H5, 1.2 mm, with CO2 at 15 L/min, 240 A, 26 V and 1.4 kJ/mm. A stock error delivered a self-shielded wire from the same ISO 17632 series, designated for no external shielding gas. The coil looked identical, so the welder ran it with the same CO2 torch and settings. Three days later, 40 web-to-flange fillets show surface porosity, slag lines and convex beads; macro sections show lack of fusion at the web root. Charpy testing of a production test plate gave 22 J at -20 °C against the specified 27 J minimum. The contract requires EN 1090-2 EXC3 and impact testing. The shop proposes to grind the visible porosity and continue production.
1. Identify the nonconformities and classify the essential-variable changes. Can the existing WPS/PQR support continued production?
Read the worked answer
The change is a process change, not a consumable brand change. The original wire is gas-shielded FCAW, while the delivered wire is self-shielded FCAW. That changes the process number, the shielding medium and the consumable classification. In ISO 15614-1 and ASME IX, the process and the shielding gas/medium are essential variables, and the consumable classification is also essential for impact-tested work.
Consequences:
- The WPS and PQR no longer cover the production weld.
- The welder has not been qualified for self-shielded FCAW.
- CO2 at 15 L/min is wrong for a wire that generates its own shielding; excess gas can disturb the arc and the flux coverage.
- The porosity and slag lines fit disturbed shielding and poor slag control; the root lack of fusion fits parameters developed for a different wire.
- The 22 J result is below the 27 J requirement, so toughness is not demonstrated.
Grinding visible defects would not restore procedure qualification, toughness or welder qualification. Production should stop and the affected welds should be quarantined. The shop cannot rely on the original WPS/PQR for this wire.
2. Set out the investigation, disposition and requalification steps before production restarts.
Read the worked answer
Traceability first. Identify every weld made with the delivered coil by coil number, date, welder and WPS. Quarantine those welds and mark them on the fabrication drawings.
Investigation and disposition:
- Perform 100% visual and surface NDT, plus UT or RT on a sampling basis agreed with the client. The macro lack of fusion is volumetric and will not be found by surface methods alone.
- Excavate representative defects to sound metal and repair only under a qualified repair procedure.
- Because the contract requires EXC3 and impact testing, qualify a new WPS with the correct consumable, or qualify the self-shielded wire properly with its own parameters and impact test.
- The PQR must cover the production thickness, position and impact energy at -20 °C.
- Requalify the welders for the process actually used.
- Issue a nonconformance report and obtain client acceptance of the disposition.
Root cause is a stores and identification failure. Introduce barcode issue, colour coding, segregated storage and a WPS check at the workstation before the first arc.
3. What immediate controls would you require in the consumable store and at the workstation?
Read the worked answer
Separate gas-shielded and self-shielded coils physically, label them with the process number and classification, and issue them only against the WPS. Scan the coil at the workstation and block use if it does not match the WPS. Put the WPS and a consumable photograph at each station. Brief the welders on the difference between gas-shielded and self-shielded FCAW, and add a first-off check by the supervisor. Retain the suspect coil and packaging for the investigation. These controls address the actual failure: a coil that looked identical was allowed into a qualified process without a check.