Many problems in fluid mechanics and material sciences deal with liquid-vapour flows. In these flows, the ideal gas assumption is not accurate and the van der Waals equation of state is usually used. This equation of state is non-convex and causes the solution domain to have two hyperbolic regions separated by an elliptic region. Therefore, the governing equations of these flows have a mixed elliptic-hyperbolic nature. Numerical oscillations usually appear with standard finite-difference space discretization schemes, and they persist when the order of accuracy of the semi-discrete scheme is increased. In this study, we propose to use a Chebyshev pseudospectral method for solving the governing equations. A comparison of the results of this method with very high-order (up to tenth-order accurate) finite difference schemes is presented, which shows that the proposed method leads to a lower level of numerical oscillations than other high-order finite difference schemes, and also does not exhibit fast-traveling packages of short waves which are usually observed in high-order finite difference methods. The proposed method can thus successfully capture various complex regimes of waves and phase transitions in both elliptic and hyperbolic regimes

Many problems in fluid mechanics and material sciences deal with liquid-vapour flows. In these flows, the ideal gas assumption is not accurate and the van der Waals equation of state is usually used. This equation of state is non-convex and causes the solution domain to have two hyperbolic regions separated by an elliptic region. Therefore, the governing equations of these flows have a mixed elliptic-hyperbolic nature. Numerical oscillations usually appear with standard finite-difference space discretization schemes, and they persist when the order of accuracy of the semi-discrete scheme is increased. In this study, we propose to use a Chebyshev pseudospectral method for solving the governing equations. A comparison of the results of this method with very high-order (up to tenth-order accurate) finite difference schemes is presented, which shows that the proposed method leads to a lower level of numerical oscillations than other high-order finite difference schemes, and also does not exhibit fast-traveling packages of short waves which are usually observed in high-order finite difference methods. The proposed method can thus successfully capture various complex regimes of waves and phase transitions in both elliptic and hyperbolic regimes

Many problems in fluid mechanics and material sciences deal with liquid-vapour flows. In these flows, the ideal gas assumption is not accurate and the van der Waals equation of state is usually used. This equation of state is non-convex and causes the solution domain to have two hyperbolic regions separated by an elliptic region. Therefore, the governing equations of these flows have a mixed elliptic-hyperbolic nature. Numerical oscillations usually appear with standard finite-difference space discretization schemes, and they persist when the order of accuracy of the semi-discrete scheme is increased. In this study, we propose to use a Chebyshev pseudospectral method for solving the governing equations. A comparison of the results of this method with very high-order (up to tenth-order accurate) finite difference schemes is presented, which shows that the proposed method leads to a lower level of numerical oscillations than other high-order finite difference schemes, and also does not exhibit fast-traveling packages of short waves which are usually observed in high-order finite difference methods. The proposed method can thus successfully capture various complex regimes of waves and phase transitions in both elliptic and hyperbolic regimes

Many problems in fluid mechanics and material sciences deal with liquid-vapour flows. In these flows, the ideal gas assumption is not accurate and the van der Waals equation of state is usually used. This equation of state is non-convex and causes the solution domain to have two hyperbolic regions separated by an elliptic region. Therefore, the governing equations of these flows have a mixed elliptic-hyperbolic nature. Numerical oscillations usually appear with standard finite-difference space discretization schemes, and they persist when the order of accuracy of the semi-discrete scheme is increased. In this study, we propose to use a Chebyshev pseudospectral method for solving the governing equations. A comparison of the results of this method with very high-order (up to tenth-order accurate) finite difference schemes is presented, which shows that the proposed method leads to a lower level of numerical oscillations than other high-order finite difference schemes, and also does not exhibit fast-traveling packages of short waves which are usually observed in high-order finite difference methods. The proposed method can thus successfully capture various complex regimes of waves and phase transitions in both elliptic and hyperbolic regimes

Dans cette thèse, on considère plusieurs problèmes issus de la modélisation macroscopique des mouvements de foule. Le premier modèle consiste en une loi de conservation avec un flux discontinu, le second est un système mixte hyperbolique-elliptique et le dernier est une équation non-locale. D'abord, on utilise le modèle de Hughes une dimension pour décrire l'évacuation d'un couloir avec deux sorties. Ce modèle couple une loi de conservation avec un flux discontinu à une équation eikonale. On implémente la méthode de suivi de fronts, qui traite explicitement le comportement de la solution non-classique au point de rebroussement, afin d'obtenir des solutions de référence. Elles serviront à tester numériquement la convergence de schémas aux volumes finis classiques. Ensuite, on modélise le croisement de deux groupes marchant dans des directions opposées avec un système de lois de conservation mixte hyperbolique-elliptique dont le flux dépend des deux densités. Le système perd son hyperbolicité pour certainement valeurs de densité. On assiste à l'apparition d'oscillations persistantes mais bornées, ce qui conduit à la reformulation du problème associé dans le cadre des mesures de probabilités. Finalement, on étudie un modèle non-local de trafic piétonnier en deux dimensions. Le modèle consiste en une loi de conservation dont le flux dépend d'une convolution de la densité. Avec ce modèle, on résout un problème d'optimisation pour une évacuation d'une salle avec une méthode de descente, évaluant l'impact du calcul explicite du gradient de la fonction coût avec la méthode de l'état adjoint plutôt que son approximation par différences finies. / In this thesis, we consider nonclassical problems brought out by the macroscopic modeling of pedestrian flow. The first model consists of a conservation law with a discontinuous flux, the second is a mixed hyperbolic-elliptic system of conservation laws and the last one is a nonlocal equation. In the first chapter, we use the Hughes model in one space-dimension to represent the evacuation of a corridor with two exits. The model couples a conservation law with discontinuous flux to an eikonal equation. We implement the wave front tracking scheme, treating explicitly the solution nonclassical behavior at the turning point, to provide a reference solution, which is used to numerically test the convergence of classical finite volume schemes. In the second chapter, we model the crossing of two groups of pedestrians walking in opposite directions with a system of conservation laws whose flux depends on the two densities. This system loses its hyperbolicity for certain density values. We assist to the rising of persistent but bounded oscillations, that lead us to the recast of the problem in the framework of measure-valued solutions. Finally we study a nonlocal model of pedestrian flow in two space-dimensions. The model consists of a conservation law whose flux depends on a convolution of the density. With this model, we solve an optimization problem for a room evacuation with a descent method, evaluating the impact of the explicit computation of the cost function gradient with the adjoint state method rather than approximating it with finite differences.

Práce je zaměřená na numerickou simulaci dvoufázového proudění. Je studován matematický model a numerická aproximace toku dvou nemísitelných nestlačitelných tekutin. Rozhraní mezi tekutinami je popsáno pomocí pomocí tzv. level set metody. Představena je diskretizace problému v prostoru a v čase. Metoda konečných prvk· se zpětnou Eulerovou metodou je aplikována na Navierovy-Stokesovy rovnice a časoprostorová nespojitá Galerkinova metoda je použita k řešení transportního problému. D·raz je kladen na analýzu chyby nespojité Galerkinovy metody přímek a časoprostorové nespojité Galerkinovy metody pro transportní problém. Jsou prezentovány numerické výsledky. 1