L’étude des seuils de tolérance du corps humain aux impacts requiert des expérimentations sur des sujets humains vivants ou post-mortem, ce qui soulève naturellement des questions d’éthique. Pour pallier à ces limitations, le développement des outils numériques a conduit, au fil des dernières années à la mise en place des mannequins numériques plus ou moins capables de reproduire fidèlement le comportement mécanique du corps humain lorsqu’ils sont soumis à divers types de sollicitations. C’est dans ce contexte que le modèle de mannequin numérique HUByx (Hermaphrodite Biomechanics yx-model) a été développé au sein du département de recherche COMM du laboratoire ICB à l’UTBM. Ce travail de thèse a pour but la validation et l’amélioration de la biofidélité de la partie thoracique du modèle HUByx, et vise à comprendre les mécanismes de lésions et à rechercher des critères de prédiction des lésions thoraciques à travers la reconstruction numérique des chargements violents dans des contextes civils et militaires. Des simulations numériques ont été réalisées dans le cadres des études de chutes libres de personnes, des impacts balistiques non-pénétrants de projectiles non létaux et dans le cadre du phénomène de l’explosion. De bonnes corrélations ont été obtenues entre les résultats numériques et expérimentaux, contribuant ainsi à renforcer la capacité du modèle HUByx à répondre de manière biofidèle aux différentes sollicitations auxquelles il est soumis. Mots-clés : Biomécanique d’impacts, éléments finis, blessures thoraciques, critère visqueux, reconstruction d’accident, chutes libres, balistique, explosion. / The study of human tolerance thresholds to impacts requires experiments on living or post mortem human subjects, which naturally raises ethical questions. To overcome these limitations, the development of numerical tools has led over the last few years to the implementation of numerical models more or less capable to accurately reproduce the mechanical behavior of the human body when subjected to various types of stresses. It is in this context that the numerical model HUByx (Hermaphrodite Biomechanics yx-model) has been developed within the research department COMM of the ICB lab at UTBM. This PhD work aims at validating and improving the biofidelity of the thoracic part of the HUByx model and also aims to understand the mechanisms of lesions and to seek criteria for the thoracic injury prediction through the numerical reconstruction of violent loadings in civil and military contexts. Numerical simulations were carried out in the framework of human free falls studies, non-penetrating ballistic impacts of non-lethal projectiles and finally in the context of explosion phenomenon. Good correlations were obtained between the numerical and experimental results, thus contributing to reinforce the capacity of the HUByx model to respond in a biofidelic manner to the different stresses to which it is subjected. Keywords : Impact biomechanics, finite elements, thoracic injuries, viscous criterion, accident reconstruction, free falls, ballistics, explosion.

The impact phenomena of high velocity micron-size particles, although commonly considered and described as detrimental in numerous engineering applications, can be used in a beneficial way if properly understood and controlled. The Cold Gas Dynamic Spray (CGDS) process, known as a surface modification, repair and additive manufacturing process, relies on such high velocity impacts. In the process, solid particles are accelerated by a supersonic gas flow to velocities up to 1200 m/s and are simultaneously heated to temperatures lower than their melting point. When propelled under proper velocity and temperature, the particles can bond onto a target surface. This bonding is caused by the resulting interfacial deformation processes occurring at the contact interface. Hence, the process relies heavily on the gas/particle and particle/substrate interactions. Although numerous experimental and/or numerical studies have been performed to describe the phenomena occurring during particle flight and impact in the CGDS process, numerous phenomena remain poorly understood. First, the effect of substrate surface topographical condition on the particle deformation and ability to successfully adhere, i.e. atomically and/or mechanically, has not been thoroughly investigated such that its influence is not well understood. Another aspect of the process that is generating the largest gap between experimental and numerical studies in the field is the lack of particle in-flight temperature measurements. Obtaining such data has proven to be technically difficult. The challenges stem from the short particle flight time, low particle temperature and small particle size preventing the use of established thermal spray pyrometry equipment. Relatedly, lack of such measurements precludes a proper experimental study of the impact related phenomena at the particle/substrate interface. As a result, the effect of particle size dependent temperature on overall coating properties and atomic bonding relies currently on estimates. Finally, the effect of particle impact characteristics on interfacial phenomena, i.e. grain size and geometry, velocity/temperature, and oxide scale thickness, on adhesion and deformation upon single particle collision has also been scarcely studied for soft particle depositions on hard substrate. Hence, the current research work aims at studying fundamental aspects of particle/gas heat transfer and particle/substrate impact features in goals to improve the understanding of the CGDS process. Different surface preparation methods will be used to create various surface roughness and topographical features, to provide a clear understanding of the target surface state influence on coating formation and adhesion. Additionally, new equipment relying on novel technology, i.e. high-speed IR camera, will be utilized to obtain particle in-flight temperature readings with sequence recordings. Subsequently, the experimental particle in-flight temperature readings will be used to develop a computational fluid dynamics model in goals to validate currently used Nusselt number correlations and heat transfer equations. The particle size-dependent temperature effect on the particle’s elastic and plastic response to its impact with a targeted surface and its ability to successfully bond and form a coating will be studied experimentally. A thorough CFD numerical work, based on experimental findings, will be included to provide full impact characteristics (velocity, temperature, size and trajectory) of successfully deposited particles. Finally, the numerical results will be utilized in the ensuing study to correlate single particle deformation, adhesion and interfacial features to impact characteristics. A finite element model will be included to investigate the effect of particle size dependent temperature on single particle interfacial pressure, temperature and bonding ability.

<p>Raman spectroscopy is a molecular spectroscopy technique that uses monochromatic light to provide a fingerprint to identify structural components and chemical composition. Depending on the changes in the unit-cell parameters and volume under the application of stress and temperature, the Raman spectrum undergoes changes in the wavenumber of Raman-active modes that allow identification of sample characteristics. Due to the various advantage of mechanical Raman spectroscopy (MRS), the use of this technique in the characterization and modeling of chemical changes under stress and temperature have gained popularity. </p> <p> Quantitative information regarding the local behavior of interfaces in an inhomogeneous material during shock loading is limited due to challenges associated with time and spatial resolution. Recently, we have extended the use of MRS to high-strain rate experiments to capture the local thermomechanical response of mock energetic material and obtain material properties during shock wave propagation. This was achieved by developing a novel method for <i>in‑situ</i> measurement of the thermo‑mechanical response from mock energetic materials in a time‑resolved manner with 5 ns resolution providing an estimation on local pressure, temperature, strain rate, and local shock viscosity. The results show the solid to liquid phase transition of sucrose under shock compression. The viscous behavior of the binder was also characterized through measurement of shock viscosity at strain rates higher than 10⁶/s using microsphere impact experiments.</p> <p> This technique was further extended to perform Raman spectral imaging over a microscale domain of the sample with a nano-second resolution. This was achieved by developing a laser-array Raman spectral imaging technique where simultaneous deconvolution of Raman spectra over the sample domain was achieved and Raman spectral image was reconstructed on post-processing. We developed a Raman spectral imaging system using a laser array and analysis was performed over the interface of sucrose crystals bonded using an epoxy binder. This study provides the Raman spectra over the microstructure domain which enabled the detection of localized melting under shock compression. The distribution of shock pressure and temperature over the microstructure was obtained using mechanical Raman analysis. The study shows the effects of an actual interface on the propagation of shock waves where a higher dissipation of shock energy was observed compared to an ideal interface. This increase in shock dissipation is accompanied by a decrease in both the maximum temperature, as well as the maximum pressure within the microstructure during shock wave propagation.</p>