Robotic Mapping for NDE
Supervisor: Professor Gordon Dobie
The RCNDE/NDEvR research priorities have identified the need for faster large area coverage for NDE as a priority, with deployment in hostile environments or those with difficult access also of great interest. Robotic implementation of NDE can meet these requirements, but is currently limited to inefficient ‘brute force’ data collection where the NDE probe is scanned over every point on the surface which generates an abundance of data with a low information content. This proposal represents the first-step in a paradigm shift from automated NDE imaging to intelligent closed-loop inspection, incorporating recent advancements in robotic mapping. It will investigate mobile Electromagnetic Acoustic Transducers (EMATs) that can send pitch-catch ultrasonic signals through the sample intelligently reconfiguring during the inspection. The real-time creation of an NDE map removes the need for manual control or complex automation. Once an area of concern has been identified, high resolution thickness and eddy current probes can be used for a high accuracy localised inspection.
There have been several recent advances in non-contact NDE which mean it is now timely for implementation into robotic crawlers. RCNDE partners have developed high frequency eddy current probes which offer excellent defect detectability on challenging materials. A prototype has been produced with built-in analysis, removing some of the difficulties with cabled operation.
Historic limitations have meant that both EMATs and robotic NDE have, to date, found relatively niche application in industry. However, by combining the non-contact nature and flexible wave generation of EMATS, with the ability of robotic crawlers to accurately relocate the probes, it is possible to create a truly next generation inspection system that is greater than the sum of its parts. The use of multiple robots allow novel imaging/mapping strategies. The transducers, robots, and data communication will be optimised to enable repeatable measurements to be undertaken. Data fusion between the different payloads will be investigated.