Ultrasonic sensor platforms for non-destructive evaluation

Dobie, Gordon and Hayward, Gordon and Pierce, Gareth (2010) Ultrasonic sensor platforms for non-destructive evaluation. PhD thesis, University Of Strathclyde.

[thumbnail of Dobie-2010-Ultrasonic-sensor-platforms-for-non-destructive-evaluation]
Text. Filename: Dobie_2010_Ultrasonic_sensor_platforms_for_non_destructive_evaluation.pdf
Final Published Version

Download (11MB)| Preview


Robotic vehicles are receiving increasing attention for use in Non-Destructive Evaluation (NDE), due to their attractiveness in terms of cost, safety and their accessibility to areas where manual inspection is not practical. A reconfigurable Lamb wave scanner, using autonomous robotic platforms is presented. The scanner is built from a fleet of wireless miniature robotic vehicles, each with a non-contact ultrasonic payload capable of generating the A0 Lamb wave mode in plate specimens. An embedded Kalman filter gives the robots a positional accuracy of 10mm. A computer simulator, to facilitate the design and assessment of the reconfigurable scanner, is also presented. Transducer behaviour has been simulated using a Linear Systems approximation (LS), with wave propagation in the structure modelled using the Local Interaction Simulation Approach (LISA). Integration of the LS and LISA approaches were validated for use in Lamb wave scanning by comparison with both analytical techniques and more computationally intensive commercial finite element/diference codes. Starting with fundamental dispersion data, the work goes on to describe the simulation of wave propagation and the subsequent interaction with artificial defects and plate boundaries. The computer simulator was used to evaluate several imaging techniques, including local inspection of the area under the robot and an extended method that emits an ultrasonic wave and listens for echos (B-Scan). These algorithms were implemented in the robotic platform and experimental results are presented. The Synthetic Aperture Focusing Technique (SAFT) was evaluated as a means of improving the fidelity of B-Scan data. It was found that a SAFT is only effective for transducers with reasonably wide beam divergence, necessitating small transducers with a width of approximately 5mm. Finally, an algorithm for robot localisation relative to plate sections was proposed and experimentally validated.