A single porosity pore, a short slag inclusion, a small UT indication—each one, evaluated on its own, may sit well within AWS D1.1 acceptance limits. But weld discontinuities rarely occur in isolation. Porosity clusters, strings of slag inclusions, and adjacent UT reflectors all involve a second question the individual-indication criteria do not answer: what happens when acceptable indications are close together?

AWS D1.1 addresses this through proximity and accumulation rules embedded in the inspection acceptance criteria of Clause 9. CWIs and NDE technicians who know only the individual-indication limits—maximum pore diameter, maximum indication length, reference level amplitude—are working with half the picture. The proximity rules are where borderline weld quality gets decided.

Why Proximity Rules Exist

Two adjacent discontinuities in a weld are not simply two independent flaws. The material between them is also reduced in cross-section, and the stress concentration effects of one flaw can interact with those of the adjacent one. A cluster of pores may not create a continuous void, but under tensile or fatigue loading the effective stress-raising effect can approach that of a continuous flaw of equivalent projected length.

AWS D1.1's proximity rules capture this by treating adjacent indications as a combined discontinuity when they are close enough that their interaction is structurally significant. The threshold varies by NDE method and discontinuity type.

Porosity: Visual and RT Accumulation Limits

For visual inspection, AWS D1.1 limits the frequency and distribution of porosity that is visible at the weld surface. Isolated surface pores must not exceed the maximum diameter permitted by the acceptance criteria for the weld category and loading type. The visual criteria do not provide an explicit accumulation formula—what they establish is that scattered, randomly distributed surface porosity above a certain density is a reason to investigate subsurface conditions via RT or UT.

For radiographic testing, the code limits porosity in two ways. Individual pores are limited by maximum diameter as a function of weld thickness. Separately, the code limits the accumulated area of porosity within a specified length of the radiograph. If the total porosity area within that inspection window exceeds the limit—even if no individual pore exceeds the individual diameter limit—the region fails. This is the accumulation rule.

The practical implication: when you read an RT film and see scattered small pores, each one individually acceptable, you must still calculate the total porosity area within the code's evaluation length and compare it against the area limit. This calculation is more demanding than a simple pore-by-pore check, and it is the step most commonly skipped on busy production radiographs. See RT acceptance criteria under AWS D1.1 for the specific area limits by weld thickness category. Rule library based on AWS D1.1:2025; verify against your governing edition.

Elongated Discontinuities: The Separation Criterion

Slag inclusions, incomplete fusion indications, and elongated porosity strings are evaluated differently from round pores. The critical dimension is length, not diameter. AWS D1.1 permits elongated indications up to a length limit based on weld thickness.

For adjacent elongated indications, the separation rule applies. Two indications must be separated by a clear gap at least equal to the length of the longer indication; if the separation is less, they are treated as a single continuous indication for purposes of the length limit. This is the interaction criterion for elongated flaws.

An example: two slag inclusions, each 3/8 in long, are separated by a clear gap of 1/4 in. Individually, each is acceptable under typical AWS D1.1 thickness-based limits. But the separation (1/4 in) is less than the longer indication (3/8 in), so by the separation rule they are evaluated as a single 1 in continuous inclusion (3/8 + 1/4 + 3/8 = 1 in). That combined length may exceed the acceptance limit.

This rule has significant practical consequences for welds with systematic lack-of-fusion patterns—root passes with intermittent bonding, or fill passes that consistently leave thin slag stringers at the fusion line. Individual segments look acceptable on the radiograph. The separation evaluation reveals a chronic quality problem.

UT Indications: Proximity and Combination

Ultrasonic testing introduces a different set of proximity considerations because UT indications are reported as amplitude (in decibels relative to reference) rather than as dimensional measurements on a film. Two UT indications that are each just at the reference level present a combined acoustic signature that can be misleading if the technician is evaluating them separately.

AWS D1.1's UT acceptance criteria in Clause 9 establish that when two indications are separated by a distance less than the length of the longer indication, they are evaluated together as a single indication of combined length. The amplitude of the combined indication is assessed against the acceptance criteria for that combined length. This mirrors the elongated flaw rule from RT but adapted to UT scanning geometry.

For phased array UT (PAUT), proximity evaluation can be performed more precisely because PAUT produces a full cross-sectional image rather than a single A-scan trace. Adjacent indications that overlap in depth can be measured with greater certainty. AWS D1.1 permits PAUT under qualified procedures per Annex N (Rule library based on AWS D1.1:2025; verify against your governing edition). See phased array UT under AWS D1.1 for qualification requirements.

The nearest-neighbor question also arises when UT is used on a weld that already has RT data. An RT film may show two separated elongated indications that individually pass the film criteria. UT performed on the same zone may detect them as a combined reflector that fails the UT criteria. When both NDE methods are used, neither overrides the other: a failure by either method is a rejection. Document both methods' results and apply each set of acceptance criteria independently.

Accumulation in Fillet Welds and Surface NDE

For fillet welds examined by magnetic particle testing or liquid penetrant testing, AWS D1.1 visual acceptance criteria apply—surface inspection methods for fillet welds generally use visual criteria as the acceptance baseline. MT and PT reveal surface-breaking and near-surface indications that may not be visible to the naked eye, but the acceptance length limits remain those of the visual criteria applied to the indication as revealed by the MT/PT indication.

Adjacent MT or PT indications on a fillet weld root or toe follow the same logic: if the separation between two linear indications is less than the length of the longer, evaluate them as continuous. This matters especially on high-restraint fillet welds—gusset plate toes, continuity plate fillet welds—where hydrogen cracks can initiate at multiple points and grow toward each other.

See MT acceptance criteria for structural welds and visual acceptance criteria under AWS D1.1 for the underlying limits these proximity rules apply to.

CWI Documentation for Adjacent Indications

When adjacent indications are identified, the NDE report must capture:

  • Location of each indication (weld station, depth if UT)
  • Individual dimensions (diameter or length) measured from the film or scan
  • Measured separation distance between indications
  • Whether the separation rule applies and if so the combined length evaluated
  • The acceptance criterion used and the result

"Two small pores, accepted" is not adequate documentation if the combined porosity area is close to the accumulation limit. Write down the numbers. If you are on the edge—combined indication length within 10% of the rejection limit—that is a CWI judgment call worth documenting with a note: what you measured, how you applied the criterion, and your determination. If the location is in a fracture-critical member or a demand-critical weld zone, escalate to the EOR before final disposition.

The weld discontinuity vs. defect distinction under AWS D1.1 article covers the terminology: a discontinuity that passes acceptance criteria is not a defect. Document both the indication and the accept determination, so there is no ambiguity in the closeout package about whether a recorded indication was accepted or simply missed.

When to Bring the EOR In

The proximity rules give CWIs a deterministic framework for most cases. Where engineering judgment genuinely enters the picture is in three situations:

Indications in fracture-critical locations. AWS D1.1 does not define fracture critical in the same way that AASHTO and AWS D1.5 do for bridges, but some structural projects explicitly designate fracture-critical members. On those, proximity-evaluated indications that pass code acceptance criteria may still warrant EOR review before sign-off.

Repeated proximity failures on the same weld. If a weld consistently produces adjacent slag inclusions that fail the separation criterion, the root cause is a process problem in the WPS or welder technique—not a quality-control documentation problem. Bring the welding supervisor and the WPS owner in alongside the EOR.

Borderline UT indications with competing RT data. When UT shows a combined-length indication that barely rejects and RT shows two separated indications that barely accept, the NDE technicians for each method should review the data together before either report is finalized. The NDE method selection for structural welds article covers how to match method to joint geometry and loading type, which can reduce these ambiguous situations upfront.

Sound WPS practice and consistent process control reduce clustered discontinuities before inspection begins. The CWI's proximity evaluation is the last line—not the first.