Reconnaissance géométrique des structures en maçonnerie ou en béton par imagerie radar multi-récepteurs approche numérique et expérimentale
In a very restrictive economic context, the built heritage managers are eager to evaluate their structures in order to know there security state and also assess the need for maintenance. Among the most common structural disorders in old masonry structures, we find the presence of voids within the masonry and especially deep unfilled joint defects. The objective of this research is to use radar technology to increase the accuracy of this technique in recognizing geometric masonry structures and particularly the detection of deep unfilled joint defects. An imaging algorithm inspired from migration methods and based on a multi-receiver acquisition was developed. To implement this method, improvement in the use of radar data, such as determining the emission time, inaccessible to conventional radar system were needed. The development of the algorithm was made from numerical experiments on simulated environments integrating various dimensions of unfilled joint defects. A sensitivity study has also been proposed. The imaging algorithm was finally tested on real masonry structures and the effectiveness in detecting small-sized voids (deep unfilled joint defects, buried pipes) was demonstrated. The determination of wave velocity in the medium combined to the exploitation of phase in the recorded echoes allowed to precisely locate and partially identify the different interfaces of the monitored structure.
- Génie – Thèses