Abstract
Hydraulic steel structures (HSS) are components of navigation, flood control,
and hydropower projects that control or regulate the flow of water.
Damage accumulates in HSS as they are operated over time, and they must
be inspected periodically. This is often accomplished using nondestructive
testing (NDT) techniques. If damage is detected, the structure’s fitness for
continued service must be evaluated, which requires information on the
location and size of discontinuities. This information can be obtained using
ultrasonic testing (UT) techniques. However, there is limited information
on the reliability of UT techniques with respect to detecting, sizing,
and characterizing flaws in HSS. This study addresses this gap. Roundrobin
experiments were carried out using phased-array ultrasonic testing
(PAUT) to scan weld specimens representing a variety of HSS geometries.
The results of the round-robin experiments were analyzed to estimate the
probability of detection (POD) and to assess the influence of factors potentially
affecting POD. Uncertainty in estimates of flaw length and height
were described, and partial safety factors were derived for use in fitnessfor-
service analyses. These results demonstrate the importance of the
technician as a factor influencing the reliability of NDT techniques applied
to HSS.