Miroirs deformables a base de ferrofluide: Simulations numeriques et etude preliminaire du controle

This work is a preliminary theoretical and experimental exploration of a new type of deformable optics that can be controlled by a magnetic field. The optical component studied is a ferrofluid based deformable mirror. The ferrofluid, a colloidal suspension of magnetic particles, is coated with a reflective layer and put over a matrix of coils. Without the mechanical constraints of solid surfaces, this mirror allows exceptional deformations compared to conventional deformable mirrors. These mirrors also offer an important economic advantage, because of the polishing step avoided and the low actuators cost.; This new class of deformable mirror is destined to be applied to adaptive optics. This project is then a first analysis of spatial and temporal behavior of the mirror. First, an overview of the current deformable mirror technology and the context of most of the applications are presented. After follows introduction about the composition of ferrofluids and the theory of magnetic fields and ferrohydrodynamics. The finite element simulations requirements, that are formulations of the equations and numerical techniques are detailed. Numerical simulations of different ferrofluids in motion under the effect of an external magnetic field have been done and are analyzed for a better comprehension of the laboratory experiments. To characterize the mirror spatially, a whole chapter is devoted to the experimental results of the static behavior of the ferrofluid-based mirror. It was demonstrated that the actuators interaction is not the same as usually found in conventional deformable mirrors. This fact leads to a need to develop a new control algorithm. Following these experimental explorations, a study of the mirror control with a new kind of influence matrix is presented. The application of this control matrix hasn't yet rigorously tested, but experimental issues are presented. The experimental and numerical results allow concluding that even if this liquid mirror needs new control algorithms, a fully functional mirror in a close future is possible and could be a proof of the use of new control algorithm for a mirror, which has non common advantages. Finally, ideas are proposed to improve the mirror in the adaptive optics context.