Le sujet visé par cette dissertation est la logique ordinale de Turing. Nous nous référons au texte original de Turing «Systems of logic based on ordinals» (Turing [1939]), la thèse que Turing rédigea à Princeton sous la direction du professeur Alonzo Church. Le principe d’une logique ordinale consiste à surmonter localement l’incomplétude gödelienne pour l’arithmétique par le biais de progressions d’axiomes récursivement consistantes. Étant donné son importance considérable pour la théorie de la calculabilité et les fondements des mathématiques, cette recherche méconnue de Turing mérite une attention particulière. Nous retraçons ici le projet d’une logique ordinale, de ses origines dans le théorème d’incomplétude de Gödel jusqu'à ses avancées dans les développements de la théorie de la calculabilité. Nous concluons par une discussion philosophique sur les fondements des mathématiques en fonction d’un point de vue finitiste. / The main subject of this dissertation is Turing’s ordinal logic, i.e. Turing’s attempt to locally overcome Gödel’s incompleteness by means of transfinite recursive progressions. We shall refer to the original 1939 text «Systems of logic based on ordinals» which is, in fact, Turing’s Ph.D thesis at Princeton University under the direction of Professor Alonzo Church. Considering its importance for the theory of computability and the foundations of mathematics, Turing’s paper certainly didn’t get enough attention in the literature. Therefore, we want to retrace Turing’s project of an ordinal logic from its very foundation in Gödel’s incompleteness theorem to its further development in calculability theory. A discussion on the foundations of mathematics from a computational point of view will conclude this memoir.

Quantum spin models play an important role in theoretical condensed matter physics and quantum information theory. One numerical technique that is frequently used in studies of quantum spin systems is exact diagonalization. In this approach, numerical methods are used to find the lowest eigenvalues and associated eigenvectors of the Hamilton matrix of the quantum system. The computational problem is thus to determine the lowest eigenpairs of an extremely large, sparse matrix. Although many sophisticated iterative techniques for the determination of a small number of lowest eigenpairs can be found in the literature, most exact diagonalization studies of quantum spin systems have employed the Lanczos algorithm. In contrast to this, other methods have been applied very successfully to the similar problem of electronic structure calculations. The well known VASP code for example uses a Block Davidson method as well as the residual-minimization - direct inversion of the iterative subspace algorithm (RMM-DIIS). The Davidson algorithm is closely related to the Lanczos method but usually needs less iterations. The RMM-DIIS method was originally proposed by Pulay and later modified by Wood and Zunger. The RMM-DIIS method is particularly interesting if more than one eigenpair is sought since it does not require orthogonalization of the trial vectors at each step. In this work I study the efficiency of the Lanczos, Block Davidson and RMM-DIIS method when applied to basic quantum spin models like the spin-1/2 Heisenberg chain, ladder and dimerized ladder. I have implemented all three methods and are currently applying the methods to the different models. In our presentation I will compare the three algorithms based on the number of iterations to achieve convergence, the required computational time. An Intel's Many-Integrated Core architecture with Intel Xeon Phi coprocessor 5110P integrates 60 cores with 4 hardware threads per core was used for RMM-DIIS method, the achieved parallel speedups were compared with those obtained on a conventional multi-core system.

CoordenaÃÃo de AperfeiÃoamento de Pessoal de NÃvel Superior / This paper deals with the recognition of TAPER using the method of matrix diagonalization 2X2. At first, it shows the definition of conics, standard equations followed by their names and geometric representations. Then follows the ideas of eigenvalues and eigenvectors of a linear transformation that are the basis for the diagonalization of matrices.Immediately after that, the linear independence eigenvector is discussed, as well as its properties of forming a basis of a vector space. The condition for any square matrix to be diagonalizable is shown below, as well as the particulars of a symmetric matrix. The demonstration that all 2Ã2 symmetric matrix is diagonalizable is made from a matrix, elegant and elemental approach. The recognition of conics is made from basic calculations using some content widely exploited in high school such as matrices, determinants, linear systems and algebraic equations. At the end it is presented a way of teaching conical school using educational software Winplot. / Este trabalho trata do reconhecimento de CÃNICAS utilizando o mÃtodo de diagonalizaÃÃo de matrizes 2X2. No inÃcio à apresentada a definiÃÃo de cÃnicas, as equaÃÃes padrÃes seguidas de seus respectivos nomes e representaÃÃes geomÃtricas. Seguem-se entÃo as ideias de autovalores e autovetores de uma transformaÃÃo linear que servem de base para a diagonalizaÃÃo de matrizes. Logo apÃs sÃo discutidas a independÃncia linear de autovetores bem como suas propriedades de formarem uma base de um espaÃo vetorial. A condiÃÃo para que toda matriz quadrada seja diagonalizÃvel à apresentada em seguida, bem como as particularidades de uma matriz simÃtrica. A demonstraÃÃo de que toda matriz simÃtrica 2X2 à diagonalizÃvel à feita a partir de uma abordagem matricial, elegante e elementar. O reconhecimento de cÃnicas à feito a partir de cÃlculos bÃsicos, utilizando alguns conteÃdos amplamente explorados no Ensino MÃdio tais como: matrizes, determinantes, sistemas lineares e equaÃÃes algÃbricas. No final à apresentada uma forma de ensinar cÃnicas na escola utilizando o software educacional Winplot.

CoordenaÃÃo de AperfeiÃoamento de Pessoal de NÃvel Superior / Este trabalho trata do reconhecimento de quÃdricas utilizando o mÃtodo de diagonalizaÃÃo de matrizes 3 x 3. No inÃcio à apresentada a definiÃÃo de quÃdricas, as equaÃÃes padrÃes seguidas de seus respectivos nomes e representaÃÃes geomÃtricas. Seguem-se entÃo as ideias de autovalores e autovetores de uma transformaÃÃo linear que servem de base para a diagonalizaÃÃo de matrizes. Logo apÃs sÃo discutidas a independÃncia linear de autovetores bem como suas propriedades de formarem uma base de um espaÃo vetorial. A condiÃÃo para que toda matriz quadrada seja diagonalizÃvel à apresentada em seguida, bem como as particularidades de uma matriz simÃtrica. A demonstraÃÃo de que toda matriz simÃtrica à diagonalizÃvel à feita a partir de uma abordagem matricial, elegante e elementar. O reconhecimento de quÃdricas à feito a partir de cÃlculos bÃsicos, utilizando alguns conteÃdos amplamente explorados no Ensino MÃdio tais como: matrizes, determinantes, sistemas lineares e equaÃÃes algÃbricas. No final à apresentada uma forma de ensinar quÃdricas na escola utilizando o software educacional Winplot. / This paper deals with the recognition of quadrics using the method of diagonalization of matrices 3 à 3. Earlier it shows the definition of quadrics, the standard equations followed by their names and geometric representations. Then follows the ideas of eigenvalues and eigenvectors of a linear transformation that are the basis for the diagonalization of matrices. Immediately after the linear independence of the eigenvectors is discussed as well as their properties of forming a basis of a vector space. The condition for any square matrix be diagonalizable is shown after, as well as the particularities of a symmetric matrix. The demonstration that all 3 à 3 symmetric matrix is diagonalizable is made from an elegant and elemental matrix approach. Recognition of quadrics is made from basic calculations using some content widely exploited in high school such as matrices, determinants, linear systems and algebraic equations. At the end it presents a way of teaching quadrics in school using educational software Winplot.

This thesis, comprising three publications, explores the efficacy of novel generalization of the fidelity susceptibility and their numerical application to the study of frustrated quantum phase transitions in two and one dimensions. Specifically, they will be used in exact diagonalization studies of the various limiting cases of the anisotropic next-nearest neighbour triangular lattice Heisenberg model (ANNTLHM). These generalized susceptibilities are related to the order parameter susceptibilities and spin stiffness and are believed to exhibit similar behaviour although with greater sensitivity. This makes them ideal for numerical studies on small systems. Additionally, the utility of the excited-state fidelity and twist boundary conditions will be explored. All studies are done through numerical exact diagonalization. In the limit of interchain couplings going to zero the ANNTLHM reduces to the well studied $J_1-J_2$ chain with a known, difficult to identify, BKT-type transition. In the first publication of this work the generalized fidelity susceptibilities introduced therein are shown to be able to identify this transition as well as characterize the already understood phases it straddles. The second publication of this work then seeks to apply these generalized fidelity susceptibilities, as well as the excited-state fidelity, to the study of the general phase diagram of the ANNTLHM. It is shown that the regular and excited-state fidelities are useful quantities for the mapping of novel phase diagrams and that the generalized fidelity susceptibilities can provide valuable information as to the nature of the phases within the mapped phase regions. The final paper sees the application of twisted boundary conditions to the anisotropic triangular model (next-nearest neighbour interactions are zero). It is demonstrated that these boundary conditions greatly enhance the ability to numerically explore incommensurate physics in small systems. / Thesis / Doctor of Science (PhD)

No description available.

A rich variety of phases can exist in quantum systems. For example, the fractional quantum Hall states have persistent topological characteristics that derive from strong interaction. This thesis uses the exact diagonalization method to investigate quantum lattice models with strong interaction. Our research topics revolve around quantum phase transitions between novel phases. The goal is to find the best schemes for realizing these novel phases in experiments. We studied the fractional Chern insulator and its transition to uni-directional stripes of particles. In addition, we studied topological Mott insulators with spontaneous time-reversal symmetry breaking induced by interaction. We also studied emergent kinetics in one-dimensional lattices with spin-orbital coupling. The exact diagonalization method and its implementation for studying these systems can easily be applied to study other strongly correlated systems. / PHD / Topological quantum states are a new type of quantum state that have properties that cannot be described by local order parameters. These types of states were first discovered in the 1980s with the integer quantum Hall effect and the fractional quantum Hall effect. In the 2000s, the predicted and experimentally discovered topological insulators triggered studies of new topological quantum states. Studies of strongly correlated systems have been a parallel research topic in condensed matter physics. When combining topological systems with strong correlation, the resulting systems can have novel properties that emerge, such as fractional charge. This thesis summarizes our work that uses the exact diagonalization method to study topological states with strong interaction.

In this dissertation, we use exact diagonalization to study a few strongly correlated systems, ranging from the Fermi-Hubbard model to the fractional quantum Hall effect (FQHE). The discussion starts with an overview of strongly correlated systems and what is meant by strongly correlated. Then, we extend cluster perturbation theory (CPT), an economic method for computing the momentum and energy resolved Green's function for Hubbard models to higher order correlation functions, specifically the spin susceptibility. We benchmark our results for the one-dimensional Fermi-Hubbard model at half-filling. In addition we study the FQHE at fillings $nu = 5/2$ for fermions and $nu = 1/2$ for bosons. For the $nu = 5/2$ system we investigate a two-body model that effectively captures the three-body model that generates the Moore-Read Pfaffian state. The Moore-Read Pfaffian wave function pairs composite fermions and is believed to cause the FQHE at $nu = 5/2$. For the $nu = 1/2$ system we estimate the entropy needed to observe Laughlin correlations with cold atoms via an ansatz partition function. We find entropies achieved with conventional cooling techniques are adequate. / Doctor of Philosophy / Strongly correlated quantum many-body physics is a rich field that hosts a variety of exotic phenomena. By quantum many-body we mean physics that is concerned with the behavior of interacting particles, such as electrons, where the quantum behavior cannot be ignored. By strongly correlated, we mean when the interactions between particles are sufficiently strong such that they cannot be treated as a small perturbation. In contrast to weakly correlated systems, strongly correlated systems are much more difficult to solve. That is because methods that reduce the many-body problem to a single independent body problem do not work well. In this dissertation we use exact diagonalization, a method to computationally solve quantum many-body systems, to study two strongly correlated systems: the Hubbard model and the fractional quantum Hall effect.The Hubbard model captures the physics of many interesting materials and is the standard toy model. Originally developed with magnetic properties in mind, it has been extended to study superconductivity, topological phases, cold atoms, and much more. The fractional quantum Hall effect is a novel phase of matter that hosts exotic excitations, some of which may have applications to quantum computing.

This work presents a new implementation of the QP algorithm, a computer method to diagonalize the extremely large matrices arising in multimode vibronic problems. Benchmark calculations are included, showing the accuracy of the program. The QP algorithm is extended to treat multiple electronic surfaces for competitive control and this is demonstrated with an Hamiltonian including three electronic states, a model of the benzene radical cation. Finally, the evolution of zeroth-order states in a simple two electronic states, two dimensional model with a conical intersection is explored, towards building a time-dependent view of overlapping resonances coherent control.

This work presents a new implementation of the QP algorithm, a computer method to diagonalize the extremely large matrices arising in multimode vibronic problems. Benchmark calculations are included, showing the accuracy of the program. The QP algorithm is extended to treat multiple electronic surfaces for competitive control and this is demonstrated with an Hamiltonian including three electronic states, a model of the benzene radical cation. Finally, the evolution of zeroth-order states in a simple two electronic states, two dimensional model with a conical intersection is explored, towards building a time-dependent view of overlapping resonances coherent control.