SAW seam porosity and arc instability after consumable change
An IWS supervises submerged arc welding of a 12 m longitudinal seam in 25 mm S355J2 plate for a 1.8 m diameter vessel. The qualified WPS uses 4.0 mm wire, DC electrode positive, 700 A, 30 V, 500 mm/min, agglomerated flux and a ceramic backing on the root. Production is two-sided: first side to about 60% penetration, back-gouge, then second side. Over the last 30 m, welders report intermittent arc stuttering, and the last 4 m show clustered porosity and irregular undercut. The wire spool was changed at the start of the shift, and a partly opened flux bag was left on the machine overnight. The flux hopper contains some caked lumps. The shop wants to continue because the vessel is due for hydrotest in six days. No other process variables were changed, the WPS is valid and the base metal certificates are correct.
1. Outline a systematic diagnosis to separate wire-feed, flux and electrical causes. Which checks would you make first, and why?
Read the worked answer
Reasoning approach
The combination of stuttering, clustered porosity and irregular undercut points to more than one possible cause. I would not change the WPS until I know whether the arc itself is unstable or the consumables are contaminated.
First checks, in order
- Wire feed: confirm feed-roll pressure and alignment, liner condition, contact-tip bore and wire cast. A new spool can still feed badly if the wire has crossed layers or the liner is kinked. Watch the feed rolls during a dry run and measure wire speed against the WPS.
- Electrical circuit: check work-return clamp position, cable connections, voltage-sensing lead and open-circuit voltage. A poor return path can cause arc stuttering and erratic penetration.
- Flux condition: inspect the hopper, screen and delivery hose. Caked lumps and an open bag overnight suggest moisture pickup. Damp flux is a strong candidate for clustered porosity.
- Parameter trace: compare actual current, voltage and travel speed with the WPS. Undercut can be voltage- or travel-related, but it can also follow arc instability.
I would run a short test plate with dry flux and the same wire, then repeat with the suspect flux. That isolates consumables from the equipment without risking the production seam.
2. Assuming the flux is damp, state the corrective actions before production restarts, including parameter and handling controls, and explain how you would verify them on a test plate.
Read the worked answer
Corrective actions
Stop welding on the affected seam. Segregate the suspect flux and the wire lot. Remove all flux from the hopper, hose and screen; do not top up a hopper with fresh flux over old flux. Use new sealed flux or flux baked strictly to the manufacturer's drying schedule, and record the bake time and temperature. Keep flux in a heated holding oven and return unused flux only if the procedure allows it.
Check the wire spool for contamination, rust or crossed winding. Confirm feed-roll pressure, liner and contact tip. Re-establish the work return close to the weld and clean the contact surfaces.
Verification
Use a test plate of the same grade and thickness with run-on and run-off tabs. Weld with the WPS parameters, then inspect visually for undercut and porosity. Radiograph or UT the test weld, and cut a macro if porosity is suspected at the root. Only restart if the test plate is clean and the arc is stable. Record the flux lot, bake record, wire lot and test result.
3. What records and traceability would you require for the affected 4 m and the restart?
Read the worked answer
The affected 4 m must be traceable to the seam, the welder, the WPS, the wire lot and the flux lot. I would require the weld log with current, voltage, travel speed and preheat/interpass where applicable; the flux baking record; the wire certificate; the NDT report; and the disposition of the affected length.
If the porosity exceeds acceptance, mark the repair area, excavate to sound metal, re-weld with a qualified repair procedure and re-inspect. Record the repair location and depth.
For the restart, retain the test-plate results, the first production weld inspection and the parameter record. The records should show that the corrective action was effective, not merely that a new flux bag was opened. This also protects the shop if the customer asks why the seam was accepted after a known consumable problem.