Suck-Back at the Top of a Small Pipe
Two welders are making 1 in. NPS Schedule 80 A106 Grade B butt joints in a fixed 6G position. The root and hot pass are GTAW; fill and cap are SMAW with 1/8 in. E7018. The root uses a 3/32 in. gap, 1/16 in. land, 1/8 in. ceriated tungsten, 1/8 in. ER70S-2 filler, 70 A, and 15 cfh argon, with no purge and an open root. At the 12 o'clock position, radiography shows scattered porosity and a 3/4 in. length of incomplete root fusion. The welder reports that the puddle 'sucks in' and the arc wanders. The joint sits in a rack 18 in. off the floor with access blocked on one side.
1. What are the most likely causes of the root defects, and what would you change first?
Read the worked answer
On 1 in. Schedule 80 the curvature is high and the heat sink changes rapidly around the circumference, so small parameter errors show up first at 12 o'clock.
Likely causes:
- Suck-back at 12 o'clock: too much heat for the wall thickness, a root gap that is too wide, or insufficient filler added at the top. Gravity pulls the molten root inward when the puddle is too fluid.
- Incomplete root fusion: travel too fast, current too low for the pass, tungsten off-center, or the arc failing to reach the root faces because of the land and gap combination.
- Porosity: loss of shielding from torch angle or drafts, contaminated tungsten or filler, moisture on the joint, or no purge allowing oxidation on the root side.
- Arc wander: a contaminated or off-center tungsten, poor tungsten grind, or excessive arc length.
First changes: reduce current slightly, tighten the gap toward 1/16 in., switch to a 3/32 in. tungsten and filler for better control, shorten the arc, and add filler more steadily at the top. Check gas flow and eliminate drafts.
2. The welder proposes raising the current to 'push through' the top. Evaluate that idea.
Read the worked answer
That is the wrong direction. Raising current on 1 in. Schedule 80 will:
- Make the puddle more fluid and worsen the suck-back at 12 o'clock.
- Increase the risk of burn-through on the thin wall.
- Not correct incomplete root fusion if the real cause is travel speed, torch angle, or joint geometry.
Higher current can help fusion only when the puddle is too cold and sluggish, which is not what the welder described. The symptoms point to excess heat and poor puddle control at the top, so the fix is lower current, better filler timing, and a tighter, more consistent gap.
3. What role do the access restrictions play, and what acceptance path should follow the repair?
Read the worked answer
- The blocked side forces the welder into an awkward torch angle, especially at 3 and 9 o'clock and as he approaches 12 o'clock. A poor torch angle disrupts gas coverage and pushes the arc off the root faces, which matches both the porosity and the incomplete fusion.
- If the procedure and customer permit, rotating the pipe to a more accessible position changes the qualification requirements, so that is a formal decision, not a convenience.
- After repair, the joint should be re-examined by the method specified for the job, typically radiography for this service. The repair itself must follow an approved repair procedure, including excavation to sound metal and re-inspection.
- If the same welder repeats the same defects, the issue is technique and parameters, not luck; requalification or retraining should be considered.