Radiograph and UT Results on a Pressure-Vessel Butt Weld
A submerged-arc double-V butt weld in 20 mm ASTM A516 Gr. 70 shell plate is 100% radiographed and ultrasonically tested. The project specification states: any crack, incomplete fusion or incomplete penetration is rejectable; slag inclusions are acceptable up to 6 mm individual and 12 mm total in any 300 mm of weld; rounded indications are acceptable up to 3 mm each, but four or more rounded indications inside any 25 mm circle are rejectable; a joint may be repaired twice before it must be cut out. The film shows (A) an 18 mm long, irregular, wavy dark line at the root on the weld centreline with varying width; (B) five rounded indications, 1.0–1.5 mm across, inside a 15 mm circle; (C) one isolated 2.5 mm rounded indication. Ultrasonic testing of the opposite side reports a 15 mm long planar indication at the fusion line of the groove face, 2 mm below the surface. The welder used 3.2 mm wire, 550 A, 30 V, 500 mm/min, and did not clean the groove faces after flame cutting.
1. Classify indications A, B and C and state the disposition of each against the stated acceptance criteria.
Read the worked answer
- A – reject and repair. The indication is 18 mm long, irregular and wavy with a changing width, which is typical of an elongated slag inclusion; incomplete penetration would normally be straighter and more uniform. Either way it fails: as slag it is 18 mm against a 6 mm individual limit, and as incomplete penetration it is rejectable regardless of length.
- B – reject and repair. Five rounded indications inside a 15 mm circle is more than four inside 25 mm, which triggers rejection regardless of the fact that each individual indication (1.0–1.5 mm) is under the 3 mm limit.
- C – accept. A single 2.5 mm rounded indication is below the 3 mm size limit, and one indication cannot form a cluster of four.
Note that A and B are separate repair areas and count separately toward the two-repair limit. If either fails again after the second repair, the joint must be cut out, so both repairs should be done to a verified procedure the first time.
2. The UT indication is judged to be lack of fusion on the groove face. What caused it and what would you change in the procedure?
Read the worked answer
Lack of sidewall fusion in SAW usually has more than one cause, and here two are visible in the record:
- Contaminated groove face. Flame-cut edges carry oxide, scale and sometimes moisture. The arc wets the oxide rather than the steel, so the weld metal sits against the sidewall without fusing it. The faces should be ground bright before fit-up.
- Bead placement. At 550 A and 500 mm/min the bead was probably running near the joint centre. SAW beads must be aimed so each pass washes the sidewall; a slight wire offset or travel angle toward the face would have helped.
Other checks: groove angle and root gap to confirm the wire can actually reach the sidewall, current density and travel speed (higher current and slower travel increase penetration), and flux handling so flux is dry and not recycled with contamination. The fix is a combination of joint preparation, bead placement and parameters, verified by a procedure test, not a parameter change alone.
3. Outline the repair method for indication A and the inspection steps afterward, keeping the two-repair limit in mind.
Read the worked answer
- Locate and excavate: mark the indication on the weld and grind or arc-gouge from the root side (the indication is at the root) to sound metal, extending about 50 mm past each end so the whole length is removed. Do not chase it with a small burr; a narrow slot traps slag.
- Verify: magnetic particle or penetrant test the excavation before welding. No indication means sound metal.
- Reweld: use the qualified WPS, with preheat appropriate to 20 mm A516 Gr. 70 (typically 50–100 °C), low-hydrogen consumables, and heat input in the qualified range.
- Re-inspect: radiograph the repaired area plus UT if the repair is deep, and repeat the original acceptance criteria.
- Record: note that this is repair attempt one of the two allowed. If it fails again, the joint is cut out, so the second attempt should be treated as final and supervised.
- Root cause: clean groove faces at fit-up prevents the inclusion recurring elsewhere on the vessel.