Acoustic wave scattering in random and non-linear media – DiAMAN

Our project brings together an industrial issue (the ultrasonic characterization of strongly scattering media like poly-crystals, steel, concrete or biological media) and a question of more fundamental interest (random matrices and wave propagation in complex non-linear media). Ultrasonic waves are commonly used for many applications –for example medical imaging, therapy or non destructive evaluation of materials- mostly in homogeneous or weakly inhomogeneous media, sometimes in a non-linear regime.
The general problem we consider is the propagation of ultrasonic waves in random and non-linear scattering media, with arrays of multiple transmitters/receivers. Aside from direct applications (non-destructive evaluation of materials, medical or seismic imaging), the project raises fundamental issues. We propose two complementary and original subjects :
- in the linear regime : ultrasonic imaging, detection and characterisation of random
scattering media by separating the « single scattering » from the « multiple scattering » contribution
- non-linear propagation of ultrasound in random heterogeneous media, in the multiple scattering regime.

The issues at stake can be briefly described as follows :
- Industrial/applied issues : for the ultrasonic inspection of heterogeneous or poly-crystalline media, « structural noise » i.e., the scattered signal resulting from random structural heterogeneities within the medium, causes a serious problem for the detection of flaws. Besides, even if there is no flaw, this so-called « noise » is actually a signature of the medium, and as such it contains potential information that could allow a better characterisation. Yet, apart from « simple » situations (low frequency approximation, small fluctuations...) it is still not well modelled, particularly in the multiple scattering regime. As an example, the detection of flaws in materials such as austenitic coarse grain steel is a crucial challenge for the control of nuclear power plants. The inspection of the outer layers of concrete structures with ultrasound is another example. In this project, we propose to test a “smart antenna” in order to improve detection techniques in such materials.

- Academic/fundamental issues : mesoscopic physics of waves (whatever their nature) in
complex media has undergone a tremendous revival in the last 20 years. The remarkable scientific production that resulted has benefited from the cross-fertilisation between, for nstance, acoustics and seismology, optics and acoustics, ultrasound and micro-waves… Yet, when it comes to ultrasonic waves in random scattering media, the simultaneous presence of non-linearity and strong multiple scattering (« coda » waves) has been so far very little explored. There is a fundamental physical interest in considering this problem, both from a theoretical and an experimental point of view.

To address these challenging issues, we have gathered a compact and efficient team of 7 permanent researchers from 3 close laboratories with complementary skills. The work programme associates experimental, numerical and theoretical aspects.