Data merging of infrared and ultrasonic images for plasma facing components inspection

For steady-state magnetic thermonuclear fusion devices which need large power exhaust capability, actively cooled plasma facing components have been developed. In order to guarantee the integrity of these components during the required lifetime, their thermal and mechanical behaviour must be assessed. Before the procurement of the ITER Divertor, the examination of the heat sink to armour joints with non-destructive techniques is an essential topic to be addressed. Defects may be localised at different bonding interfaces. In order to improve the defect detection capability of the SATIR technique, the possibility of merging the infrared thermography test data coming from SATIR results with the ultrasonic test data has been identified. The data merging of SATIR and ultrasonic results has been performed on Carbon Fiber Composite (CFC) monoblocks with calibrated defects, identified by their position and extension. These calibrated defects were realised with machining, with “stop-off” or by a lack of CFC activation techniques, these last two representing more accurately a real defect. A batch of 56 samples was produced to simulate each possibility of combination with regards to interface location, position and extension and way of realising the defect. The use of a data merging method based on Dempster–Shafer theory improves significantly the detection sensibility and reliability of defect location and size.