How to set up ultrasonic testing scan movements, coverage angles, and documentation for structural CJP groove welds under AWS D1.1:2025 — a practical CWI reference.
Ultrasonic testing (UT) has become the dominant NDE method for structural CJP groove welds in fabrication shops and field erection alike. It is faster than radiography, more sensitive to planar discontinuities like lack of fusion, and practical on the thick plate sections common in structural steel. But the quality of a UT examination is entirely dependent on correct scan setup — wrong angles, incomplete coverage, or poor calibration produces a result that looks like a pass while real flaws go undetected.
This article covers how AWS D1.1:2025 structures the UT examination for structural CJP welds: the required scan movements, angle selection, calibration, and the documentation your CWI inspection record must show.
Why Scan Pattern Matters More Than Indication Rating
Most UT disputes on structural jobs are not about whether an indication exceeds the reject threshold — they are about whether the scan actually covered the volume that matters. A porosity cluster at the weld root in a tension chord splice, or lack of fusion at the fusion line in a column splice CJP, is invisible to a UT beam aimed at the wrong angle or moving in the wrong direction.
AWS D1.1 Annex S provides the UT examination procedure for standard structural welds. When the contract references D1.1 without specifying a project-specific UT procedure, Annex S is the default. It defines four basic scan movements — labeled A through D — and the minimum coverage each must achieve.
The Four Standard Scan Movements
Scan A (transverse to weld axis, beam toward weld centerline): The transducer moves parallel to the weld length while the beam is directed perpendicular to the weld axis (across the joint). This is the primary scan for detecting longitudinal discontinuities — lack of fusion along the weld length, aligned porosity chains, and undercut at the weld root. For most plate thicknesses, this scan is performed from both sides of the weld.
Scan B (parallel to weld axis, beam toward weld): The transducer moves perpendicular to the weld while the beam is directed along the weld axis. This is the primary scan for transverse cracks — a less common but critical failure mode in high-restraint joints and hardened HAZ conditions. Scan B is required when the contract calls for full weld volume examination and is particularly important for splices in members subject to transverse tension loading.
Scan C (angled from one side of the weld): The transducer is positioned on one side of the weld and rotated ±15° from perpendicular to capture oblique reflectors that the straight A-scan misses. This is particularly important for detecting lack of sidewall fusion along the weld bevel face, which is oriented at an angle to the beam in a standard V-groove.
Scan D (angled from the opposite side): The mirror image of Scan C, performed from the other side of the weld centerline. Together, Scans C and D ensure coverage of both fusion lines in a single-V or double-V groove weld.
For CJP groove welds in standard structural configurations (single V, double V, single bevel), all four scan movements are typically required. The specific combination depends on the joint geometry and access.
Refracted Angle Selection
The refracted shear wave angle is the single most important setup parameter. AWS D1.1 Annex S specifies angle requirements based on plate thickness:
For plate thinner than approximately 1 inch [25 mm], a 70° refracted angle typically provides the best beam coverage across the narrow weld width. The steep angle means the beam path is nearly vertical, which keeps it within the weld zone and maximizes sensitivity on the fusion line.
For plate between 1 and 3 inches [25 and 75 mm], a combination of 60° and 70° — or 45° and 70° — provides better coverage of the thicker weld cross-section. A single angle leaves a dead zone that increases with plate thickness; two angles with overlapping coverage close that gap.
For plate thicker than 3 inches [75 mm] — common in column flanges, box column corner welds, and heavy plate girder splices — a three-angle scan using 45°, 60°, and 70° ensures full coverage. At these thicknesses the weld width is large, the bevel faces are widely separated, and beam spread alone cannot compensate for poor angle selection.
Rule library based on AWS D1.1:2025; verify against your governing edition (the AHJ or contract may specify 2020 or earlier).
For guidance on UT acceptance criteria once indications are found, see UT acceptance criteria for AWS D1.1 structural welds.
Calibration Block Requirements
Scan setup is referenced to a steel calibration block machined to the same nominal material, heat treatment condition, and surface finish as the production weld base metal. AWS D1.1 Annex S specifies:
- The block contains a 1/16 in [1.6 mm] diameter flat-bottom hole and/or a side-drilled hole at the reference depth
- The transducer frequency must match the production scan (typically 2.25 MHz for structural plate, with 5 MHz available for thin material)
- Calibration is confirmed before each inspection session, whenever the operator changes equipment, and after any scan interruption exceeding 30 minutes
- The reference sensitivity (gain at which the calibration reflector produces an 80% full-screen height response) is recorded on the inspection report
If calibration verification at the end of a session shows the sensitivity has drifted more than 6 dB from the opening calibration, all readings from that session are invalid and the examination must be repeated.
Coverage Documentation on the Inspection Report
The CWI inspection report must be traceable to a specific weld joint, examination date, operator, equipment, and calibration. AWS D1.1 does not prescribe a form, but the minimum content for an enforceable UT record includes:
- Joint identification (weld map reference number, member mark, heat/piece)
- Scanning surface and direction for each scan movement performed
- Transducer angle(s), frequency, and wedge angle
- Calibration block ID and calibration results
- Indication location (distance from reference end of joint and through-thickness position)
- Indication length, amplitude relative to reference level, and classification (A, B, C, or D)
- Accept/reject determination per the D1.1 acceptance criteria table
- Inspector signature, certification level, and date
Without this documentation, the UT examination has no legal standing in the event of a dispute or audit. AISC-certified fabricators are expected to maintain complete NDE records as part of their quality system. For the broader context of how NDE records fit into the weld documentation package, see weld inspection daily report documentation for CWI.
Common Coverage Failures in Production UT
Backing bar shadow. Steel backing bars on CJP groove welds create a beam discontinuity at the root. A transducer scanning from the top surface can miss root lack of fusion because the signal reflects off the backing rather than the root pass. The solution is additional scanning from the backing side (if accessible) or using a higher frequency with smaller element size to improve near-surface resolution.
Weld crown interference. Convex weld reinforcement greater than about 3/16 in [5 mm] prevents the transducer from resting flush on the base metal adjacent to the weld. The dead zone immediately below the transducer then overlaps the weld crown area. AWS D1.1 addresses this with the near-surface resolution requirement — verify your transducer can meet it on the actual joint geometry before inspection.
Incomplete C and D scan from tight access. Field-bolted/field-welded connections often have gusset plates, clip angles, or deck framing that prevents transducer travel along the scan path. Document access restrictions on the inspection report and coordinate with the EOR for disposition when a portion of the weld cannot be scanned.
Scan A only on beam splice CJP. A common shortcut is running only Scan A on beam flange splice CJP welds and skipping Scan B. For members loaded in bending, the tension flange is subject to transverse tensile stress — exactly the condition that makes transverse cracks critical. Scan B is not optional on tension-loaded CJP splice joints.
When to Use Phased Array UT
AWS D1.1 does not have a dedicated phased array UT procedure in Annex S, but it allows alternative UT procedures when the alternative has been qualified and accepted by the engineer. Phased array UT (PAUT) with electronic scanning can cover multiple angles simultaneously, reduce examination time significantly, and produce a permanent B-scan image record. On large structural projects with many CJP groove welds, the time savings from PAUT often justify the equipment cost and the procedure qualification effort.
If PAUT is used, a project-specific procedure qualified in accordance with AWS D1.1 requirements (or SNT-TC-1A Level II supervision) must be in place before examination begins. The PAUT system must produce results traceable to the same acceptance criteria as the conventional Annex S procedure.
For shops managing UT programs across multiple projects with different structural codes, WPS Welding provides a centralized platform for tracking NDE requirements by project, managing inspector qualification records, and maintaining complete weld examination documentation.
Rule library based on AWS D1.1:2025; verify against your governing edition (the AHJ or contract may specify 2020 or earlier).