This thesis investigates analysis and simulation of the backscattering enhancement phenomenon from randomly distributed point scatterers. These point scatterers are randomly distributed within a cube or a sphere and then the backscattering enhancement phenomenon from both cubical and spherical distributions are examined throughout the thesis. The general characteristic differences between cubical and spherical distribution about the scattering phenomenon are observed. T-matrix method is used for analytic investigations of the backscattering enhancement and also a certain number of approximate formulas are obtained. As for Monte Carlo simulation method, it is used for simulated investigations of the backscattering enhancement. Some Monte Carlo simulations are prepared by using MATLAB programming language and verified by showing their confidence intervals. Both analytic and simulated investigations of the backscattering enhancement due to single and double scattering are analyzed / however, only simulated investigation of the backscattering enhancement due to multiple scattering are analyzed because of its computational complexity. The thesis traces differences between single scattering and multiple scattering from randomly distributed point scatterers. Effects of both incident field frequency and point scatterer density on the backscattering enhancement are indicated. The thesis seeks answers to questions such as which conditions cause the backscattering enhancement phenomenon from randomly distributed point scatterers, why we need to consider multiple scattering to examine the backscattering phenomenon and how we can discriminate the backscattering enhancement from the specular enhancement.

Dans le cadre d'un consortium entre centres de recherche publics et industriels, ce travail de thèse de doctorat s'est attaché à démontrer l'intérêt des détecteurs photo-organiques (OPS) pour la caractérisation des suspensions et écoulements diphasiques. Les principes de plusieurs granulomètres permettant la caractérisation de ces milieux lorsqu'ils sont confinés dans une cuve cylindrique transparente (configuration standard du Process Analytical Technology) ont été proposés. Pour évaluer et optimiser les performances de ces systèmes, un code de simulation de type Monte-Carlo a été spécifiquement développé. Ce dernier permet de prendre en compte les nombreux paramètres du problème comme le profil du faisceau laser, les différentes surfaces spéculaires composant le montage, la composition du milieu particulaire (concentration, diamètre moyen, écart-type, matériau,...), la forme et la position des OPS. Les propriétés de diffusion des particules sont traitées à l'aide des théories de Lorenz-Mie et de Debye, de même qu'un modèle hydride prenant en compte les contributions géométriques et physiques. Pour les milieux dilués (diffusion simple), l'analyse repose sur l'inversion des diagrammes de diffusion obtenus sur une large plage angulaire ou au voisinage de singularités optiques remarquables (arc-en-ciel, diffusion critique, diffraction). Pour les milieux denses (diffusion multiple), les pistes étudiées reposent sur l'analyse des caractéristiques de la tache de rétrodiffusion. / As part of a consortium between academic and industry, this PhD work investigates the interest and capabilities of organic photo-sensors (OPS) for the optical characterization of suspensions and two-phase flows. The principle of new optical particle sizing instruments is proposed to characterize particle systems confined in a cylinder glass (standard configuration for Process Analytical Technologies). To evaluate and optimize the performance of these systems, a Monte-Carlo model has been specifically developed. This model accounts for the numerous parameters of the system: laser beam profile, mirrors, lenses, sample cell, particle medium properties (concentration, mean & standard deviation, refractive indices), OPS shape and positions, etc. Light scattering by particles is treated either by using Lorenz-Mie theory, Debye, or a hybrid model (that takes into account the geometrical and physical contributions). For diluted media (single scattering), particle size analysis is based on the inversion of scattering diagrams obtained over a wide angular range or near optical singularities (rainbow, critical scattering, diffraction). For dense media (multiple scattering), the solutions foreseen are based on the analysis of the backscattering spotlight characteristics.