Phased array ultrasonic inspection of embedded defects in hydropower turbine runner welds

In-service detection of embedded defects in hydropower turbine runner welds is regarded as a challenging problem due to the limited inspection space and complex structure of the component, which affect conventional ultrasonic testing. In this paper, a phased array approach is adopted and studied to prepare for field application. Two blocks including artificial defects are made and used to demonstrate the detection capability. A defect identification and location method is developed, which uses the superimposition of the sectional image on a 2D geometry model of the component and the mirror point of the echo calculated by acoustic line propagation theory. An automatic scanning robot is made and used on real turbine weld inspections and the result is satisfactory.