Root Defect on the Last Pair of Plates
You are supervising an eight-trainee shielded metal arc welding practical. The job is a single-V butt weld in 8 mm mild steel plate: 60° included angle, 1.5 mm root face, 2.0 mm root gap, E6013 electrodes of 3.15 mm diameter, recommended range 90–110 A, plates tacked at both ends. At fit-up inspection you find trainee No. 5's plates have closed to a 0.5 mm gap at the finishing end and show 2.5 mm high-low misalignment. He ran the root pass at 135 A with a fast travel speed and a long arc. After chipping and brushing, the reverse side shows a narrow root bead with dark oxide and unfused edges over roughly 60 mm. He has 40 minutes left, one spare pair of plates, and needs a pass to complete the module. The shop acceptance sheet lists incomplete fusion or penetration at the root as a rejectable defect; one repair is permitted with a 10% mark penalty.
1. Decide whether to accept, reject-and-repair, or reject outright, and trace the defect back through the process to justify the decision.
Read the worked answer
Decision: reject the weld and authorise the single permitted repair, with the 10% penalty recorded.
Why it is rejectable. The reverse side shows a narrow root bead with dark oxide and unfused edges. That signature is incomplete root penetration with lack of sidewall fusion - the arc never reached the root face and never washed into both sidewalls. It is a planar defect at the most highly stressed point of the joint, and the shop sheet names it explicitly.
Cause chain - three faults stacked.
- Fit-up. The gap closed to 0.5 mm because both ends were tacked first; shrinkage then pulled the finishing end shut. With a 0.5 mm gap the arc cannot bridge to the root face, so penetration is impossible regardless of current.
- Current and arc length. 135 A on a 3.15 mm E6013 is above the top of the range. High current with a long arc and fast travel gives a wide, shallow, fast-freezing puddle: the arc energy goes into spatter and arc blow rather than into the root, and the sidewalls stay cold. Fast travel removes the puddle before fusion can occur.
- Misalignment. 2.5 mm high-low changes the root face thickness along the joint, so even a correct current would vary in penetration from one end to the other.
Repair route. Gouge or grind out the full length of the defect until sound metal is reached and the groove is re-established, verify by visual and, if available, dye penetrant, re-fit to a 2.0 mm gap checked with a gap gauge, re-tack using a back-step sequence, then weld the root at 90–105 A with a short arc. Note in the record that the repair used the only permitted attempt.
2. He has 40 minutes and one spare pair of plates. What do you actually do in that window, and what do you teach him rather than do for him?
Read the worked answer
Plan for the 40 minutes.
- 0–5 min: Show him the reverse side of his own weld under a magnifier and ask him to name what he sees. Do not name it for him; the recognition is the learning.
- 5–10 min: Ask him to state his current, arc length and travel speed, then compare each against the electrode packet range. Let him find that 135 A is out of range himself.
- 10–15 min: Have him measure the actual gap at both ends with a gap gauge and explain why it closed. Draw the shrinkage direction on the plate with soapstone so the cause is visible.
- 15–35 min: He re-fits and re-tacks the spare pair - tacks at one end first, then check the gap at the far end before the second tack, and use a back-step or bridge tack if it closes. He sets the current himself and shows you the meter before striking.
- 35–40 min: He runs the root pass on the spare pair. You observe, you do not take the stinger.
What you teach, not do. The transferable rules are: fit-up governs penetration; check the gap after every tack; the electrode range on the packet is a limit, not a suggestion; short arc, steady travel. If you re-fit the plates yourself, he learns that the instructor fixes fit-up and he will repeat the same fault next week.
Safety check before he restrikes. Screen in place, gauntlets, chipping goggles before any slag removal, and the electrode stubs go into the bin, not the floor.
3. How do you assess the module outcome so that the mark reflects competence rather than just a repaired weld?
Read the worked answer
Assessment principle. The repaired weld passes the joint geometrically, but the mark must record that the first attempt failed. Assess on three separate judgements, not one.
- Product (majority of marks). The repaired joint is inspected against the acceptance sheet: root penetration visible and continuous on the reverse side, no undercut deeper than the sheet allows, leg length and reinforcement within limits, misalignment within 1.5 mm. If the repair also fails, the module is not passed on that evidence alone.
- Process knowledge (short oral, 5 min). Three questions he must answer without prompting: what caused the lack of penetration, what current range he should have used and why, and how he would tack the next joint to hold a 2.0 mm gap. A trainee who cannot explain the cause has not learned - he has been lucky with a second plate.
- Diagnostic behaviour. Credit him if he measured the gap and set the current himself during the re-fit. This is the behaviour that transfers to the workshop floor.
Applying the penalty. The 10% is deducted from the product mark, not from the whole module, and the reason is written on the record: 'incomplete root penetration, cause fit-up and current'. A mark with a reason is usable evidence for the next instructor; a bare mark is not.
Fairness check. Compare his repaired weld with the other seven trainees' first attempts before finalising. If his repair after gouging and re-fitting is cleaner than a classmate's un-repaired pass, the penalty is still correct - he had a second attempt the classmate did not need.