Browse Projects > Detailed View

Refinement of Steel Bridge Ultrasonic Testing Techniques


NCHRP Research Report 908 reported issues with ultrasonic testing (UT) per AWS D1.5, including phased-array ultrasonic testing (PAUT) that need to be addressed. The issues relate to material calibration and to the correlation of the current acceptance criteria to those based on fracture mechanics. Accurate calibration is needed for the proper application of NDE methods to account for variations in material characteristics. The simplified calibration procedures adopted by AWS for UT in the 1960’s is insufficient to address all of the manufacturing processes used for modern bridge steels. NCHRP Project 14-35 discovered that current calibration requirements are lacking because they sometimes fail to adequately account for variations in sound velocity and attenuation between the calibration block and test object. Current methods were developed with calibration blocks made of A36 steel from producing practices of the 1960’s. NCHRP Project 14-35 found differences in sound velocity and attenuation between A36, A709, and nickel plated 1018 steels (which are generally used for calibration today). In particular, steels made using the thermo-mechanical controlled processing (TMCP) are known to have a grain structure that affects the rate at which sound flows through the material. While it is apparent that these differences should be accounted for during calibration, it is unknown how much variability is present within one grade of material across all possible suppliers, dimensions, and processing techniques. For example, it is unknown whether the calibration differences can be addressed by specifying that the calibration block match the grade and processing technique of the test object or whether the acoustic properties of the test object need to be measured on a case by case basis.

Regarding acceptance criteria, NCHRP Project 14-35 found that current scanning procedures and acceptance criteria may not detect and reject all critical discontinuities according to fracture mechanics. It was also noted that the acceptance criteria in the current edition of AWS D1.5 Annex K has less sensitive amplitude limits than the acceptance criteria for traditional UT which have been implemented since 1969. While current NDE practices have resulted in high reliability for bridges in service, it is unknown whether this is due to the current scanning procedures and acceptance criteria resulting in rejection of all critical defects or whether it is instead due to the low likelihood of critical defects remaining in a bridge at a critical location under critical demands. While AWS D1.5 does not allow any detectable indications classified as cracks to remain in the completed weld, accurate classification of a flaw as a crack was not demonstrated during NCHRP project 14-35. NCHRP Project 14-35 developed scanning procedures and acceptance criteria in order to detect and reject critical weld indications in any reasonable orientation, location, and skew within the weld and heat affected zone. While the proposed scanning procedures and acceptance criteria were grounded in fracture mechanics including anticipated bridge demands, weld residual stresses, and fracture toughness is it is not known how they will compare to overall detection and rejection rates compared to traditional methods of combined RT and UT. Round robin inspection data should be collected utilizing the AWS D1.5 Annex K and NCHRP Report 908 procedures on realistic shop fabrication flaws in order to evaluate the overall performance of PAUT as a replacement for combined RT and conventional UT. The detection and rejection rates can then be compared to assess the individual criteria for their effectiveness in evaluation of bridge welds.


The objectives of these study are

  1. Develop calibration procedures for UT and PAUT to effectively account for velocity and attenuation differences in bridge steels.

  2. If necessary, develop alternate procedure for anisotropic materials.

  3. Collect round robin inspection data utilizing the AWS D1.5 Annex K and NCHRP Report 908 procedures on realistic shop fabrication flaws in order to evaluate the overall performance of PAUT as a replacement for combined RT and conventional UT.

Refine or develop PAUT acceptance criteria to ensure successful performance of bridge welds, including consideration of the performance testing criteria proposed in NCHRP Report 14-35.


This project will improve nondestructive evaluation of welds, improving the accuracy of test results and increasing confidence in NDE methods and bridge welds. Improved confidence will facilitate the replacement of RT and traditional UT with PAUT. PAUT is a better method for evaluating welds because

  1. Compared to traditional UT, it facilitates evaluation at a broad variety of angles, improving the characterization of discontinuities. Further, with encoding, PAUT creates a permanent electronic file of the test results that can be reviewed at any time; by contrast, with traditional UT only the UT technician’s reports and notes are available.

  2. PAUT produces more detailed information compared to RT, including discontinuity height and length whereas RT shows length only. Additionally, RT is likely to miss thin planar discontinuities, such as incomplete fusion or tight straight cracks, that are parallel or near parallel to the X-ray or gamma ray propagation or indications without significant change in density along the axis of the radiation such as lack of fusion. Although PAUT can also miss discontinuities, RT is more likely to miss the types of discontinuities that are more detrimental to the welded member.

PAUT is less intrusive and time consuming for inspection in a fabrication shop than RT. Thus, the move to PAUT will make fabrication more efficient.

Related Research:

The following are found in the existing body of knowledge:

  1. As described above, NCHRP Project 14-35 (NCHRP Research Report 908) studied the use of PAUT for bridge welds at Purdue University. The study 1) provided recommendations for critical flaw sizes, including recommendations for addressing stress concentrations in thickness transitions; 2) raised concerns about the qualification of PAUT technicians based on round robin testing and provided recommendations for performance testing of PAUT technicians; 3) discovered calibration issues related to sound velocity and attenuation (described above); and 4) provided recommended changes for scanning procedures and acceptance criteria to D1.5 Annex K (described above). The results of this study are published in NCHRP Report 908.

  2. 2.

Florida DOT conducted a comparative study of RT, UT and PAUT and found good agreement, including what they describe as promising results

The FHWA studied four welds with built-in defects using PAUT. Calibration was performed on a calibration block made of A1018 steel. They found that scanning from two index points would probably be necessary to get weld coverage and plan to perform a second phase to evaluate the 6-db drop method of evaluating welds (Development of Phased-Array Ultrasonic Testing Acceptability Criteria (Phase 1), FHWA Publication No. HRT-14-074).


Anticipated research tasks include

  1. Conduct a literature review and light synthesis of current practice, including which fabricators and owners have used PAUT on bridge projects

  2. Measure the velocity and attenuation of a broad sample of the types of steel used in bridge fabrication, including the variety of processing. From this sampling, establish how much variation in velocity and attenuation can be expected in the different types of bridge steels. This will include both traditional bridge steels (ASTM A709 Gr50, Gr50W, HPS grades, both Q&T and TMCP, etc.) as well as non-traditional bridge steels (QST, CR50, and duplex stainless steels, etc.).

  3. Fabricate flawed and unflawed test specimens for evaluation of PAUT procedures and technicians. Conduct round-robin testing using PAUT on bridge steels including both the D1.5 Annex K and NCRHP Report 908 criteria as well as traditional methods including combined RT and conventional UT.

  4. Compare the PAUT results and assess the criteria for their effectiveness in evaluation of bridge welds. This comparison should evaluate both the ability of PAUT to recreate the inspection results provided by traditional methods (combined RT and conventional UT) as well as the ability of PAUT to reject flaws which are deemed critical based on their known size using a fitness-for-service (fracture mechanics) approach such as that developed in NCHRP Report 908.

  5. Develop proposed changes to D1.5 based on the results of the study.

  6. Publish the report and recommendations.


The results of this study will be presented to the AASHTO/AWS Bridge Welding Code committee as well as AASHTO CBS committee T-17 for consideration of adoption into D1.5.

Sponsoring Committee:AKC70, Fabrication and Inspection of Metal Structures
Research Period:24 - 36 months
Research Priority:High
RNS Developer:Curtis Schroeder, PhD, PE Associate III Wiss, Janney, Elstner Associates, Inc. 330 Pfingsten Rd. Northbrook, IL 60062 cschroeder@wje.com Ronnie Medlock VP Technical Services High Steel Structures, LLC 1915 Old Philadelphia Pike Lancaster, PA 17602 rmedlock@high.net
Date Posted:03/31/2020
Date Modified:05/11/2020
Index Terms:Steel bridges, Ultrasonic tests, Calibration, Sound attenuation, Weld tests, Welds, Acceptance tests,
Cosponsoring Committees:AKB20, Steel Bridges
Bridges and other structures

Please click here if you wish to share information or are aware of any research underway that addresses issues in this research needs statement. The information may be helpful to the sponsoring committee in keeping the statement up-to-date.