Root response in a thin-wall stainless tube weld
A 25.4 mm outside diameter, 2.77 mm wall 304 stainless steel tube butt weld is examined with a 5 MHz, 70° shear-wave probe with a 6 mm element and a curved shoe. The internal root bead is visible and the cap is ground flush. A root indication is found that reaches DAC +4 dB. When the probe is moved circumferentially, the response is broad and its amplitude changes gradually; when the probe is rotated 180°, it remains at about DAC +1 dB. A second, sharper response appears at the same depth but only over 3 mm and only from one side. The specification requires detection of lack of penetration and root concavity is not a rejectable defect unless it exceeds 1 mm depth. You must decide which response is which and whether the weld is acceptable.
1. How would you distinguish root concavity from lack of penetration using manual UT?
Read the worked answer
Root concavity is a smooth, curved reflector. Its echo is broad, changes gradually as the probe moves, and is less sensitive to probe direction; it often shows a diffuse maximum rather than sharp tips. Lack of penetration is planar and sharp: the response appears abruptly, is stronger from the side that gives a favourable beam angle, and may drop markedly or disappear when the probe is reversed. In this case the broad response that persists at about DAC +1 dB on reversal is consistent with root concavity, while the 3 mm sharp response seen from one side is the suspicious one. Confirm with a second angle, 45°, and from the opposite side of the weld, and inspect the root visually or with a borescope. For 2.77 mm wall, keep the probe index close and use a small element so the root is not confused with the internal bead.
2. What acceptance decision and further actions would you take?
Read the worked answer
Do not accept the weld on the present evidence. The broad response should be assessed as root concavity: measure its depth by profile or by comparing the root profile with a reference, not by echo amplitude. If it is 1 mm or less, it is acceptable under the stated specification. The sharp 3 mm response must be confirmed as lack of penetration before acceptance; a planar root defect is rejectable even if short. Scan from both sides, use 45° and 70°, and if it persists, reject and repair. If access permits, use PAUT or RT to resolve the root geometry. Re-examine the repair with the same technique and record the result. If the sharp response disappears completely at a second angle and is shown to be part of the root bead, document that evidence before accepting.