Martin ROBIN
Monday 15 July 2019
Amphitheatre IEMN Dpt OAE - UPHF - Valenciennes
Jury:
- Bertrand AUDOIN, Professor at the University of Bordeaux (Rapporteur)
- Mounsif ECHERIF EL KETTANI, Professor at the University of Le Havre (Rapporteur)
- Hossep ACHDJIAN, Senior Lecturer at INSA Centre Val De Loire (Examiner)
- Ms Dorothée CALLENS-DEBAVELAERE, Senior Lecturer at UPHF (Examiner)
- Frédéric JENOT, Professor at the UPHF (Co-Director)
- Mohammadi OUAFTOUH, Professor at the UPHF (Co-Director)
Summary:
Non-destructive characterisation of the adhesion of film-on-substrate structures is a major industrial and academic challenge. This type of sample is used for many applications and its lifespan depends largely on the quality of adhesion of the films to the substrate. This significantly modifies the dispersive behaviour of surface acoustic waves propagating in this type of structure. To generate and detect these waves, an Ultrasound-Laser device was chosen.
Firstly, we sought to overcome the difficulties of interpretation usually encountered in the control of adhesion by surface acoustic waves. Variations in layer thickness can have an influence on wave dispersion comparable to that caused by adhesion. To achieve this, polymer films of virtually constant thickness are used and applied to an aluminium substrate. These films also have the added property of being transparent. This means that the laser pulse generating the acoustic waves can be focused through the film directly onto the surface of the substrate, placing the acoustic source at the film-substrate interface. The influence of the position of the source on the dispersive behaviour of surface acoustic waves and consequently on the control of adhesion quality is then studied experimentally and through finite element simulations.
Finally, the adhesion of the various samples is characterised using the dispersion curves obtained using the Matrix-Pencil method applied to the experimental results. Using an inversion algorithm, the interface stiffnesses characteristic of the adhesion of the samples analysed are estimated.