Weld — Definition & NDT Use

From a materials standpoint, Weld affects how an NDT signal propagates, scatters, or returns — which is why method selection, frequency, and reference blocks are tied so tightly to material specification. The magnetising current creates a field that runs continuous through the part; at a discontinuity the lines of flux squeeze around the gap and break the surface as a leakage field, where dry powder or wet-suspension particles cluster and outline the flaw to the inspector's eye. On piping, the inspection is usually a circumferential band of UT thickness readings, a girth-weld RT or PAUT shot, and a follow-up MT/PT on the toes — each method picking up a different failure mode at the same weld. On a pressure vessel, the procedure follows the API 510 inspection plan: thickness monitoring on shell and head, internal visual on a cycle, NDE on nozzle welds, and a fitness-for-service review whenever a reading falls below a calculated minimum. Material specification, heat treatment, and manufacturing route all leave fingerprints in the NDT signal; reference blocks cut from the same heat as the part are used wherever those fingerprints might be confused with a real flaw.

Material data drives method selection long before the inspector arrives on site: a coarse-grained austenitic weld and a clean ferritic plate produce very different ultrasonic responses and demand very different setups. On welded fabrication it is most often paired with VT and one volumetric method (RT or UT) so surface and internal defects are both addressed. On in-service piping the inspection is scheduled against an API 570 corrosion monitoring plan rather than a one-off; trend data is what matters.

Treating a material as if it were homogeneous when its grain or heat-treat condition says otherwise is the cause of false calls and missed flaws far more often than equipment failure.