On considère le problème d’advection-diffusion stationnaire v(∇u, ∇v)+( β•∇u, v) = (f, v) et non stationnaire d/dt (u(t), v) + v(∇u, ∇v)+( β•∇u, v) = (g(t), v), ainsi que le problème d’advection (β•∇u, v) = (f, v) sur un domaine polygonal borné du plan. Le terme de diffusion est approché par des éléments de Crouzeix Raviart et le terme de convection par une méthode upwind sur des volumes barycentriques finis avec un maillage triangulaire. Pour le problème stationnaire d’advection-diffusion, la L²-stabilité (c’est-à-dire indépendante du coefficient de diffusion v) est démontrée pour la solution du problème approché obtenue par cette méthode d’éléments finis et de volumes finis. Pour cela une condition sur la géométrie doit être satisfaite. Des exemples de maillages sont donnés. Toujours avec cette condition géométrique sur le maillage, une inégalité de stabilité (où la discrétisation en temps n’est pas couplée à une condition sur la finesse du maillage) est obtenue pour le cas non-stationnaire. La discrétisation en temps y est faite par un schéma d’Euler implicite. Une majoration de l’erreur, proportionnelle au pas en temps et à la finesse du maillage, est ensuite proposée et exprimée explicitement en fonction des données du problème. Pour le problème d’advection, une approche utilisant la théorie des graphes est utilisée pour obtenir l’existence et l’unicité de la solution, ainsi que le résultat de stabilité. Comme pour la stabilité du problème d’advection-diffusion, une condition géométrique - qui est équivalente pour les points intérieurs du maillage à celle du problème d’advection-diffusion - est nécessaire. / We consider the stationary linear convection-diffusion equation v(∇u, ∇v)+( β•∇u, v) = (f, v), the time dependent d/dt (u(t), v) + v(∇u,∇v)+( β•∇u, v)= (g(t), v) equation and the linear advection equation (β•∇u, v) = (f, v) on a two dimensional bounded polygonal domain. The diffusion term is discretized by Crouzeix-Raviart piecewise linear finite elements, and the convection term by upwind barycentric finite volumes on a triangular grid. For the stationary convection-diffusion problem, L²-stability (i.e. independent of the diffusion coefficient v) is proven for the approximate solution obtained by this combined finite-element finite-volume method. This result holds if the underlying grid satisfies a condition that is fulfilled, for example, by some structured meshes. Using again this condition on the grid, stability is shown for the time dependent convection-diffusion equation (without any link between mesh size and time step). An implicit Euler approach is used for the time discretization. It is shown that the error associated with this scheme decays linearly with the mesh size and the time step. This result holds without any link between mesh size and time step. The dependence of the corresponding error bound on the diffusion coefficient is completely explicit. For the stationary advection equation, an approach using graph theory is used to obtain existence, uniqueness and stability. As in the stationary linear convection-diffusion equation, the underlying grid must satisfy some geometric condition. / Gegenstand der Arbeit ist die zweidimensionale stationäre Konvektion-Diffusionsgleichung v(∇u, ∇v)+( β•∇u, v) = (f, v), die zeitabhängige Konvektion-Diffusionsgleichung d/dt (u(t), v) + v(∇u,∇v)+( β•∇u, v)= (g(t), v), sowie die Konvektionsgleichung (β•∇u, v) = (f, v). Der Diffusionsterm ist diskretisiert mittels Crouzeix-Raviart stückweise lineare Finite Elemente. Das Gebiet ist in Dreiecke unterteilt und der Konvektionsterm ist mittels einer upwind Methode auf Baryzentrische Finite Volumenelemente definiert. Für die stationäre Konvektion-Diffusionsgleichung, wird (d.h. von v unabhängige) L²-Stabilität der numerischen Lösung bewiesen. Voraussetzung dafür, ist die Erfüllung gewisser geometrischer Bedingungen an die Unterteilung des Gebiets. Beispiele von Unterteilungen die diese Bedingungen erfüllen, werden gegeben. Wieder an dieser geometrischen Bedingung geknüpft, wird Stabilität (d.h. die Zeitdiskretisierung ist entkoppelt von der Netzweite) für die zeitabhängige Konvektion-Diffusionsgleichung, bewiesen. Für die Zeitableitung wird dabei eine Implizite Euler Diskretisierung verwendet. Eine obere Schranke für den Diskretisierungsfehler, proportional zum Zeitdiskretisierungsparameter und zur Netzfeinheit, ausgedrückt als Funktion der Daten der Differenzialgleichung, wird gezeigt. Für die Konvektionsgleichung wird ein graphentheoretischer Zugang verwendet, der es ermöglicht Existenz, Eindeutigkeit und Stabilität, zu bekommen. Für die Stabilität, werden ähnliche geometrische Bedingungen an die Unterteilung des Gebiets gestellt, wie beim stationären Konvektion-Diffusionsproblem.

Residual-distribution methods offer several potential benefits over classical methods, such as a means of applying upwinding in a multi-dimensional manner and a multi-dimensional positivity property. While it is apparent that residual-distribution methods also offer higher accuracy than finite-volume methods on similar meshes, few studies have directly compared the performance of the two approaches in a systematic and quantitative manner. In this study, comparisons between residual distribution and finite volume are made for steady-state smooth and discontinuous flows of gas dynamics, governed by hyperbolic conservation laws, to illustrate the strengths and deficiencies of the residual-distribution method. Deficiencies which reduce the accuracy are analyzed and a new nonlinear scheme is proposed that closely reproduces or surpasses the accuracy of the best linear residual-distribution scheme. The accuracy is further improved by extending the scheme to fourth order using established finite-element techniques. Finally, the compact stencil, arithmetic workload, and data parallelism of the fourth-order residual-distribution scheme are exploited to accelerate parallel computations on an architecture consisting of both CPU cores and a graphics processing unit. Numerical experiments are used to assess the gains to efficiency and possible monetary savings that may be provided by accelerated architectures.

Optimal control problems (OCPs) governed by convection dominated diffusion-convection-reaction equations arise in many science and engineering applications such as shape optimization of the technological devices, identification of parameters in environmental processes and flow control problems. A characteristic feature of convection dominated optimization problems is the presence of sharp layers. In this case, the Galerkin finite element method performs poorly and leads to oscillatory solutions. Hence, these problems require stabilization techniques to resolve boundary and interior layers accurately. The Streamline Upwind Petrov-Galerkin (SUPG) method is one of the most popular stabilization technique for solving convection dominated OCPs. The focus of this thesis is the application and analysis of the SUPG method for distributed and boundary OCPs governed by evolutionary diffusion-convection-reaction equations. There are two approaches for solving these problems: optimize-then-discretize and discretize-then-optimize. For the optimize-then-discretize method, the time-dependent OCPs is transformed to a biharmonic equation, where space and time are treated equally. The resulting optimality system is solved by the finite element package COMSOL. For the discretize-then-optimize approach, we have used the so called allv at-once method, where the fully discrete optimality system is solved as a saddle point problem at once for all time steps. A priori error bounds are derived for the state, adjoint, and controls by applying linear finite element discretization with SUPG method in space and using backward Euler, Crank- Nicolson and semi-implicit methods in time. The stabilization parameter is chosen for the convection dominated problem so that the error bounds are balanced to obtain L2 error estimates. Numerical examples with and without control constraints for distributed and boundary control problems confirm the effectiveness of both approaches and confirm a priori error estimates for the discretize-then-optimize approach.

Residual-distribution methods offer several potential benefits over classical methods, such as a means of applying upwinding in a multi-dimensional manner and a multi-dimensional positivity property. While it is apparent that residual-distribution methods also offer higher accuracy than finite-volume methods on similar meshes, few studies have directly compared the performance of the two approaches in a systematic and quantitative manner. In this study, comparisons between residual distribution and finite volume are made for steady-state smooth and discontinuous flows of gas dynamics, governed by hyperbolic conservation laws, to illustrate the strengths and deficiencies of the residual-distribution method. Deficiencies which reduce the accuracy are analyzed and a new nonlinear scheme is proposed that closely reproduces or surpasses the accuracy of the best linear residual-distribution scheme. The accuracy is further improved by extending the scheme to fourth order using established finite-element techniques. Finally, the compact stencil, arithmetic workload, and data parallelism of the fourth-order residual-distribution scheme are exploited to accelerate parallel computations on an architecture consisting of both CPU cores and a graphics processing unit. Numerical experiments are used to assess the gains to efficiency and possible monetary savings that may be provided by accelerated architectures.

Wind turbines installed in the open atmosphere experience much more complex and highly-varying flow than their counterparts in wind tunnels or numerical simulations. In particular, aerodynamic stall—which occurs often on stall-regulated wind turbines in such variable flow conditions—can affect both wind turbine blade lifespan and noise generation. A field test site was therefore installed at the outer limits of the city of Waterloo, Ontario to study a small-scale 30 kW stall-regulated wind turbine. Experimental equipment was installed to monitor parameters such as wind speed and direction, electrical power output, blade pitch angle, rotor rotational speed, and wind turbine yaw orientation. Extensive hardware and software was developed and installed to wirelessly collect data from all instrumentation. Tufts and a remote-operated camera were also installed on one of the two blades of the 10 m diameter horizontal-axis turbine. In a variation on the tuft flow visualisation technique, video files were analysed using a novel digital image processing code. The code was developed in MATLAB to calculate the fraction of the blade which was stalled by determining the position and angle of each tuft in every video frame. The algorithm was able to locate on average 85% of the visible tufts and correctly tagged those which were stalled with a bias of only −5% compared to the typical manual method. When the algorithm was applied to 7 h of tuft video at the outboard 40% of the blade, the total average fraction of stalled tufts varied from 5% at 5 m/s to 40% at 21 m/s. This trend was expected for the stall-regulated design since, as the wind speed is increased, the stall progresses from inboard to outboard regions and from trailing edge to leading edge. The 7 h time period represents at least a two order-of-magnitude increase compared with time periods analysed using previous manual methods. This work has demonstrated a digital implementation of tuft flow visualisation which lends statistical validity (through long-time-period averaging) to a common tool for researching wind turbine stall. The speed and ease with which the tuft method can be implemented, combined with the high cost per energy of small-scale wind turbines, suggest that this digital algorithm is a highly beneficial tool for future studies.

No description available.

This Ph.D. thesis aims to develop numerical models for two-dimensional and three-dimensional shallow water systems over mobile beds. In order to accomplish the goal of this dissertation, the following sub-projects are defined and completed. 1: The first sub-project consists in developing a robust two-dimensional coupled numerical model based on an unstructured mesh, which can simulate rapidly varying flows over an erodible bed involving wet–dry fronts that is a complex yet practically important problem. In this task, the central-upwind scheme is extended to simulation of bed erosion and sediment transport, a modified shallow water system is adopted to improve the model, a wetting and drying scheme is proposed for tracking wet-dry interfaces and stably predict the bed erosion near wet-dry area. The shallow water, sediment transport and bed evolution equations are coupled in the governing system. The proposed model can efficiently track wetting and drying interfaces while preserving stability in simulating the bed erosion near the wet-dry fronts. The additional terms in shallow water equations can improve the accuracy of the simulation when intense sediment-exchange exists; the central-upwind method adopted in the current study shows great accuracy and efficiency compared with other popular solvers; the developed model is robust, efficient and accurate in dealing with various challenging cases. 2: The second sub-project consists in developing a novel numerical scheme for a coupled two-dimensional hyperbolic system consisting of the shallow water equations with friction terms coupled with the equations modeling the sediment transport and bed evolution. The resulting 5*5 hyperbolic system of balance laws is numerically solved using a Godunov-type central-upwind scheme on a triangular grid. A spatially second-order and temporally third-order central-upwind scheme has been derived to discretize the conservative hyperbolic sub-system. However, such schemes need a correct evaluation of local wave speeds to avoid instabilities. To address such an issue, a mathematical result by the Lagrange theorem is used in the proposed scheme. Consequently, a computationally expensive process of finding all of the eigenvalues of the Jacobian matrices is avoided: The upper/lower bounds on the largest/smallest local speeds of propagation are estimated using the Lagrange theorem. In addition, a special discretization of the bed-slope term is proposed to guarantee the well-balanced property of the designed scheme. 3: The third sub-project consists in designing a novel scheme to estimate bed-load fluxes which can produce more accurate results than the previously reported coupled model. Using a pair of local wave speeds different from those used for the flow, a novel wave estimator in conjunction with the central upwind method is proposed and successfully applied to the coupled water-sediment system involving a rapid bed-erosion process. It was demonstrated that, in comparison with the decoupled model, applying the proposed novel scheme to approximate the bed-load fluxes can successfully avoid the numerical oscillations caused by simple and less stable schemes, e.g. simple upwind methods; in comparison with the coupled model using same flux-estimator for both hydrodynamic and morphological systems, the proposed numerical scheme successfully prevents excessive numerical diffusion for prediction of bed evolution. Consequently, the proposed scheme has advantages in terms of accuracy which are shown in several numerical tests. In addition, analytical expressions have been provided for calculating the eigenvalues of the coupled shallow-water-Exner system, which greatly enhances the efficiency of the proposed method. 4: The fourth sub-project consists in developing a three-dimensional numerical model for the simulation of unsteady non-hydrostatic shallow water flows on unstructured grids using the finite volume method. The free surface variations are modeled by a characteristics-based scheme which simulates sub- and super-critical flows. Three-dimensional velocity components are considered in a collocated arrangement with a sigma coordinate system. A special treatment of the pressure term is developed to avoid the water surface oscillations. Convective and diffusive terms are approximated explicitly, and an implicit discretization is used for the pressure term. The unstructured grid in the horizontal direction and the sigma coordinate in the vertical direction facilitate the use of the model in complicated geometries. 5: The fifth sub-project consists in developing a well-balanced three-dimensional shallow water model which is able to simulate shock waves over dry bed. Due to the hydrostatic simplification of the vertical momentum equation, the governing system of equations is not hyperbolic and can not be solved using standard hyperbolic solvers. That is, one can not use a high-order Godunov-type scheme to compute all fluxes through cell-interfaces. This may cause the model to fail in simulations of some unsteady-flows with discontinuities, e.g., dam-break flows and floods. To overcome this difficulty, a novel numerical scheme for the three-dimensional shallow water equations is proposed using a relaxation approach in order to convert the system to a hyperbolic one. Thus, a high-order Godunov-type central-upwind scheme based on the finite volume method can be applied to approximate the numerical fluxes. The proposed model can also preserve the ``lake at rest'' state and positivity of water depth over irregular bottom topographies based on special reconstruction of the corresponding parameters. 6: The sixth sub-project consists in extending the result of the fifth sub-project to development of a three-dimensional numerical model for shallow water flows over mobile beds, which is able to simulate morphological evolutions under shock waves, e.g. dam-break flows. The hydrodynamic model solves the three-dimensional shallow water equations using a finite volume method on prismatic cells in sigma coordinates based on the scheme prposed in sub-project 5. The morphodynamic model solves an Exner equation consisting of bed-load sediment transportation. The performance of the proposed model has been demonstrated by several laboratory experiments of dam-break flows over mobile beds.

Yes / Based on the common approach, the adaptation length in sediment transport is normally estimated in the temporal independence. However, this approach might not be theoretically justified as the process of reaching of the sediment transport equilibrium stage is affected by the flow conditions in time, especially for those fast sediment moving flows, such as scour-hole developing flow. In this study, the 2D shallow water formulation together with a sediment continuity-concentration (SCC) model were applied to flow with mobile sediment boundary. A time-varying approach was proposed to determine the sediment transport adaptation length to treat the flow sediment erosion-deposition rate. The proposed computational model was based on the Finite Volume (FV) method. The Monotone Upwind Scheme of Conservative Laws (MUSCL)-Hancock scheme was used with the Harten Lax van Leer-contact (HLLC) approximate Riemann solver to discretize the FV model. In the flow applications of this paper, a highly discontinuous dam-break fast sediment transport flow was used to calibrate the proposed time-varying sediment adaptation length model. Then the calibrated model was further applied to two separate experimental sediment transport flow applications documented in literature, i.e. a highly concentrated sediment transport flow in a wide alluvial channel and a sediment aggradation flow. Good agreements with the experimental data were presented by the proposed model simulations. The tests prove that the proposed model, which was calibrated by the discontinuous dam-break bed scouring flow, also performed well to represent the rapid bed change and the steady sediment mobility conditions. / The National Natural Science Foundation of China NSFC (Grant Number 20101311246), Major State Basic Research Development Program (973 program) of China (Grant Number 2013CB036402) and Open Fund of the State Key Laboratory of Hydraulics and Mountain River Engineering, Sichuan University of China (Grant Number SKLH-OF-1103).

Neste trabalho são apresentados os resultados do desenvolvimento e teste de esquemas upwind de alta resolução para o controle da difusão numérica em leis de conservação gerais e problemas em dinâmica dos fluidos. Em particular, são derivados dois novos esquemas: o ALUS (Adaptive Linear Upwind Scheme) e o TOPUS (Third-Order Polynomial Upwind Scheme). Esses esquemas são testados no transporte de escalares, em equações 1D tipo convecção-difusão, em sistemas hiperbólicos 1D, nas equações de Euler 2D da dinâmica dos gases e nas equações de Navier-Stokes incompressíveis 2D/3D. Os esquemas são então associados a uma modelagem algébrica não linear para a simulação de problemas de escoamentos incompressíveis turbulentos 2D com/sem superfícies livres / In this work, results of the development and testing of high-resolution upwind schemes for controlling of the numerical diffusion for general conservation laws and fluid dynamics problems are presented. In particular, two new high-resolution upwind schemes are derived, namely, the ALUS (Adaptive Linear Upwind Scheme) and the TOPUS (Third-Order Polynomial Upwind Scheme). These schemes are tested in scalar transport, 1D convection-diffusion equations, 1D hyperbolic systems, 2D Euler equations of the gas dynamics, and in 2D/3D incompressible Navier-Stokes equations. The schemes are then combined with a nonlinear Reynolds stress algebraic equation model for the simulation of 2D incompressible turbulent flows with/without free surfaces

Apesar de as EDPS que modelam leis de conservação e problemas em dinâmica dos fluídos serem bem estabelecidas, suas soluções numéricas continuam ainda desafiadoras. Em particular, há dois desafios associados à computação e ao entendimento desses problemas: um deles é a formação de descontinuidades (choques) e o outro é o fenômeno turbulência. Ambos os desafios podem ser atribuídos ao tratamento dos termos advectivos não lineares nessas equações de transporte. Dentro deste canário, esta tese apresenta o estudo do desenvolvimento de um novo esquema \"upwind\" de alta resolução e sua associação com modelagem da turbulência. O desempenho do esquema é investigado nas soluções da equação de advecção 1D com dados iniciais descontínuos e de problemas de Riemann 1D para as equações de Burgers, Euler e águas rasas. Além disso, são apresentados resultados numéricos de escoamentos incompressíveis 2D e 3D no regime laminar a altos números de Reynolds. O novo esquema é então associado à modelagem \'capa\' - \'epsilon\' da turbulência para a simulação numérica de escoamentos incompressíveis turbulentos 2D e 3D com superfícies livres móveis. Aplicação, verificação e validação dos métodos numéricos são também fornecidas / Althought the PDEs that model conservation laws and fluid dynamics problems are well established, their numerical solutions have presented a continuing challenge. In particular, there are two challenges associated with the computation and the understanding of these problems, namely, formation of shocks and turbulence. Both challenges can be attributed to the nonlinear advection terms of these transport equations. In this scenario, this thesis presents the study of the development of a new high-resolution upwind scheme and its association with turbulence modelling. The performance of the scheme is investigated by solving the 1D advection equation with discontinuous initial data 1D Riemann problems for Burgers, Euler and shallow water equations. Besides, numerical results for 2D and 3D incompressible laminar flows at high Reynolds number are presented. The new scheme is then associated with the \'capa - \' epsilon\' turbulence model for the simulation of 2D and 3D incompressible turbulent flows with moving free surfaces. Application, verification and validation of the numerical methods are also provided