Why Bridge Ultrasonics Defeats Technicians Trained on Pressure Equipment

What infrastructure actually examines

Bridge steel is not pressure equipment with different paint. The inspected population is plate girders and box girders, floor beams and stringers, tie girders, truss members and their gusset plates, splice welds, bearing assemblies, pins and hangers, anchor rods and cables. The welds that matter most are complete joint penetration groove welds in members carrying tension, because a bridge is a cyclically loaded structure and tension plus cycles is what drives fatigue cracking.

Ancillary structures belong in the same conversation and are frequently forgotten in training. Sign, signal and high-mast luminaire supports have a well-documented fatigue problem at the pole-to-baseplate socket weld, where cracks initiate at the weld root and propagate around the pole wall. Ultrasonic examination of those socket welds requires its own technique and its own understanding of the expected flaw orientation, and it is nothing like scanning a butt weld in a girder flange.

The consequence structure differs too. A pressure vessel flaw is usually assessed against a leak-before-break intuition and a fitness-for-service framework with time to react. A crack in a nonredundant bridge tension member has no such cushion. That single fact is what shapes the codes, the acceptance tables, the repair restrictions and the personnel expectations that follow, and a technician who does not internalise it will treat bridge acceptance criteria as arbitrary strictness rather than as a rational response to consequence.

AWS D1.5 versus D1.1: which code is actually on the contract

AWS D1.5 is the Bridge Welding Code, issued jointly with AASHTO and written for welded highway bridge construction. AWS D1.1 is the general structural steel welding code and governs buildings and most non-bridge structural work. They overlap heavily in method but not in severity. D1.5 restricts prequalified joints more tightly, imposes stricter welding procedure qualification, controls filler metals, preheat and interpass temperature more closely, and carries its inspection provisions in Clause 6 with the fracture control plan for nonredundant members in Clause 12.

A technician's first professional obligation on any infrastructure job is to establish which document the contract invokes and in which edition, then to read the owner's special provisions on top of it. State transportation departments routinely add requirements: specific procedure approvals, additional personnel qualification, reporting formats, hold points and independent verification. A procedure that satisfies bare D1.5 and ignores the DOT special provision will be rejected at the first audit, and the rework falls on the fabricator's schedule.

Numbering is a practical trap. The ultrasonic provisions of D1.1 have moved between clause numbers across editions — work previously cited under Clause 6 now sits under Clause 8 Part F. Quoting the wrong clause in a procedure or a report is not a substantive error but it reads as carelessness to a reviewer, and on fracture-critical work carelessness in documentation is treated as evidence about the rest of the work.

The decibel rating and the attenuation factor

The bridge codes score an indication with explicit arithmetic. The technician records the indication level in decibels, subtracts the reference level established on the calibration reflector, and subtracts an attenuation factor derived from the measured sound path. The result is a signed decibel rating. A more negative rating means a weaker reflector relative to the reference; a positive or near-zero rating means a strong one. That number, not the raw screen height, is what the acceptance table judges.

The attenuation factor is where the discipline lives. It grows with sound path, so an indication found on the second or third leg of a scan through a thick web is penalised relative to one found close to the entry surface. Technicians arriving from codes where attenuation is absorbed into the DAC curve routinely forget it, and the failure is one-directional: they under-rate deep indications and pass things they should have rejected. It is the single most common substantive error in bridge ultrasonic reports written by cross-over technicians.

The reference itself also differs. Bridge work establishes the reference level on a side-drilled hole in an IIW-type block rather than building a multi-point curve from a block matched to the part thickness. That means the reference is a fixed, portable anchor rather than a part-specific one, and the code compensates for part thickness in the acceptance table instead of in the calibration. Understanding that trade — where the codes choose to put the compensation — is what turns a button-pusher into a technician who can be trusted on a fracture-critical member.

Acceptance classes and why thickness changes the answer

Indications are sorted into classes. Class A covers large discontinuities, which are unacceptable regardless of length. Classes B and C are unacceptable when their length exceeds the limits the code specifies, with Class B held to a much shorter permissible length than Class C. Class D covers minor discontinuities that are acceptable regardless of length. The class an indication falls into is read from a table indexed by its decibel rating, the weld thickness and the refracted angle used to find it.

This is the point where cross-over technicians most often argue with the code. The identical reflector, giving the identical screen response, can be a rejectable Class B in a thin flange and an acceptable Class D in a heavy web, because the tables shift with thickness. That is not inconsistency; it reflects the relationship between flaw size, section size and the stress field. A technician who cannot explain why will eventually apply one remembered threshold across every thickness on the job.

Two further rules catch people out. First, the tables split by loading condition — statically loaded and cyclically loaded work are scored differently, and bridges sit firmly in the cyclic column, where tension members attract the strictest criteria. Second, closely spaced discontinuities must be assessed together rather than individually when their separation falls below the code's limit, so two individually acceptable indications can combine into a rejectable one. Both rules reward a technician who reads the tables rather than memorising a number.

Fracture-critical members and the fracture control plan

A fracture-critical member is a tension member or tension component whose failure would be expected to result in collapse of the structure. The defining property is the absence of a redundant load path. Once a member is designated fracture-critical, the D1.5 fracture control plan governs it, and the plan touches almost every process input: which welding processes and consumables are permitted, how base metal and consumables are traced, how preheat and interpass temperature are controlled and recorded, how tack welds are handled, and how repairs are approved and limited.

Repair restriction is the provision fabricators feel most. The plan limits how repairs may be made and constrains repeated repair at the same location, because each repair cycle re-imposes a thermal history on material whose toughness is the whole point of the designation. That has a direct effect on the inspection technician: an ultrasonic call that triggers an excavation is consequential, and both false rejects and missed indications carry real cost. Calls have to be defensible in writing, with the sound paths, ratings and arithmetic recorded.

Personnel expectations rise accordingly. Base qualification to a written practice under SNT-TC-1A, CP-189 or ISO 9712 is the floor, not the ceiling. Owners commonly require a documented fracture-control endorsement, an owner-administered or third-party examination on the bridge code specifically, and evidence of recent bridge experience. Technicians should assume that a general ASNT Level II certificate in ultrasonics, on its own, will not get them onto fracture-critical work for a state transportation department.

Scanning patterns, access and the two-face problem

Bridge ultrasonics is prescriptive about how the probe moves. Before any angle-beam work, the scanning surface is examined with a straight beam for laminations, because a lamination in the plate will block or deflect the angle-beam sound path and produce a clean, entirely false report of no relevant indications. Skipping the lamination scan is a documentation and a technical failure at once, and it is easy for a reviewer to detect because the record simply is not there.

Angle-beam scanning then follows defined patterns designed to interrogate the weld for longitudinal and transverse discontinuities, with movement combining advance along the weld, transverse oscillation toward and away from the weld centreline, and rotation. Scanning is performed at a gain above the reference level so that marginal indications announce themselves, and the gain is then returned to reference for evaluation. Getting the scanning-versus-evaluation gain relationship wrong in either direction produces either a flood of noise or a quietly incomplete examination.

Access is the practical constraint that shapes everything else. Girders come with stiffeners, cope holes, bearing assemblies and connection plates that block the ideal scanning position, and in-service members are often accessible from one face only. That forces the technician into second- and third-leg sound paths where beam spread has widened, the attenuation factor has grown and resolution has fallen. Knowing when the geometry has defeated the technique — and saying so in the report rather than producing a confident null result — is a mark of competence, not an admission of failure.

In-service infrastructure: fatigue, section loss and the field environment

New fabrication is only half the infrastructure workload. The other half is examining structures that have been carrying traffic for decades. The dominant damage mechanism is fatigue: cracking initiating at weld toes, at cover-plate terminations, at the ends of longitudinal attachments, and at cope holes and re-entrant corners. Distortion-induced cracking in web gaps near diaphragm and cross-frame connection plates is a separate and extremely common family, driven by out-of-plane movement the original detail never anticipated.

Corrosion adds the second theme. Leaking deck joints concentrate chloride-laden water at girder ends and bearing seats, and the resulting section loss is measured ultrasonically where access permits. Pin-and-hanger assemblies, anchor rods at ancillary structures and post-tensioning components each demand their own technique and their own understanding of what a valid signal looks like in a component with complex geometry and threaded or machined features that generate their own reflections.

The field environment is unforgiving in ways a fabrication shop is not. Surfaces are painted, sometimes with thick or unbonded coatings that destroy coupling; temperature swings across a day change velocity and couplant behaviour; access is by snooper truck, rope or lift with limited time on station. And the target is often a tight, closed fatigue crack oriented near-normal to the surface — which an angle beam from the correct side can find well and a straight beam can miss entirely. Surface methods still have their place here, and a technician who understands that ultrasonics is not always the right answer is worth more than one who does not.

What a technician arriving from oil and gas gets wrong

The errors are predictable enough to teach directly. Applying DAC logic and calling anything above the curve. Omitting the attenuation factor, and so under-rating deep indications. Calibrating on the wrong reference — a step wedge or a part-thickness calibration block rather than the code's reference reflector. Treating the acceptance criteria as thickness-independent. Confusing the statically loaded and cyclically loaded tables. Assessing closely spaced indications individually when the code requires them to be combined.

Then there is process discipline. Skipping the base-metal lamination scan. Ignoring the calibration re-verification interval because the previous code allowed a full shift between checks. Failing to re-verify after changing cable, transducer or operator. Setting up in a shop and never re-checking after moving onto the structure and through a twenty-degree temperature change. Each of these individually looks small; together they are the difference between a report a state transportation department will accept and one it will not.

Finally there is register. Pressure-equipment reporting culture leans on fitness-for-service language and remaining-life reasoning. Bridge reporting is a code-compliance document: sound path, probe angle, indication level, reference level, attenuation factor, decibel rating, length, location, class, accept or reject. A technician who writes narrative where the reviewer wants arithmetic will have their reports returned, and repeated returns are how a competent inspector acquires a reputation they do not deserve.

Building an infrastructure-ready ultrasonic technician

Training that works for this sector is built backwards from the report. Candidates read the actual code clauses rather than a summary of them. They drill the decibel-rating arithmetic until it is fluent, including the attenuation factor, on worked examples where the sound path varies. They score the same reflector across several thicknesses and angles so that the acceptance tables stop feeling arbitrary. And they complete the reporting form the owner actually uses, because a correct call recorded badly still fails an audit.

Practical work has to use bridge-representative geometry: groove welds in plate of realistic thickness, stiffener and connection-plate details, restricted access that forces multi-leg scanning, and specimens containing the flaw types the sector genuinely produces. Calibration verification is examined as a discipline of its own, on the clock, with equipment changes introduced deliberately. For candidates heading toward fracture-critical work, the fracture control plan and its personnel provisions are taught explicitly rather than mentioned.

Atlantis delivers ultrasonic Level I, Level II and Level III training and examination preparation to ASNT SNT-TC-1A and ISO 9712, with syllabus content built around the acceptance framework the client's sector actually applies — including the structural and bridge welding codes for infrastructure fabricators and inspection contractors. We also draft and audit employer written practices so that certification survives an owner's review. Delivery is classroom, on-site at your fabrication shop, or blended. Affordable, accessible and fully customisable; consultation or quote on request at info@atlantisndt.com.

Request a consultation