Ultrasonic immersion testing for depth sizing of crack-like defects in pipes
R Rachev, A Velichko and P Wilcox
The structural integrity of oil pipelines is of vital importance in the petrochemical industry. The work presented focuses on inspections of pipes with ultrasonic phased arrays (PAs). The scans are performed in immersion from the inside of the pipe with emphasis on detection and sizing of axially orientated surface-breaking cracks. The project aims to propose a new data capture and processing technique applicable on oil pigs performing in-line inspections. Pigs perform inspection runs of hundreds of kilometres moving at very high velocities (2 m/s) resulting in a limited acquisition time at a specific location.
The selected post-processing algorithm is plane wave imaging (PWI). The technique offers high performance in resolution and sensitivity with respect to data acquisition volume, which makes it particularly applicable on oil pigs.
Currently, only a few published examples are available of PWI of components with non-planar surfaces in immersion. In these cases, inspections were performed by adapting the transmission delays in order to produce a plane wave inside the component rather than in the couplant. Here, this is termed PWI adapted in transmission (PWAT). PWAT requires prior knowledge of the inspection configuration, which is not always possible due to the uncertainty in the standoff and orientation of pig PAs. The work presented in the current paper proposes a new implementation termed PWI adapted in post-processing (PWAPP), focused particularly on ultrasonic immersion testing of components with non-planar surfaces. The algorithm does not require prior knowledge of the inspection configuration, while forming images of comparable sensitivity to PWAT. To apply PWAPP, a plane wave is transmitted inside the couplant, without accounting for the front surface geometry. The method utilises multiple post-processing stages to firstly reconstruct the surface of the component and then adapt the ray tracing of the PWI to account for the distortion of the incident plane waves inside the specimen.
Simulation and experimental data are produced from an immersed sample with a concave front surface featuring point-like scatterers. These have been processed with PWI with no adaptation (PWNA), PWAT and PWAPP. For a surface with a small diameter, curvature adaptation is beneficial. Near equivalent array performance indices are observed for the adapted algorithms, which both outperform PWNA. All three techniques perform similarly on components with large diameter surfaces.
An analytical forward model with flat parallel component and PA surfaces is presented and used to simulate ultrasonic phased array data in an inspection geometry close to that experienced by a pipeline inspection gauge. The data has been imaged using the total focusing method (TFM) and PWI. Image-based depth sizing has been applied using a minimum bounding box around pixels above a threshold intensity. Initial results are presented from a half-skip transverse wave imaging and sizing of crack-like defects of depths between 1 mm and 6 mm and orientation angles between –25° and 25°. TFM results show good agreement with nominal values in the –10° to 10° angular range for all depths. PWI outputs similar results with less than a sixth of the data.
The selected post-processing algorithm is plane wave imaging (PWI). The technique offers high performance in resolution and sensitivity with respect to data acquisition volume, which makes it particularly applicable on oil pigs.
Currently, only a few published examples are available of PWI of components with non-planar surfaces in immersion. In these cases, inspections were performed by adapting the transmission delays in order to produce a plane wave inside the component rather than in the couplant. Here, this is termed PWI adapted in transmission (PWAT). PWAT requires prior knowledge of the inspection configuration, which is not always possible due to the uncertainty in the standoff and orientation of pig PAs. The work presented in the current paper proposes a new implementation termed PWI adapted in post-processing (PWAPP), focused particularly on ultrasonic immersion testing of components with non-planar surfaces. The algorithm does not require prior knowledge of the inspection configuration, while forming images of comparable sensitivity to PWAT. To apply PWAPP, a plane wave is transmitted inside the couplant, without accounting for the front surface geometry. The method utilises multiple post-processing stages to firstly reconstruct the surface of the component and then adapt the ray tracing of the PWI to account for the distortion of the incident plane waves inside the specimen.
Simulation and experimental data are produced from an immersed sample with a concave front surface featuring point-like scatterers. These have been processed with PWI with no adaptation (PWNA), PWAT and PWAPP. For a surface with a small diameter, curvature adaptation is beneficial. Near equivalent array performance indices are observed for the adapted algorithms, which both outperform PWNA. All three techniques perform similarly on components with large diameter surfaces.
An analytical forward model with flat parallel component and PA surfaces is presented and used to simulate ultrasonic phased array data in an inspection geometry close to that experienced by a pipeline inspection gauge. The data has been imaged using the total focusing method (TFM) and PWI. Image-based depth sizing has been applied using a minimum bounding box around pixels above a threshold intensity. Initial results are presented from a half-skip transverse wave imaging and sizing of crack-like defects of depths between 1 mm and 6 mm and orientation angles between –25° and 25°. TFM results show good agreement with nominal values in the –10° to 10° angular range for all depths. PWI outputs similar results with less than a sixth of the data.
