RNS
Browse Projects > Detailed View

Improved Non-Destructive Evaluation of Full Penetration Steel Bridge Welds


I. Problem Title: Improved Non-Destructive Evaluation of Full Penetration Steel Bridge Welds

II. Research Problem Statement

There are two non-destructive evaluation (NDE) methods used for volumetric (through-thickness) evaluation of complete joint penetration (CJP) welds in steel bridges: radiographic (RT) and ultrasonic (UT). Recent advances in ultrasonic testing include the development of phased-array ultrasonic technology (PAUT), which allows for efficient detection and characterization of flaws, with the option of automated data collection and imaging. Criteria for categorizing discontinuities found by RT and conventional UT as acceptable or rejectable are codified in the AASHTO/AWS Bridge Welding Code (D1.5) based on workmanship standards. However, these acceptance criteria do not reflect full use of the capability of PAUT, and furthermore are not based on the effect that these discontinuities might have on the performance of the structure, particularly the resistance to fatigue and fracture. Some discontinuities that are not allowed by AWS workmanship criteria are not at all harmful to the structure. Therefore, research is needed to improve methods and interpretation of these techniques, including:

1) Development of fitness-for-service based acceptance criteria. Present D1.5 acceptance criteria for UT and RT results are not based on fitness for service, but rather on the size, location and distribution of discontinuities that can be reliably detected. As such, welds and base metal are repaired without knowing if the repair is required for structural purposes. Better knowledge is needed with respect to what must be repaired and what can be left in place.

2) Evaluation and codification of phased-array automatic ultrasonic testing (PAUT). Evaluating the potential benefits of using PAUT for moving towards a more accurate and refined fitness for purpose approach. This will include a cost/benefit analysis to assess the added cost potentially associated with the ability of PAUT to detect very small discontinuities that do not threaten the structure integrity. In addition, PAUT techniques will be codified based on best practices in the bridge industry as well as other industries around the US.

3) Statistical sampling. At present, volumetric NDE is conducted on 100% of tension CJP butt splices, but based on broad owner experience, this inspection may be excessive. Once sufficient NDE demonstrates that a routine fabrication process is operating consistently and controls are functioning effectively, statistical sampling of welds may demonstrate sufficient confidence with a fraction of the testing. Appropriate statistical criteria, including sampling rate, need to be established for common NDE situations.

III. Literature Search Summary

1) The FHWA sponsored an international steel bridge fabrication technology scan in 1999, and the scan team noted international practices related to this topic. The findings are presented in Report FHWA-PL-01-018, “Steel Bridge Fabrication Technologies in Europe and Japan”.

2) The FHWA NDE Validation Center conducted parametric studies comparing UT and AUT for steel bridges and may have useful information.

3) NCHRP report # 335 on “Acceptance criteria for Steel Bridge Welds” provides acceptance criteria for welded connections based fatigue crack growth under different levels of effective stress ranges.

III. Research Objective

The objective of this research is to improve NDE of CJP welds on steel bridges by accomplishing the following goals: 1) Develop fitness-for-service based acceptance criteria for internal discontinuities in CJP welds and base metal 2) Conduct parametric study of PAUT, comparing results to RT 4) Develop statistical sampling criteria to reduce unnecessary NDE

IV. Estimate of Problem Funding and Research Period

Recommended Funding and Research Period: Research period: 3 years Funding: $600,000

V. Urgency, Payoff Potential, and Implementation

Fitness-for-service acceptance criteria and statistically based sampling criteria will optimize testing and avoid unnecessary repairs that are costly and potentially more detrimental than the original discontinuities. RT is currently the preferred butt weld inspection method, but UT offers improved safety, better detection of certain flaws, faster production, lower costs, and easier data maintenance. UT test results can be stored electronically, while RT requires storing film. Digitizing RT film is expensive and may incorporate film artifacts or have poor legibility, especially at thickness transitions. Implementation will be accomplished by adoption of new code provisions related to the research findings by the AASHTO/AWS Bridge Welding Code committee.

VI. Person(s) Developing the Problem Statement

Hussam Mahmoud (corresponding) Colorado State University Phone: 970 491 6605 Email: hussam.mahmoud@colostate.edu

Ronnie Medlock VP - Technical Services High Steel Structures, Inc. phone: 717.209.4647 email: rmedlock@high.net Email: rmedlock@high.net

Heather E. Gilmer Quality Assurance Manager Tampa Tank/Florida Structural Steel Phone: 813 241 4261 x205 Email: hgsteelfab@gmail.com

Tom Schlafly AISC Phone: 312 670 5412 Email: Schlafly@aisc.org

VII. Date and Submitted by Originally written in 2007

Re-written in March 2013

VIII. Sponsorship

Alex Bardow, Massachusetts DOT (T-17 chairman)

Steve Duke, Florida DOT


Sponsoring Committee:AKC70, Fabrication and Inspection of Metal Structures
Date Posted:05/05/2013
Date Modified:05/07/2013
Index Terms:Steel bridges, Nondestructive tests, Evaluation and assessment, Ultrasonics, Data collection, Penetration resistance, Butt welds, Phased arrays,
Cosponsoring Committees: 
Subjects    
Highways
Construction
Design
Materials
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.