In many environmental and climate applications, extreme data are spatial by nature, and hence statistics of spatial extremes is currently an important and active area of research dedicated to developing innovative and flexible statistical models that determine the location, intensity, and magnitude of extreme events. In particular, the development of flexible sub-asymptotic models is in trend due to their flexibility in modeling spatial high threshold exceedances in larger spatial dimensions and with little or no effects on the choice of threshold, which is complicated with classical extreme value processes, such as Pareto processes. In this thesis, we develop new flexible sub-asymptotic extreme value models for modeling spatial and spatio-temporal extremes that are combined with carefully designed gradient-based Markov chain Monte Carlo (MCMC) sampling schemes and that can be exploited to address important scientific questions related to risk assessment in a wide range of environmental applications. The methodological developments are centered around two distinct themes, namely (i) sub-asymptotic Bayesian models for extremes; and (ii) flexible marked point process models with sub-asymptotic marks. In the first part, we develop several types of new flexible models for light-tailed and heavy-tailed data, which extend a hierarchical representation of the classical generalized Pareto (GP) limit for threshold exceedances. Spatial dependence is modeled through latent processes. We study the theoretical properties of our new methodology and demonstrate it by simulation and applications to precipitation extremes in both Germany and Spain. In the second part, we construct new marked point process models, where interest mostly lies in the extremes of the mark distribution. Our proposed joint models exploit intrinsic CAR priors to capture the spatial effects in landslide counts and sizes, while the mark distribution is assumed to take various parametric forms. We demonstrate that having a sub-asymptotic distribution for landslide sizes provides extra flexibility to accurately capture small to large and especially extreme, devastating landslides.

Cette étude s’inscrit dans les efforts faits par l’industrie aéronautique pour la protection environnementale.Pour cela, un des objectifs principaux est de réduire les émissions de polluants et les émissions sonores des aéronefs. Les émissions polluantes sont liées à la qualité de la combustion qui dépend elle même de la conception des chambres de combustion. Les émissions sonores sont réduites grâce à des traitements passifs comme les structures absorbantes placées en paroi des moteurs pour diminuer le bruit de soufflante. Pour ces deux applications différentes, on utilise des matériaux perforés. En effet, les matériaux absorbants sont constitués d’une structure en nid d’abeilles surmontée d’une plaque perforée. Dans les chambres de combustion, les perforations permettent l’injection d’un air froid protégeant les parois des hautes températures, mais leur présence peut modifier la caractérisation acoustique de la chambre. L’objectif de cette thèse est de modéliser la réponse acoustique d’un matériau perforé. La taille des perforations étant petite devant les longueurs d’ondes sonores, des techniques de modélisation asymptotique adaptées à la résolution de problèmes multi-échelle peuvent être mises en œuvre. En effet, ces méthodes permettent de faire le lien entre les effets présents à l’intérieur d’une perforation et la réponse acoustique homogénéisée de la plaque perforée. Dans ces travaux,ce sont les effets visqueux présents dans les perforations qui ont été essentiellement étudiés. Ensuite,des simulations numériques directes ont été réalisées pour vérifier la validité des hypothèses émises lors de la modélisation asymptotique. Ces travaux de thèse ont permis d’améliorer la compréhension de la modélisation de la réponse acoustique des matériaux perforés. De plus, la méthode analytique présentée peut être mise en œuvre pour des perforations de géométrie complexe. / This study is part of the effort made by the aeronautic industry to protect the environment. Thus,one of the main objectives is to reduce aircraft polluting emissions and sound emissions. The pollutinge missions are linked to the combustion quality which depends on the combustion chamber’s design.Noise pollution is reduced thanks to passive treatments such as absorbing structures placed on engine walls. For both applications, perforated plates are used. Indeed, absorbing materials are composed of honey comb structures topped by a perforated plate. In combustion chambers, cold air is injected through perforations to protect the walls from high temperatures, but the perforations can modify the chamber’s acoustic behavior. The objective of this thesis is to model the acoustic response of aperforated material. The perforation size is smaller than the sound wave length, therefore asymptotic modeling techniques adapted to resolving multiscale problems can be implemented. Indeed, thesemethods allow to link the effects inside the perforation with the homogenized acoustic response of theperforated plate. In this study, the viscous effects inside the perforation have been analysed. Moreover,direct numerical simulations have been performed to verify the asymptotic modeling hypotheses. The findings of this thesis allow to understand the acoustic modeling of perforated materials. Furthermore,the developed analytical method can be implemented for perforations with complex shape.

Composites are materials which cater to the present and future needs of many demanding industries, such as aerospace, as they are weight-sensitive for a given requirement of strength and stiff ness, corrosion resistant, potentially multi-functional and can be tailored according to the application. However, they are in particular difficult to join as they cannot be easily machined, without introducing damages which can eventually grow. Any structure is as strong as its weakest joint. Most of the joints belong to the category of mechanically-fastened joints and they pose enormous challenges in modeling due to contact phenomena, nonlinearity and stress concentration factors. It is therefore a necessity to construct an efficient model that would include all the relevant contact phenomena in the joints, as it has been pointed out in literature that damage typically initiates near the joint holes. The focus of this work is to describe the construction of an asymptotically-correct model using the Variational Asymptotic Method (VAM). Amongst its many potential applications, VAM is a well-established analytical tool for obtaining the stress and strain fields for beams and shells. The methodology takes advantage of the small parameter that is inherent in the problem, such as the ratio of certain characteristic dimensions of the structure. In shells and beams, VAM takes advantage of the dimension-based small parameter(s), thereby splitting the problem into 2-D + 1-D (for beams) and 1-D + 2-D (for shells), in turn offering very high computational efficiency with very little loss of accuracy compared to dimensionally unreduced 3-D models. In this work, the applicability of VAM is extended to two-dimensional (2-D) and three-dimensional (3-D) frictionless contact problems. Since a generalised VAM model for contact has not been pursued before, the `phantom0 step is adopted for both 2-D and 3-D models. The development of the present work starts with the construction of a 2-D model involving a large rectangular plate being pressed against a rigid frictionless pin. The differential equations governing the problem and the associated boundary conditions are obtained by minimizing the reduced strain energy, augmented with the appropriate gap function, by using a penalty method. The model is developed for both isotropic and orthotropic cases. The boundary value problem is solved numerically and the displacement field obtained is compared with the one obtained using commercial software (ABAQUSr) for validation at critical regions such as the contact surfaces. Banking on the validation of the 2-D model, a 3-D model with a pin and a finite annular cylinder was constructed. The strain energy for the finite cylinder was derived using geometrically exact 3-D kinematics and VAM was applied leading to the reduction in the strain energy for isotropic and orthotropic materials in rectangular and cylindrical co-ordinates. As in the 2-D case, the reduced strain energy, subject to the inequality constraint of the gap function, is minimized with respect to the displacement field and the corresponding boundary value problem is solved numerically. The displacements of the contact surface and the top surface of the annular cylinder are compared with those from ABAQUS and thus validated. The displacement fields obtained using the current 2-D and 3-D models show very good agreement with those from commercial finite element software packages. The model could be re ned further by using the gap function derived in this work and applying it to a plate model based on VAM, which could be explored in the future.