Lab & Field Testing of AUT Systems for Steel Highway Bridges

The results of the fourth set of field tests at HSS are summarized in table 11. There were 10 field specimens inspected; 5 of the 10 specimens were designated as FCMs. Table 11 indicates that three FCM and three non-FCM specimens were accepted with no rejectable defects by all of the inspection methods. The 76.2-mm- (3-inch-) thick field specimen FG40M-TF1-Curved-FCM (figures 120 and 121) is a plate specially fabricated for a railroad bridge. Unlike conventional plates that are generally fabricated flat, specimen FG40M-TF1-Curved-FCM was fabricated at a right angle. RT accepted the weld with no rejectable defects. Since specimen FG40M-TF1-Curved-FCM is an FCM, it is required to be inspected by UT. Two rejectable indications were detected by manual UT and one rejectable indication was detected by AUT. Figure 122 shows a radiographic image of the weld section where UT detected rejectable indications. Figures 123 through 130 show the P-scan images created by scanning the weld with a 45-degree probe. The rejectable indication is located between Y = 228.6 mm and Y = 457.2 mm (Y = 9 inches and Y = 18 inches) (figures 125 and 126). According to table 6.2 in the AASHTO/AWS D1.5M/D1.5: 2002 Bridge Welding Code,⁽¹⁾ a 76.2-mm- (3-inch-) thick plate is required to be scanned by 45-degree and 70-degree probes. However, a P-scan inspection of the weld with a 70-degree probe was not performed because of scheduling constraints. The 50.8-mm- (2-inch-) thick field specimen FG1A-TF2-BottF-FCM (figures 131 and 132) was rejected by all three inspection methods. RT rejected the specimen based on a rejectable slag inclusion in section A‑B. According to table 6.2 in the AASHTO/AWS D1.5M/D1.5: 2002 Bridge Welding Code,⁽¹⁾ a 50.8-mm- (2-inch-) thick plate must be scanned by 60-degree and 70-degree probes. Manual UT and AUT rejected the specimen based on an indication detected with the 60-degree and 70-degree probes. The P-scan images of specimen FG1A-TF2-BottF-FCM, scanned with 60‑degree and 70-degree probes, are shown in figures 133 through 135. Specimen G3VHW-CF1-BottF (Figure 136) was rejected by RT based on a slag inclusion found at location Y = 250.952 mm (Y = 9.88 inches). The radiographic image of this weld section is shown in figure 137, but the defect is not clearly visible. Manual UT and AUT inspection with a 70-degree probe found an acceptable indication at Y = 250.952 mm (Y = 9.88 inches). Figure 138 shows the P-scan images from Y = 0 mm to Y = 304.8 mm (Y = 0 inches to Y = 12 inches) for specimen G3VHW-CF1-BottF. Field specimen G5VHW-CF1-BottF (figures 139 and 140) was rejected by RT, but was accepted by manual UT and AUT. RT found a 6.35-mm- (0.25-inch-) long slag inclusion at Y = 639.83 mm (Y = 25.29 inches) (Figure 141). The defect is not clearly visible in the radiographic image because of the poor quality of the radiographic film. Specimen G5VHW-CF1-BottF was not required to be tested by manual UT; however, the fabricator used UT to determine the depth of the defect for repair purposes. Manual UT and AUT identified the defect found by RT and this indication was found to be acceptable. The P-scan images from segment 4 in specimen G5VHW-CF1-BottF were produced using 45-degree and 70-degree probes (figures 142 through 144). Figure 142 illustrates that the indication detected by AUT was within code requirements.

The results of the field tests at Stupp are summarized in table 12. There were four field specimens tested, with all three inspection methods accepting all four specimens.

Figure 120. Field specimen FG40M-TF1-Curved-FCM: Top view of joint.

Figure 121. Field specimen FG40M-TF1-Curved-FCM: Side view of joint.

Figure 127. P-scan images of field specimen FG40M-TF1-Curved-FCM: From TSC side of centerline between 457.2 and 685.8 mm (18 and 27 inches).

Figure 128. P-scan images of field specimen FG40M-TF1-Curved-FCM: From BSC side of centerline between 457.2 and 685.8 mm (18 and 27 inches).

Figure 129. P-scan images of field specimen FG40M-TF1-Curved-FCM: From TSC side of centerline between 685.8 and 993.8 mm (27 and 39.125 inches).

Figure 142. P-scan images of field specimen G5VHW-CF1-BottF: From TSC side of centerline using 45-degree probe.

For comparison, the laboratory and field specimens with rejectable defects were grouped according to plate thickness in 25.4-mm (1-inch) increments (tables 13 and 14). These groupings summarized the results and helped determine the relationship between the effectiveness of each inspection method with respect to the plate thickness. The first column of each table shows the thickness range of the grouped plates. The second column shows the total number of specimens inspected. The third column indicates the number of specimens accepted with no rejectable defects by all three inspection methods. The fourth column shows the number of specimens rejected with rejectable defects by at least one inspection method. The fifth column identifies the rejected specimen. The sixth column shows the number of rejectable defects found. The seventh, eighth, and ninth columns indicate the inspection methods that rejected the particular specimen. The checkmark () stands for "Rejected," indicating that the specimen was rejected by the employed inspection method; X stands for "Accepted," indicating that the specimen was accepted by the employed inspection method; and DWC stands for "Detected/Within Code," indicating that the employed inspection method identified an indication that was within code requirements. Table 13 shows that the three inspection methods produced consistent results in the laboratory. In table 14, there are 8 field specimens with thicknesses ranging from 0 to 50.8 mm (0 to 2 inches) that contained 12 rejectable defects. RT rejected all eight specimens, while UT rejected only three specimens. Note that UT identified defects in the five remaining specimens, but found that the defects were acceptable. In addition, 5 specimens with thicknesses ranging from greater than 50.8 to 101.6 mm (2 to 4 inches) containing 8 rejectable defects also were identified. RT rejected only two specimens, while UT rejected all five specimens. Note that FG36K-TF2-TopF-FCM was not rejected by AUT since the defect was located at the curvature region of the width transition and the P-scan system was not configured to inspect this type of region.

Tables 15 and 16 compare the results of the RT and AUT inspections. The tables are arranged to illustrate: (1) the total number of specimens inspected within each 25.4-mm (1-inch) thickness increment, (2) the total number of specimens that were rejected by at least one inspection technique, (3) the number of specimens rejected by each inspection method, and (4) the number of specimens accepted with identifiable indications.

Table 13. Consolidating the results of laboratory testing using laboratory specimens with rejectable defects.

Thickness Range (inch)	Total No. of Laboratory Specimens	No. of Laboratory Specimens Accepted	No. of Laboratory Specimens Rejected by at Least One Technique	Rejected Laboratory Specimen	Ind. No.	Rejected by RT	Rejected by UT	Rejected by AUT
0-1	6	0	6	S033 (0.5" thick)	1
					2
				S034 (0.5" thick)	1
					2			X1
				S125 (1" thick)	1
					2
					3
				S126 (1" thick)	1
					2
					3
					4		DWC
				S135 (1" thick)	1
					2
					3
				S136 (1" thick)	1
					2
>1-2	4	2	2	S132 (1.5" thick)	1
				S133 (1.5" thick)	1

*AUT is not configured to detect transverse defects or inspect width transition regions.

1 inch = 25.4 mm

Table 15. Assessing the detectability and rejectability of three inspection methods in the laboratory.

Table 16. Assessing the detectability and rejectability of three inspection methods in the field.

Thickness Range (inch)	Total No. of Field Specimens	No. of Field Specimens With Rejectable Defects	No. of Specimens Rejected by RT	No. of Specimens Rejected by AUT	No. of Specimens Accepted by RT Within Code	No. of Specimens Accepted by AUT Within Code
0-1	5	2	2 (100%)	1 (50%)	0	1
>1-2	23	6	6 (100%)	2 (33%)	0	4
2-3	13	4	2 (50%)	4 (100%)	0	0
3-4	3	1	0 (0%)	1 (100%)	0	0

ARTICULATION ANGLES

To supplement field testing and establish the importance of the code-prescribed transducer articulation, specimen S033 was tested at a series of fixed angles (0 degrees, 2.5 degrees, 5 degrees, 7.5 degrees, and 10 degrees). The angles were measured with respect to a line normal to the weld centerline. The main purpose of articulating the transducer is to obtain the highest possible amplitude echo from a given defect. To accomplish this, the transducer must be articulated so that the direction of the shear wave propagating within the specimen is perpendicular to a plane of the defect. As mentioned earlier, specimen S033 (shown in figures 7 through 10) contained manufactured crack-like defects oriented along the centerline in the weld bevels.

The testing is performed by the P-scan system using a test frame (Figure 145) and a series of wedges designed to keep the transducer at a fixed angle during scanning (Figure 146). Figure 146 shows an angled wedge attached to a vertical sliding bar which is oriented transverse to the weld axis. The transducer is placed against the angled wedge, holding the articulation angle fixed relative to the weld axis. As the sliding bar is advanced along the length of the weld, the transducer is held at the fixed angle of interest.

The maximum echo amplitudes from the defect were determined from the P-scan images and were plotted against the transducer articulation angles (Figure 147). The abscissa in Figure 147 represents the transducer articulation angles, and the ordinate represents the indication rating (d). The indication rating of negative dB is indicative of a high-amplitude echo and the indication rating of positive dB is indicative of a low-amplitude echo. It is evident from Figure 147 that the amplitude of the echo decreases from a high value of -4 dB to a low value of +14 dB as the articulation angle increases from 0 to 10 degrees, respectively. The 0-degree articulation angle, which produces the highest indication rating of -4 dB, reveals that the direction of the incident traveling wave is aligned normal to the plane of the crack thus reflecting the highest amount of ultrasonic energy. For articulation angles other than 0 degrees, the incident traveling wave is no longer perpendicular to the plane of the crack, leading to less ultrasonic energy reflection. The difference in the echo amplitude between the articulation angles of 0 degrees and 2.5 degrees, 2.5 degrees and 5 degrees, 5 degrees and 7.5 degrees, 7.5 degrees and 10 degrees are 5 dB, 6 dB, 5 dB, and 2 dB, respectively. The 2-dB to 6‑dB difference in the echo amplitude is significant in rejecting or accepting a weld based on the indication rating criteria set forth in the UT acceptance-rejection criteria in tables 6.3 and 6.4 in the AASHTO/AWS D1.5M/D1.5: 2002 Bridge Welding Code.⁽¹⁾ Therefore, a rejectable defect may be misrepresented as an acceptable indication if articulation of the transducer is eliminated.

Figure 145. Transducer articulation testing: Test setup.

 

Figure 146. Transducer articulation testing: Various articulation angle wedges.

 

Figure 147. Influence of transducer articulation angle on the maximum amplitude of the reflected signal.