Phased Array Shot Scenario and Shot Sequence Optimization for Crack Detection Inline Inspection Tools

Abstract

Ultrasonic crack detection (UC) technology is a well-established methodology to address crack-like flaws in liquid product pipelines. The UC technology is based on shear bulk waves propagating in the pipe wall. For monolithic transducers, the angle of incidence (AOI) is defined by the mechanical inclination of the transducers with respect to the internal pipe wall surface. Changes in the AOI due to the differences in acoustic properties of the different liquid products therefore require changes in the mechanical layout. Ultrasonic inspection tools based on phased arrays offer a much higher level of flexibility regarding the ultrasound characteristics (e.g., AOI, beam width, focus) compared to tools based on monolithic transducers. In particular, multiple array elements can be grouped together in so-called virtual sensors and placed with element-pitch granularity on the array. This flexibility allows to develop a shot scenario optimized for the specific requirements presented in an inspection project, e.g., challenging long seam weld geometries. However, exploiting this configuration flexibility comes with a drawback: Each configuration requires an in-depth analysis of sound propagation paths between array and pipeline wall in order to avoid spurious sound propagation that can lead to undesired crosstalk. This analysis becomes even more relevant if multiple virtual sensors must be used simultaneously to guarantee high tool inspection speed. The authors present the methodology used in the phased array inline inspection tools to enable simultaneous usage of several virtual sensors on each array and to minimize the impact of spurious sound propagation on subsequent usages of virtual sensors. In particular, the authors demonstrate the application of a model of the inspection setup / shot scenario that allows computerized optimization of possible configuration options. We outline the configuration optimization procedure and present experimental validation results from pump tests confirming that data quality and inspection speed are well aligned with customer expectations.