This thesis, made in the non-destructive testing field, presents two new adaptive methods, looking for real-time imaging of structures with complex and irregular surfaces. These methods have been developed for immersion testing (contactless) where water ensure the transmission of the ultrasonic waves in the solid. The general principle of adaptive imaging is to acoustically measure the surface under the probe, then to determine the ultrasonic paths through the complex surface to produce an image inside the material. One of the methods that can be applied to adaptive imaging is the SyntheticTransmit Aperture (STA) method. It provides high quality images and offers the possibility of using different reconstruction modes to improve the defect characterization, depending on their geometry and orientation. However, it suffers from two major drawbacks: a large amount of data to be stored and processed, and a higher sensibility to random or structural noise than the conventional imaging methods. To overcome these drawbacks, we propose the Plane Wave Imaging (PWI) method, based on plane wave transmissions. The results on a plane surface show that the two methods give similar results in terms of spatial resolution, but PWI is less sensitive to random noise. They also reveal that PWI improves greatly image quality when defects are cracks, and that very few transmissions are needed compared to STA. The STA and PWI methods are then generalized to complex surfaces, and coupled with surface measurement methods to perform adaptive imaging. The two methods give the same images that those obtained when all the control parameters are known and the adaptive PWI keeps the advantages introduced previously: low sensibility to random noise and reduced number of transmission.
展开▼