What is the difference between a weld discontinuity and a weld defect?
A weld discontinuity is any interruption in the uniform structure of a weld or the metal around it, while a weld defect is a discontinuity of a type, size, location or frequency that the applicable code will not accept, and which therefore has to be repaired or the component rejected. Weld discontinuities include porosity, slag inclusions, lack of fusion, incomplete penetration, undercut, overlap, arc strikes and cracking. The two words are not interchangeable: the same pore in two different fabrications can be either one.
Which code decides whether a weld discontinuity is a rejectable defect?
The acceptance criteria of the code governing the fabrication decide whether a weld discontinuity is a rejectable defect, and those acceptance criteria are not universal. AWS D1.1 governs structural steel welding; AWS D1.5 governs highway bridge welding to a noticeably more demanding standard; pressure work is evaluated against the relevant section of the ASME Boiler and Pressure Vessel Code, with ASME Section VIII for unfired pressure vessels and procedure and performance qualification under ASME Section IX.
Within a single welding code the acceptance criteria can differ again by service. AWS D1.1 applies different limits to statically loaded and cyclically loaded connections, because a weld discontinuity that is harmless under a steady load can be a fatigue initiation site under a fluctuating one. Establishing which code, which edition and which service category applied is the necessary first step in deciding whether a weld was rejectable, and it is frequently contested.
Why are planar and volumetric weld discontinuities treated differently?
Planar and volumetric weld discontinuities are treated differently because they are not comparable: volumetric discontinuities reduce load-bearing area without concentrating stress sharply, while planar discontinuities behave like pre-existing cracks in the structure. Porosity and slag inclusions are volumetric discontinuities, rounded voids in the weld. Lack of fusion, incomplete joint penetration and cracks are planar discontinuities — flat, sharp-edged features.
Welding code acceptance criteria reflect the difference between planar and volumetric weld discontinuities. Scattered porosity is tolerated within stated limits; cracks are generally not tolerated at all, in any size or location. Between those poles sit the planar fusion-type discontinuities, where the limits are tighter than for volumetric ones and where the orientation of the flaw relative to the applied stress matters as much as its size.
How does the weld inspection method limit what could have been found?
The weld examination method actually used bounds what could have been found. Liquid penetrant testing under ASTM E1417 reveals only weld discontinuities open to the surface. Magnetic particle testing under ASTM E1444, with technique guidance in ASTM E709, finds surface and near-surface indications in ferromagnetic material and nothing deeper. Neither liquid penetrant testing nor magnetic particle testing can detect a lack of fusion buried in the middle of a weld joint.
Radiography images volumetric weld discontinuities well and can miss a tight planar flaw whose plane is unfavorably oriented to the beam. Ultrasonic examination, with contact-method practice described in ASTM E164, is generally the more capable technique for planar weld discontinuities, but ultrasonic examination depends far more heavily on operator technique, calibration and physical access.
Why does inspection coverage matter as much as the inspection method for welds?
Inspection coverage matters as much as the inspection method because many projects require volumetric examination of a stated proportion of weld joints rather than all of them, selected by category, criticality or sampling. A weld joint that was never examined has no inspection record and no finding, and the absence of an inspection report is not evidence that the joint was sound. Reconstructing what proportion of the welding was inspected, by which method, and how the sample was chosen is often more informative than the individual inspection reports.
Does a weld that passed inspection at fabrication stay acceptable?
Not necessarily: acceptance of a weld at fabrication is a snapshot in time, because the weld was accepted in the condition it was in on the day it was examined. Hydrogen-assisted cold cracking develops after welding, which is why welding codes impose a delay before final examination on susceptible materials: inspect too early and a crack that will exist tomorrow does not exist yet. In-service mechanisms, meanwhile, may later initiate at a weld discontinuity that was entirely acceptable when it was made.
Does a rejectable weld defect mean the weld caused the failure?
Not by itself: a rejectable weld is not the same as a causal one, and that distinction is the one most often lost in a dispute over a weld. Welding code acceptance criteria are workmanship and quality-control thresholds, set conservatively across a whole class of fabrications. Welding code acceptance criteria are not statements that a flaw exceeding them will fail, nor that a flaw within them cannot.
Whether a particular weld discontinuity governed a particular failure is a fracture-mechanics and fitness-for-service question — flaw size and orientation, applied and residual stress, material toughness — answered by different methods from those that establish code compliance. A weld can be non-conforming and not causal, or fully compliant and still the origin.
What records decide whether a weld was rejectable?
Beyond the weld itself, the records that decide whether a weld was rejectable are the contract documents identifying the governing code and edition, the examination procedures and technique sheets, calibration records and reference blocks, inspector qualification and certification, the radiographs or ultrasonic data as recorded rather than as summarized, weld maps identifying which joint is which, and the repair log. Where a weld was repaired, the repair is itself a weld, with its own procedure, its own thermal cycle and its own inspection requirement.
How is the evidence of a weld discontinuity usually lost?
The evidence of a weld discontinuity is usually lost when a suspect joint is ground out and re-welded, which removes the discontinuity, its geometry, its position within the joint and the surrounding microstructure in a single operation. Grinding out and re-welding a suspect joint is the ordinary, sensible response of a fabricator who wants to fix a problem. Radiographic film and ultrasonic data files also have retention periods that expire. Preserving the weld joint as found, and requesting the raw inspection records early, costs very little at the point where it is still possible.
This discussion of weld discontinuities and defects is general technical orientation, not a failure analysis, an engineering opinion, or advice on any specific matter. Determining the cause of a particular incident requires hands-on examination by a credentialed expert.