top of page

Phased Array Ultrasonic Testing of Structural Steel Welds to AWS D1.5 and AWS D1.1

In 2015, the American Welding Society (AWS) D1.5 Bridge Welding Code [1,2,3] adopted the annex, "Advanced Ultrasonic Examination," that provides the recommended practice to perform phased array ultrasonic testing (PAUT) on steel bridge joints in lieu of radiographic testing. The annex is an ultrasonic testing alternative to the standard shear wave ultrasonic testing described in AWS D1.5 BWC that has been used for decades. This article describes the basic approach to perform phased array testing in compliance with AWS D1.5 and provides some example results and comparisons between PAUT and radiographic testing. The document was reissued in 2020.

Description of PAUT Equipment for Structural Steel Weld Inspection

Phased array testing of steel bridge welds is performed using a PAUT pulser/receiver, encoder, Plexiglas wedge, and PAUT transducer. The PAUT instrumentation varies from 16:64 to 32:128 configurations depending on the complexity of the weld inspection scan plan with the former number being the maximum active aperture and the letter number the total number of channels available. AWS D1.5 requires that unfocused PAUT inspection are performed. As a result, the PAUT inspections must be performed in the far field and full weld coverage in the first and second leg be demonstrated.

An example PAUT scan plan for a 0.625” joint is shown below. This isa two group setup using an Olympus 32:128 X3 and 5L64-A12 transducer, with 1 mm pitch, and SA12-N55S wedge. One group of 16 elements is used to provide a root inspection of the double vee weld. A second ground of elements is used to provide a full volume PAUT inspection. The two groups are used for the PAUT inspection at one index offset from the weld center line.

Example PAUT multigroup scan plan for AWS D1.5 Bridge Welding Code.
Figure 1: Example PAUT multigroup scan plan for AWS D1.5 Bridge Welding Code.

The phased array inspections of bridge welded joints for weld and heat affected zone defects are generally encoded in the axial direction of the weld. Supplemental base metal lamination scans can be performed with encoded linear scans or raster scanned manually. Additional manual scans are required for towards edges of the welded plate and for transverse oriented flaws. String and magnetic wheel encoders lend themselves to PAUT AWS D1.5 weld inspections. An example string encoder scan is shown below. The encoder is magnetically anchored to an adjacent steel block and the string is typically mechanically fixed to the PAUT wedge at the fastener used to attach the PAUT probe to the wedge. Prior to starting the PAUT line scan, the encoder string is confirmed to be parallel with the scan direction. Upon confirmation, the scan is initialized on the Olympus X3 and the C-scan is populated at the encoder resolution – typically 1mm or 0.040”. The string encoded generally allows for the scan to start and finish very close to the plate edge. A straight edge is generally required to maintain the transducer front face at the desired index offset while minimizing and transducer orientation.

Figure 1: String encoder phased array scan of steel bridge weld to AWS D1.5.

An example magnetic wheel encoder inspection is shown in the video below. In this configuration, the PAUT transducer and wedge are mechanically fastened to the encoder arm using aluminum fixtures. One or two magnetic wheel encoders are used at similar encoder resolutions. Once aligned with the weld center line, the encoder maintains a straight line scan but may required periodic adjustments. The magnetic wheel encoders attract metal debris and can become difficult to manually push at constant speed. Data gaps may occur and these areas can be rescanned by reversing the direction of the scanner.

Figure 2: Magnetic wheel encoder phased array scan of steel bridge weld to AWS D1.5.

Phased Array Ultrasonic Testing – AWS Bridge Welding Code Data Analysis

Analyzing PAUT data acquired from compression flange welds, tension flange welds, and tension zones in welds requires consideration of joint type (compression versus tension for instance), plate thickness, A-scan indication amplitude from the welded and heat affected zone, and indication length. Example PAUT weld inspection data is shown below from a 2.25” thick compression flange over a 6” long encoded scan. Shown on the upper left is the A-scan from the active S-scan focal law. The top left is the 45 to 70 degree S-scan. At the index offset used, the weld is inspected completely in the first, second and third legs.

Figure 3: Analyzing PAUT steel bridge data to AWS D1.5.

The PAUT C-scan on the bottom shows a clear indication from the weld in the 1 to 1.4” range. The first step is to assess the peak amplitude of the reflector. This is done by ensuring the red gate is positioned properly. This can be assessed in the PAUT A-scan or S-scan but is easier in the S-can because the gate position can be viewed across all angles. In this example, the detection gate starts slightly before the plate bottom and extends beyond the third leg. The maximum amplitude of the indication shown in the C-scan is assessed by moving the blue horizontal and vertical data cursors while observing the amplitude measured in % full screen height (%FSH). The maximum amplitude is approximately 40% FSH which is between the Standard Sensitivity Level (SSL) and Disregard Level (DRL) which results in a Class C categorization.

The next step is to assess the length of the weld indication using the PAUT C-scan data using the blue data cursors, red reference cursor and green measurement cursor. AWS D1.1 and 1.5 require that standard ultrasonic and phased array ultrasonic testing data be sized in the length direction using the 6 dB sizing technique. Since the peak amplitude was 40 %FSH, the blue horizontal data cursor is used to identify the 6 dB point to the left and right of the peak amplitude location in the PAUT C-scan. The 6 dB location to the left and right are marked with the red and green data cursors, respectively. The resulting flaw length is close to 0.20”. AWS D1.5 allows a compression stress joint Class C indication to be up to 2” in length. As a result, this indication is acceptable.


Encoded phased array ultrasonic testing of compression flange welds, tension flange welds, and tension zones in flange welds is an acceptable alternative to radiography in AWS structural and bridge welding codes. The inspector must understand the accept/reject criteria as it relates to flaw class, type of joint, and indication length.


  1. American Welding Society (AWS) Bridge Welding Code D1.5M/D1.5:2015

  2. American Welding Society (AWS) Structural Welding Code D1.1/D1.1M 24th Edition, 2020

  3. Haldipur, P. ; Boone, Shane D., “Development of phased array ultrasonic testing in lieu of radiography for testing complete joint penetration (CJP) welds”, Proceedings of the SPIE, Volume 9063, id. 90632A 9 pp. (2014).

75 views0 comments


bottom of page