Mathematical models involving ordinary differential equations (ODEs) play a critical role in scientific and engineering applications. Advances in computing hardware and numerical methods have allowed these models to become larger and more sophisticated. Increasingly, problems can be described as multiphysics and multiscale as they combine several different physical processes with different characteristics. If just one part of an ODE is stiff, nonlinear, chaotic, or rapidly-evolving, this can force an expensive method or a small timestep to be used. A method which applies a discretization and timestep uniformly across a multiphysics problem poorly utilizes computational resources and can be prohibitively expensive. The focus of this dissertation is on "multimethods" which apply different methods to different partitions of an ODE. Well-designed multimethods can drastically reduce the computation costs by matching methods to the individual characteristics of each partition while making minimal concessions to stability and accuracy. However, they are not without their limitations. High order methods are difficult to derive and may suffer from order reduction. Also, the stability of multimethods is difficult to characterize and analyze. The goals of this work are to develop new, practical multimethods and to address these issues. First, new implicit multirate Runge–Kutta methods are analyzed with a special focus on stability. This is extended into implicit multirate infinitesimal methods. We introduce approaches for constructing implicit-explicit methods based on Runge–Kutta and general linear methods. Finally, some unique applications of multimethods are considered including using surrogate models to accelerate Runge–Kutta methods and eliminating order reduction on linear ODEs with time-dependent forcing. / Doctor of Philosophy / Almost all time-dependent physical phenomena can be effectively described via ordinary differential equations. This includes chemical reactions, the motion of a pendulum, the propagation of an electric signal through a circuit, and fluid dynamics. In general, it is not possible to find closed-form solutions to differential equations. Instead, time integration methods can be employed to numerically approximate the solution through an iterative procedure. Time integration methods are of great practical interest to scientific and engineering applications because computational modeling is often much cheaper and more flexible than constructing physical models for testing. Large-scale, complex systems frequently combine several coupled processes with vastly different characteristics. Consider a car where the tires spin at several hundred revolutions per minute, while the suspension has oscillatory dynamics that is orders of magnitude slower. The brake pads undergo periods of slow cooling, then sudden, rapid heating. When using a time integration scheme for such a simulation, the fastest dynamics require an expensive and small timestep that is applied globally across all aspects of the simulation. In turn, an unnecessarily large amount of work is done to resolve the slow dynamics. The goal of this dissertation is to explore new "multimethods" for solving differential equations where a single time integration method using a single, global timestep is inadequate. Multimethods combine together existing time integration schemes in a way that is better tailored to the properties of the problem while maintaining desirable accuracy and stability properties. This work seeks to overcome limitations on current multimethods, further the understanding of their stability, present new applications, and most importantly, develop methods with improved efficiency.

Numerical assessment of soil structure interaction problems require heavy computational efforts because of the dynamic and iterative (nonlinear) nature of the problems. Furthermore, modeling soil-structure interaction may require

Three- and four-derivative k-step Hermite-Birkhoff-Obrechkoff (HBO) methods are constructed for solving stiff systems of first-order differential equations of the form y'= f(t,y), y(t0) = y0. These methods use higher derivatives of the solution y as in Obrechkoff methods. We compute their regions of absolute stability and show the three- and four-derivative HBO are A( 𝜶)-stable with 𝜶 > 71 ° and 𝜶 > 78 ° respectively. We conduct numerical tests and show that our new methods are more efficient than several existing well-known methods.

Le crissement de frein est une nuisance sonore qui représente des coûts importants pour l'industrie automobile. Il tire son origine dans des phénomènes complexes à l'interface frottante entre les plaquettes de frein et le disque. L'analyse de stabilité reste aujourd'hui la méthode privilégiée dans l'industrie pour prédire la stabilité d'un système de frein malgré ses aspects sur- et sous-prédictifs.Afin de construire un système de frein robuste, il est nécessaire de trouver la technologie qui permette de limiter les instabilités malgré certains paramètres incertains présents dans le système. Ainsi, l'un des objectifs de la thèse est de développer une méthode permettant de traiter et de propager l'incertitude et la variabilité de certains paramètres dans le modèle éléments finis de frein avec des coûts numériques abordables.Dans un premier temps, l'influence d'un premier groupe de paramètres correspondant à des contacts internes au système a été étudiée afin de mieux comprendre les phénomènes physiques mis en jeu et leurs impacts sur le phénomène de crissement. Une approche basée sur l'utilisation d'un algorithme génétique a été également mise en place afin d'identifier le jeu de paramètres le plus défavorable en terme de propension au crissement sur le système.Dans un second temps, différentes méthodes de méta-modélisation ont été proposées afin de prédire la stabilité du système de frein en fonction de différents paramètres qui peuvent être des paramètres de conception ou des paramètres incertains liés à l'environnement du système.Dans un troisième temps, une méthode d'analyse non-linéaire complémentaire de l'analyse de stabilité a été proposée et développée. Elle se base sur le suivi de la stabilité d'une solution vibratoire approchée et permet d'identifier les modes instables présents dans la réponse dynamique du système. Cette méthode a été appliquée sur un modèle simple avant d'illustrer sa faisabilité sur le modèle éléments finis de frein complet. / Brake squeal is a noise nuisance that represents significant costs for the automotive industry. It originates from complex phenomena at the frictional interface between the brake pads and the disc. The stability analysis remains the preferred method in the industry today to predict the stability of a brake system despite its over- and under-predictive aspects.In order to build a robust brake system, it is necessary to find the technology that limits instabilities despite some uncertain parameters present in the system. Thus, one of the main objectives of the PhD thesis is to develop a method to treat and propagate the uncertainty and variability of some parameters in the finite element brake model with reasonable numerical costs.First, the influence of a first group of parameters corresponding to contacts within the system was studied in order to better understand the physical phenomena involved and their impacts on the squealing phenomenon. An approach based on the use of a genetic algorithm has also been implemented to identify the most unfavourable set of parameters in terms of squeal propensity on the brake system.In a second step, different meta-modelling methods were proposed to predict the stability of the brake system with respect to different parameters that may be design parameters or uncertain parameters related to the environment of the brake system.In a third step, a non-linear analysis method complementary to the stability analysis was proposed and developed. It is based on the tracking of the stability of an approximate vibrational solution and allows the identification of unstable modes present in the dynamic response of the system. This method was applied to a simple academic model before demonstrating its feasibility on the complete industrial brake finite element model under study.

Cette thèse présente l'étude de la réduction de modèles pour les réseaux électriques et les réseaux de transmission. Un point de vue mathématique a été adopté pour la réduction de modèles. Les réseaux électriques sont des réseaux immenses et complexes, dont l'analyse et la conception nécessite la simulation et la résolution de grands modèles non-linéaires. Dans le cadre du développement de réseaux électriques intelligents (smart grids) avec une génération distribuée de puissance, l'analyse en temps réel de systèmes complexes tels que ceux-ci nécessite des modèles rapides,fiables et précis. Dans la présente étude, nous proposons des méthodes de réduction de de modèles à la fois a priori et a posteriori, adaptées aux modèles dynamiques des réseaux électriques.Un accent particulier a été mis sur la dynamique transitoire des réseaux électriques, décrite par un modèle oscillant non linéaire et complexe. La non-linéarité de ce modèle nécessite une attention particulière pour bénéficier du maximum d'avantages des techniques de réduction de modèles.Initialement, des méthodes comme POD et LATIN ont été adoptées avec des degrés de succès divers. La méthode de TPWL, qui combine la POD avec des approximations linéaires multiples, a été prouvée comme étant la méthode de réduction de modèles la mieux adaptée pour le modèle dynamique oscillant.Pour les lignes de transmission, un modèle de paramètres distribués en domaine fréquentiel est utilisé. Des modèles réduits de type PGD sont proposés pour le modèle DP des lignes de transmission. Un problème multidimensionnel entièrement paramétrique a été formulé, avec les paramètres électriques des lignes de transmission inclus comme coordonnées additionnelles de la représentation séparée. La méthode a été étendue pour étudier la solution du modèle des lignes de transmission pour laquelle les paramètres dépendent de la fréquence. / This thesis presents the study of the model order reduction for power grids and transmission networks. The specific focus has been the transient dynamics. A mathematical viewpoint has been adopted for model reduction. Power networks are huge and complex network, simulation for power grid analysis and design require large non-linearmodels to be solved. In the context of developing “SmartGrids” with the distributed generation of power, real time analysis of complex systems such as these needs fast,reliable and accurate models. In the current study we propose model order reduction methods both a-priori and aposteriori suitable for dynamic models of power grids.The model that describes the transient dynamics of the power grids is complex non-linear swing dynamics model. The non-linearity of the swing dynamics model necessitates special attention to achieve maximum benefit from the model order reduction techniques. In the current research, POD and LATIN methods were applied initially with varying degrees of success. The method of TPWL has been proved as the best-suited model reduction method for swing dynamics model ; this method combines POD with multiple linear approximations.For the transmission lines, a distributed parameters model infrequency-domain is used. PGD based reduced-order models are proposed for the DP model of transmission lines. A fully parametric problem with electrical parameters of transmission lines included as coordinates of the separated representation. The method was extended to present the solution of frequency-dependent parameters model for transmission lines.

A major concern for geotechnical engineers is the ability to predict how a soil will react to large ground motions produced by earthquakes. Of all the different types of soil, liquefiable soils present some of the greatest challenges. The ability to quantify the degradation of a soil's shear modulus as it undergoes liquefaction would help engineers design more reliably and economically. This thesis uses ground motions recorded by an array of downhole accelerometers on Port Island, Japan, during the 1995 Kobe Earthquake, to quantify the shear modulus of sand as it liquefies. It has been shown that the shear modulus of sand decreases significantly as it liquefies, apparently decreasing in proportion to the increasing excess pore water pressure ratio (Ru). When completely liquefied, the shear modulus of sand (Ru = 1.0) for a relative density of 40 to 50% is approximately 15% of the high-strain modulus of the sand in its non-liquefied state, or 1% of its initial low-strain value. Presented in this thesis is an approach to modeling the shear modulus degradation of sand as it liquefies. This approach, called the "degrading shear modulus backbone curve method" reasonably predicts the hysteretic shear stress behavior of the liquefied sand. The shear stresses and ground accelerations computed using this method reasonably matches those recorded at the Port Island Downhole Array (PIDA) site. The degrading shear modulus backbone method is recommended as a possible method for conducting ground response analyses at sites with potentially liquefiable soils.