Ultrasonic Testing of Welds: When the Code Requires UT and What You Get

This page is for the people who have to buy, schedule or specify that testing: owners and their representatives, general contractors, fabricators and erectors, and engineers of record. It covers when UT is required, which welds are tested and how many, who may perform it, and what the report must show. The full range of methods is on the NDT inspection page.

On this page

• What is ultrasonic testing of welds used for?
• When do the CBC and AISC 360 require UT of welds?
• Which welds get ultrasonic testing, and how many?
• How can the UT rate go down or up during a job?
• What decides whether a weld passes UT?
• Who is qualified to perform ultrasonic weld inspection?
• What should a UT report show?
• How do you schedule UT on a steel project?
• What are the limits of ultrasonic testing?
• How SierraPoint Industries can help
• Frequently asked questions

What is ultrasonic testing of welds used for?

Ultrasonic testing of welds is used to judge the inside of a weld. A trained technician couples a probe to the steel and reads how sound reflects from discontinuities such as cracks, lack of fusion and incomplete penetration. It is the method the building code names for CJP groove welds that carry tension across the joint, because those are the welds whose hidden flaws matter most.

UT does not replace visual inspection. AISC 360 states that "observation of welding operations and visual inspection of in-process and completed welds shall be the primary method" of confirming that welding conforms (Section N5.4). UT is the supplement that looks below the surface. Magnetic particle testing (MT) and penetrant testing (PT) look at or just below the surface, and radiographic testing (RT) records a radiographic image of the weld volume.

On a building, ultrasonic weld inspection is part of structural steel special inspection, not a separate program. UT results feed the structural steel special inspection record, so the UT reports and the steel construction inspection reports go to the same people and follow the same Statement of Special Inspections.

This page deals with what UT is required for and what you should receive. How a technician calibrates the instrument and scans a weld is a separate topic, and the project's written UT procedure governs it.

When do the CBC and AISC 360 require UT of welds?

The requirement starts in the building code and is completed by the AISC standards it adopts. The 2025 California Building Code (CBC) took effect on 1 January 2026 and follows the 2024 International Building Code (IBC) for steel. Section 1705.2.1 reads:

"Special inspections and nondestructive testing of structural steel elements in buildings, structures and portions thereof shall be in accordance with the quality assurance inspection requirements of AISC 360." (2024 IBC and 2025 CBC Section 1705.2.1)

The 2025 CBC adopts AISC 360-22 and AISC 341-22 in Chapter 35. AISC 360 Chapter N then sets the UT rates by risk category in Section N5.5b:

"For structures in risk category III or IV, UT shall be performed by QA on all complete-joint-penetration (CJP) groove welds subject to transversely applied tension loading in butt, T- and corner joints, in material 5/16 in. (8 mm) thick or greater. For structures in risk category II, UT shall be performed by QA on 10% of CJP groove welds in butt, T- and corner joints subject to transversely applied tension loading, in materials 5/16 in. (8 mm) thick or greater." (AISC 360, Section N5.5b)

A user note to that section says NDT of CJP groove welds is not required in Risk Category I, or in material thinner than 5/16 in. in any risk category. AISC's comparison of the 2022 and 2016 editions lists no change to Section N5.5.

For the seismic force-resisting system, CBC Section 1705.13.1.1 sends steel in Seismic Design Categories B through F to the quality assurance requirements of AISC 341. AISC 341-22 Section J7.2a requires UT of 100% of CJP groove welds in material 5/16 in. or thicker, and MT of 25% of beam-to-column CJP groove welds, both to AWS D1.8. In ordinary moment frames in Risk Category I or II, the UT and MT apply only to demand critical welds.

California adds limits for healthcare work. Under the [OSHPD 1R, 2 & 5] banner, CBC Section 1705.2.1 does not permit AISC 360 Section N5 Item 3 (coordinated inspection) or Section N6 (approved fabricators and erectors), and Section 1705.13.1.1 says its exceptions are "Not permitted by OSHPD." Buildings regulated under Chapter 17A follow Section 1705A.2.1 instead, so confirm which chapter governs before the UT scope is written.

Which welds get ultrasonic testing, and how many?

The code-required UT applies to a defined set of welds, not to every weld on the job. Under AISC 360 Section N5.5b, a weld is in the required group only when all four of these are true:

1. It is a CJP groove weld.
2. It is in a butt, T- or corner joint.
3. It is subject to transversely applied tension loading.
4. The material is 5/16 in. (8 mm) thick or greater.

Fillet welds and partial-joint-penetration (PJP) groove welds are outside the Section N5.5b group and are inspected visually unless the contract documents ask for more. The project specifications can always add UT beyond the code minimum, for example on heavy transfer girders or column splices.

The base metal near a weld can matter too. Laminations, which are flat separations inside rolled plate, run parallel to the plate surface and can affect T- and corner joints loaded through the plate thickness. ASTM A578/A578M is a straight-beam ultrasonic specification for rolled steel plate 3/8 in. and thicker that detects internal discontinuities parallel to the rolled surface. Whether plate gets this check is the engineer's call in the specifications; Chapter N does not require it by itself.

Worked example: the UT count for a Risk Category II building

A Risk Category II office building's weld schedule lists 92 CJP groove welds in butt, T- and corner joints subject to transverse tension. Twelve of them join 1/4 in. plate. Those 12 are below the 5/16 in. threshold, so the qualifying group is 92 − 12 = 80 welds.

Under Section N5.5b, QA tests 10% of the group: 80 × 0.10 = 8 welds. If the same building were Risk Category III or IV, QA would test all 80. The 12 thin-plate welds still receive visual inspection.

Worked example: a special moment frame under AISC 341

The same building has a steel special moment frame with 12 welded beam-to-column connections, each with a top and bottom flange CJP weld. That is 12 × 2 = 24 beam-to-column CJP groove welds, all in material over 5/16 in.

Where AISC 341 applies, Section J7.2a requires UT of all 24 and MT of 25% of them: 24 × 0.25 = 6 welds. These rates apply in place of the 10% Risk Category II sample for those welds.

How can the UT rate go down or up during a job?

AISC 360 tracks each welder's record and lets the UT rate change. The rules sit in Sections N5.5d to N5.5f and apply per individual welder, not to the whole crew.

• Counting long welds. For the rejection rate, a continuous weld over 3 ft long counts as one weld for each 12 in. or fraction of its length where the effective throat is 1 in. or less, and for each 6 in. or fraction where the throat is greater than 1 in. (Section N5.5d).
• Reduction. Where the rate is 100%, it may drop to 25% for a welder whose rejection rate is 5% or less over at least 40 completed welds, with the approval of the engineer of record and the authority having jurisdiction. Projects with 40 or fewer welds get no reduction (Section N5.5e).
• Increase. Where the rate is 10%, it rises to 100% for a welder whose rejection rate exceeds 5% over at least 20 completed welds. It returns to 10% once that welder's rejection rate is 5% or less over at least 40 completed welds (Section N5.5f).

Worked example: qualifying a welder for the 25% rate

Take a Risk Category III building with about 300 qualifying CJP welds, so the starting rate is 100% and every weld is tested. Welder C has completed 38 short CJP welds and one continuous CJP weld 60 in. long with a 3/4 in. effective throat.

The long weld is over 3 ft and its throat is 1 in. or less, so it counts in 12 in. units: 60 ÷ 12 = 5 welds. Welder C's count is 38 + 5 = 43 welds. UT rejected two of them, one short weld and one 12 in. segment of the long weld, so the rejection rate is 2 ÷ 43 = 4.65%.

That is 5% or less over at least 40 welds, so welder C's rate may drop to 25% if the engineer of record and the building official approve. Of welder C's next 40 qualifying welds, 40 × 0.25 = 10 would then be tested. With three rejections instead of two, the rate would be 3 ÷ 43 = 6.98%, which exceeds 5%, and welder C would stay at 100%.

What decides whether a weld passes UT?

AISC 360 Section N5.5a says the UT, MT, PT and RT it requires are performed by QA in accordance with AWS D1.1/D1.1M. The acceptance criteria therefore come from D1.1, or from AWS D1.8 for the seismic welds that AISC 341 covers. Chapter N itself sets how many welds are tested, not how big a flaw may be.

The edition matters. The 2025 CBC Chapter 35 adopts AWS D1.1/D1.1M-20, while AWS has since published D1.1/D1.1M:2025. By the 2020 edition the inspection requirements were in Clause 8; AISC's edition comparison records the D1.1 clauses it cites changing from 6.12.2 and 6.13.2 to 8.12.2 and 8.13.2. Published summaries of the 2025 edition report further changes to Clause 8, so tables should be cited only from the edition the specifications name.

Secondary sources describe separate D1.1 UT acceptance criteria for statically and cyclically loaded connections, so the contract documents should say which set applies to each weld. Ultrasonic weld inspection rates each indication against the stated criteria; it does not judge whether a weld is strong enough for its load. A rejected weld is reported to the fabricator or erector and, under AISC 360 Section N7, to the engineer of record and the authority having jurisdiction, and the repair and its re-test should be agreed with the engineer of record.

Who is qualified to perform ultrasonic weld inspection?

AISC 360 Section N4 separates welding inspectors from NDT personnel. NDT personnel, other than for visual inspection, are qualified under their employer's written practice, which must meet ASNT SNT-TC-1A or ANSI/ASNT CP-189. According to ASNT, SNT-TC-1A (2024) is employer guidance for in-house qualification and certification programs, while CP-189 (2024) "establishes minimum requirements for employer-based NDT personnel certification."

Both are employer-based schemes. The employer, not ASNT, certifies the UT technician to Level I, II or III under its written practice. Ask the agency for its written practice and each technician's certification record for UT before testing starts.

AWS D1.1 also addresses UT personnel, and published summaries of the 2025 edition report changes to its NDT personnel rules, so check the edition the specifications name. Separately, CBC and IBC Section 1704.2.1 require the agency to document its inspectors' competence to the building official before construction starts.

What should a UT report show?

AISC 360 Section N5.5g requires the NDT to be documented. Each NDT report identifies the tested weld by piece mark and location. When a weld is rejected on the basis of NDT, the report shows the location of the defect and the basis for rejection.

Under CBC and IBC Section 1704.2.4, special inspection reports go to the building official and the registered design professional in responsible charge. Discrepancies go to the contractor first for correction, and to the building official and the design professional if they are not corrected. A final report is submitted at a point agreed before work starts.

The welder UT rate log is not a named code form, but Sections N5.5d to N5.5f cannot be applied without one. Agree its format at the start of the job.

How do you schedule UT on a steel project?

The owner, or the owner's authorized agent other than the contractor, engages the agency under CBC and IBC Section 1704.2. The contractor usually controls the sequence, so both need the same information early. A UT weld test booked late can hold up decking, fireproofing or enclosure.

Send these when requesting a proposal or booking ultrasonic testing services:

1. The structural drawings and specifications, with the Statement of Special Inspections.
2. The risk category and seismic design category, which set the UT rate and whether AISC 341 applies.
3. The CJP weld list with joint type, loading and material thickness, so the qualifying group can be counted.
4. Where each weld is made: fabricator's shop, field, or both, and the shop's location.
5. The fabrication and erection schedule, including when welds will be covered.
6. The acceptance criteria and AWS D1.1 or D1.8 edition the specifications name.

Owners, contractors and engineers should also settle a few points before the first test:

• Who tests in the shop. AISC 360 assigns NDT to QA, except where Section N6 lets an approved fabricator or erector perform it. CBC Section 1705.2.1 does not permit Section N6 under the [OSHPD 1R, 2 & 5] banner, and Section 1705A.2.1 does not permit it in Chapter 17A.
• Visual first. Book UT after the weld has passed visual inspection, so time is not spent scanning welds that will be repaired anyway.
• Access. The technician needs to reach both sides of the joint where the procedure calls for it, before fireproofing, decking or finishes cover it. The agency's written procedure states the surface condition it needs.
• Repairs. Agree how rejected welds are reported, who approves the repair and how quickly the re-test is booked.

What are the limits of ultrasonic testing?

UT is a strong method for internal flaws in CJP groove welds, but it has limits that owners and engineers should know. ASTM E164, the practice for contact ultrasonic testing of weldments, is intended for thicknesses from 0.250 in. to 8 in. and does not include practices for fillet welds or spot welds. The same practice says that when UT is the basis for acceptance, the parties should agree the reference standards and limits in advance.

The material can affect the result. A December 2024 FHWA memorandum on bridge welds reports research finding that differences in shear wave velocity between the calibration block and the steel, and between the rolling directions of some plates, can mislocate flaws or reduce their indications. It notes that shear wave velocity is not reported on mill test reports, and that the effect was most prominent in TMCP steels. That memo concerns bridges under AWS D1.5, but the physics applies to any shear-wave UT.

UT results also depend on the technician, the written procedure and the access to the joint. In Risk Category II the code requires a 10% sample, so a clean sample does not certify the welds that were not tested. Final responsibility stays where the code puts it: the fabricator and erector for their work, and the engineer of record for design and for decisions on nonconforming welds.

Phased array ultrasonic testing (PAUT) is a UT technique that uses a multi-element probe to steer and focus the beam and can record encoded scans. AWS added an annex addressing PAUT to D1.1 in the 2020 edition. Where the specifications call for PAUT, confirm the annex, procedure and personnel requirements of the edition in force with the agency before work starts.

How SierraPoint Industries can help

SierraPoint Industries is accredited by the International Accreditation Service (IAS) as Special Inspection Agency SIA-442 (Type A) under IAS AC291 and ISO/IEC 17020:2012, with a scope written to 2025 CBC Chapter 17A that includes NDT by MT, UT and PT (1705A.2.7) and structural steel construction, welding and high-strength bolting (1705A.2). The details are on the accreditations page. To scope ultrasonic testing of welds for a project, send the drawings, Statement of Special Inspections and CJP weld list to schedule an inspection, or read more about NDT inspection and special inspection services.

Sources

• International Code Council, 2025 California Building Code, Title 24 Part 2, Chapter 17, Sections 1704.2, 1704.2.1, 1704.2.4, 1705.2.1 and 1705.13.1 (effective 1 January 2026)
• International Code Council, 2025 California Building Code, Chapter 17A, Section 1705A.2.1
• International Code Council, 2025 California Building Code, Chapter 35 Referenced Standards (AISC 360-22, AISC 341-22, AWS D1.1/D1.1M-20, AWS D1.8/D1.8M-2021)
• AISC, ANSI/AISC 360-22 Specification for Structural Steel Buildings, Chapter N
• AISC, Comparison of ANSI/AISC 360-22 to ANSI/AISC 360-16
• AISC, ANSI/AISC 341-22 Seismic Provisions for Structural Steel Buildings, Chapter J
• American Welding Society, AWS D1.1/D1.1M Structural Welding Code, Steel
• ASNT, Standards: SNT-TC-1A (2024) and ANSI/ASNT CP-189 (2024)
• ASTM E164-24, Standard Practice for Contact Ultrasonic Testing of Weldments
• ASTM A578/A578M, Standard Specification for Straight-Beam Ultrasonic Examination of Rolled Steel Plates for Special Applications
• Federal Highway Administration, Recommendations to Account for Acoustic Dissimilarities in Steel Ultrasonic Weld Inspection, 13 December 2024