top of page

Phased Array, TOFD, and TFM/FMC: Comparing Methods for Weld Lack of Fusion Detection

Updated: Sep 13, 2023

Phased array ultrasonic testing (PAUT), time-of-flight diffraction (TOFD) ultrasonic testing and full matrix capture (FMC) / total focusing method (TFM) are the three advanced ultrasonic non-destrictive testing techniques used routinely for full penetration weld inspections.  In this article, the application of PAUT, TOFD, and FMC/TFM to a single vee steel plate butt weld is studied with emphasis on inspection setup fundamentals, data presentation, and performance.  While these three techniques use the the same ultrasonic waves,they are applied with different nuances that are important to understand.  This article provides the reader with an excellent first introduction to PAUT, TOFD, and FMC/TFM  performance comparisons.

Description of Steel Plate Weld

The single vee weld joined two 0.375" thick carbon steel plates.  The welding process included the implementation of lack of fusion (LOF) along the bevel  in the upper half of the plate.  The LOF was approximately 0.30" long and 0.15" high.  A second lack of fusion was inserted in the root along the vertical surface.  The root LOF was 0.50" long and 0.15" high.  This weld flaw broke the bottom surface of the plate. 

Phased Array Ultrasonic Testing for Lack of Fusion

The PAUT weld inspection was setup using an Olympus 5L16 probe with an S-scan angular range of 45 to 70 degrees. THE PAUT scan plan is shown below. The first leg was used mainly for a root pass inspection. The volume of the weld was covered in the second leg which also included the heat affected zone (HAZ).

Figure 1: Phased array (PAUT) scan plan for 0.375” thick steel welded plate.

The PAUT scan length shown is approximately 3.5 inches and encompasses both the root lack of fusion and mid-thickness lack of fusion. Both weld defects were detected clearly with the phased array setup. The root lack of fusion is observable from roughly 0.80” to 1.27”. In the S-scan, it is easiest to discriminate between the root LOF and the actual root in the third leg. At the end of the first leg, it is difficult to spatially differentiate between the root and root LOF. The third leg state provides excellent data redundancy for confirmation. The height of the root LOF was measured at 0.12”. The lack of fusion in the upper half of the plate was detected with the PAUT focal laws in the 50-60 degree range over the 2.70 to 3.10 scan axis length. The height was measured to be 0.070” in the S-can.

Time-of-Flight Diffraction Testing for Lack of Fusion

The time-of-flight diffraction (TOFD) inspection was performed with 10 MHz, 0.25” diameter, transmitter and receiver. The TOFD probe center spacing was calculated to be 1.5 for the 0.375” thick welded plate. The target focal point was set to 2/3 the plate thickness.

Figure 2: Time-of-flight diffraction (TOFD) scan plan for 0.375” thick steel welded plate.

TOFD detected both the root and mid thickness lack-of-fusions. The TOFD data shown includes the lateral wave, backwall, and shear wave converted backwall at 6.5 us, 7.3 us, and 9 us, respectively. The individual waveform phases are important to consider. The lateral wave and backwall should have opposite phases. In the A-scan, this is confirmed via the first strong lateral wave negative amplitude peak and the first strong backwall positive amplitude peak. The lateral wave – backwall opposite phase phenomenon is also observed in TOFD B-scan. Notice the black line at the start of the lateral wave denoting negative amplitude and that a white strip, indicating positive amplitude, is observed at the start of the backwall.

The data analysis is performed mainly between the TOFD lateral and backwall waves, and depending on the thickness, can be a tight window to discriminate between different diffractors. In some circumstances the TOFD data present between the L-wave backwall and shear wave refracted backwall may be analyzed or at least provide data redundancy. The two LOFs, are clearly visible between the TOFD lateral wave and backwall. The root lack-of-fusion, is detected from 0.743” to 1.191” on the B-scan. The mid-wall lack-of-fusion is detected from 2.60” to 3.00”. The height of both LOFs is difficult to determine since there is no clear differentiation between a possible top and bottom diffractions. Additionally, there is no flaw depth information available.

Total Focusing Method / Full Matrix Capture Testing for Lack of Fusion

A multi-scan approach was required for the total focusing method / full matrix capture methods. The first objective was to establish a sound field that directed ultrasonic energy towards the bottom of the weld in order to detect the root lack of fusion. The established zone is shown below and encompasses the entire thickness and about 0.30” in the index axis dimension. A pulsed-echo TT wave set was used to detect the root LOF. The root LOF corner trap was detected clearly at high intensity as was the tip of the LOF. The length and height were measured be 0.80” long and 0.120”, respectively. Interesting, the mid-wall LOF was detected at the incorrect height using the TFM/FMC pulse-echo TT set. The second LOF was detected in the lower half of the plate but is actually positioned mid-wall to upper half.

Figure 3: and total focusing method / full matrix capture (TFM/FMC) scan plan for 0.375” thick steel welded plate.

Due to the position of the mid-wall LOF along the fusion line of the base and weld metal, a separate TFM/FMC wave set is required for detection. This type of weld defect is best detected in the second leg if the transducer is also located on the same side of the weld. In this case, a pulse-echo TT-TT wave set was used to detection the mid-wall LOF. The scan shows the LOF aligning with the weld fusion line over a 0.50” length. The LOF height was measured at 0.20”.


This article compared phased array ultrasonic testing (PAUT), time-of-flight diffraction (TOFD), and total focusing method / full matrix capture (TFM/FMC) setup requirements and performance a welded plate with root lack of fusion and mid-wall lack-of-fusion. All three methods detected both LOF and provide accurate length measurement along the scan axis. PAUT measured the flaw heights accurately. Flaw height measurement was not possible with TOF since there was note enough separation between the bottom and top tips of LOFs. Two separate setups were required to detect the LOFs when using TFM/FMC. A TT wave set was required to the root lack of fusion but also detected the second mid-wall LOF at an incorrect depth due largely to the orientation of the flaw. A second TT-TT wave set was created to detect the LOF in the second leg. The TFM data from this setup aligned nicely with the fusion line of the base and weld metal.

993 views0 comments


bottom of page