Η διακριτοποίηση των μερικών διαφορικών εξισώσεων (ΜΔΕ) αποτελεί κεντρικό βήμα στην αριθμητική τους επίλυση, και ως εκ τούτου είναι ένα από τα βασικά θέματα στα σύγχρονα μαθηματικά. Η μετάβαση από τη συνεχή ΜΔΕ στο αντίστοιχο διακριτό σύστημα μπορεί να γίνει με διάφορες αριθμητικές μεθόδους, ωστόσο δεν είναι όλες οι μέθοδοι εξίσου κατάλληλες και οφείλουμε πάντα να αναζητήσουμε την αρμόζουσα διακριτοποίηση για το εκάστοτε πρόβλημα. Στο 1ο κεφάλαιο γίνεται φανερό, μέσω του απλού παραδείγματος της λογιστικής εξίσωσης, πως μια αφελής διακριτοποίηση δύναται να αλλάξει δραματικά τη φύση του προβλήματος και των λύσεών του. Ιδιαίτερη προσοχή απαιτεί η διατήρηση (πριν και μετά τη διακριτοποίηση) των συμμετριών και των αναλλοίωτων μεγεθών του προβλήματος. Στην παρούσα διπλωματική εργασία μελετάμε την περίπτωση της εξίσωσης sine-Gordon, εστιάζοντας στις σολιτονικές της λύσεις. Στο 2ο κεφάλαιο παρουσιάζεται αναλυτικά η εξίσωση αυτή. Στο 3ο κεφάλαιο μέσω δύο διαφορετικών μεθόδων διακριτοποίησης, δείχνουμε τί ακριβώς πρέπει να προσέξει κανείς έτσι ώστε να δέχεται και το διακριτό σύστημα σολιτονικές λύσεις. Ως γνωστόν οι σολιτονικές λύσεις οφείλουν να πληρούν την ιδιότητα να παραμένουν αναλλοίωτες, διατηρώντας την ταχύτητα και το πλάτος τους πριν και μετά την αλληλεπίδρασή τους. Στο 4ο κεφάλαιο παρουσιάζονται συνοπτικά τα συμπεράσματα της παρούσας εργασίας ενώ συγκρίνουμε και τις δύο μεθόδους αριθμητικής επίλυσης που αναφέραμε. / The discretization of partial differential equations (PDEs) is a key step in their numerical solution, and therefore is one of the main issues in modern mathematics. The transition from continuous PDEs to their discrete counterparts can be done by various numerical methods, though not all methods are equally suitable; for this reason one should be careful to use an appropriate discretization method for each specific problem. In the first chapter it becomes clear, through the simple example of the logistic equation, that a naive discretization may dramatically change the nature of the problem and its solutions. Particular attention needs to be paid to the preservation (before and after the discretization) of the symmetries and invariant quantities of the problem. In the present work we study the case of the famous sine-Gordon equation, focusing on its soliton solutions. The second chapter presents a step-by-step derivation of the aforementioned equation. In the third chapter we show, by means of two different discretization schemes, which conditions must be met in order to guarantee that also the discrete system will admit soliton solutions. As is well known, soliton solutions are required to remain unchanged when they interact with each other, maintaining their speed and amplitude before and after the interaction. In the fourth chapter we summarize the conclusions of this work and draw a comparison between the two numerical schemes we have studied.

The goal of this Thesis is to give a presentation of some key issues regarding the non-perturbative renormalization of the periodic scalar field theories. As an example of the non-perturbative methods, we use the differential renormalization group approach, particularly the Wegner-Houghton and the Polchinski renormalization group equations, in order to investigate the renormalization of a one-component periodic scalar field theory. The Wegner-Houghton equation provides a resummation of the loop-expansion, and the Polchinski equation is based on the resummation of the perturbation series. Therefore, these equations are exact in the sense that they contain all quantum corrections. In the framework of these renormalization group equations, field theories with periodic self interaction can be considered without violating the essential symmetry of the model: the periodicity. Both methods - the Wegner-Houghton and the Polchinski approaches - are inspired by Wilson's blocking construction in momentum space: the Wegner-Houghton method uses a sharp momentum cut-off and thus cannot be applied directly to non-constant fields (contradicts with the "derivative expansion"); the Polchinski method is based on a smooth cut-off and thus gives rise naturally to a "derivative expansion" for varying fields. However, the shape of the cut-off function (the "scheme") is not fixed a priori within Polchinski's ansatz. In this thesis, we compare the Wegner--Houghton and the Polchinski equation; we demonstrate the consistency of both methods for near-constant fields in the linearized level and obtain constraints on the regulator function that enters into Polchinski's equation. Analytic and numerical results are presented which illustrate the renormalization group flow for both methods. We also briefly discuss the relation of the momentum-space methods to real-space renormalization group approaches. For the two-dimensional Coulomb gas (which is investigated by a real-space renormalization group method using the dilute-gas approximation), we provide a systematic method for obtaining higher-order corrections to the dilute gas result.

Investigamos a existência de uma fase de Kosterlitz-Thouless no estado fundamentai de um sistema de rotores quânticos em uma dimensão. Provamos que o modelo não exibe uma transição de primeira ordem, já que a estimativa de McBryan-Spencer é válida para os rotores. Obtemos a função de partição do modelo nas representações de Lie-Trotter, de sine-Gordon, e na representação das cargas. Nessa última, provamos um limite inferior para a função de correlação entre cargas externas. Ainda na representação das cargas, damos uma nova prova do decaimento polinomial do limite superior para a função de correlação de um gás de dipolos (caroço duro) na presença de cargas externas. / We investigate the existence of a Kosterlitz-Thoules phase in the ground state of a one-dimensional array of quantum rotators. We prove that this model does not exhibit a first-order phase transition since a McBryan-Spencer bound holds for it. We obtain the partition function of the rotators in the Lie-Trotter, sine-Gordon, and charge representations. In this latter representation, we give a new proof of the upper bound polynomial decay on the external charges correlation function of a dipole gas with hardcore.

Submitted by Simone Souza (simonecgsouza@hotmail.com) on 2018-04-20T15:11:04Z No. of bitstreams: 1 DISS_2014_Leonides da Rocha Mota.pdf: 1237393 bytes, checksum: 143ba9211bb27429e5a7a12c52f58839 (MD5) / Approved for entry into archive by Jordan (jordanbiblio@gmail.com) on 2018-05-14T16:48:32Z (GMT) No. of bitstreams: 1 DISS_2014_Leonides da Rocha Mota.pdf: 1237393 bytes, checksum: 143ba9211bb27429e5a7a12c52f58839 (MD5) / Made available in DSpace on 2018-05-14T16:48:32Z (GMT). No. of bitstreams: 1 DISS_2014_Leonides da Rocha Mota.pdf: 1237393 bytes, checksum: 143ba9211bb27429e5a7a12c52f58839 (MD5) Previous issue date: 2014-02-07 / CAPES / Neste trabalho, examinamos a integrabilidade do modelo generalizado de sine-Gordon (GsG), no contexto do teste de Painlevé para equações diferenciais parciais (EDPs). Mostramos que o modelo (GsG) possui certos submodelos como o modelo duplo sine-Gordon (DsG), Bukhvostov-Lipatov (BL) e os modelos integráveis sine-Gordon. O modelo BL possui algumas direções integráveis no espaço dos campos. Classi camos as massas das soluções tipo sólitons (kinks) do modelo (GsG) através dos pesos máximos da álgebra de Lie sl(3), e mostramos que essas massas pertencem a determinados multipletos neste esquema de representação. Abordamos o modelo integrável NLS defocusing e estudamos a colisão de dois sólitons dark, em particular estudamos a mudança de fase após a sua colisão. / In this work the integrability of the generalized sine-Gordon model (GsG) is examined in the context of the Painlevé test for partial di erential equations (PDEs). We show that the (GsG) model possesses certain submodels such as the double sine-Gordon (DsG), Bukhvostov-Lipatov (BL) and the integrable sine-Gordon models. The BL model possesses some integrable directions in the eld space. Moreover, we classify the kink type solutions of the (GsG) model through the highest weight representations of the underlying sl(3) Lie algebra, and we show that these masses belong to certain multiplets in that representation scheme. We discussed the integrable defocusing NLS model and study the collision of two dark solitons, in particular we study the phase shift after their collision.

Investigamos a existência de uma fase de Kosterlitz-Thouless no estado fundamentai de um sistema de rotores quânticos em uma dimensão. Provamos que o modelo não exibe uma transição de primeira ordem, já que a estimativa de McBryan-Spencer é válida para os rotores. Obtemos a função de partição do modelo nas representações de Lie-Trotter, de sine-Gordon, e na representação das cargas. Nessa última, provamos um limite inferior para a função de correlação entre cargas externas. Ainda na representação das cargas, damos uma nova prova do decaimento polinomial do limite superior para a função de correlação de um gás de dipolos (caroço duro) na presença de cargas externas. / We investigate the existence of a Kosterlitz-Thoules phase in the ground state of a one-dimensional array of quantum rotators. We prove that this model does not exhibit a first-order phase transition since a McBryan-Spencer bound holds for it. We obtain the partition function of the rotators in the Lie-Trotter, sine-Gordon, and charge representations. In this latter representation, we give a new proof of the upper bound polynomial decay on the external charges correlation function of a dipole gas with hardcore.

The goal of this Thesis is to give a presentation of some key issues regarding the non-perturbative renormalization of the periodic scalar field theories. As an example of the non-perturbative methods, we use the differential renormalization group approach, particularly the Wegner-Houghton and the Polchinski renormalization group equations, in order to investigate the renormalization of a one-component periodic scalar field theory. The Wegner-Houghton equation provides a resummation of the loop-expansion, and the Polchinski equation is based on the resummation of the perturbation series. Therefore, these equations are exact in the sense that they contain all quantum corrections. In the framework of these renormalization group equations, field theories with periodic self interaction can be considered without violating the essential symmetry of the model: the periodicity. Both methods - the Wegner-Houghton and the Polchinski approaches - are inspired by Wilson's blocking construction in momentum space: the Wegner-Houghton method uses a sharp momentum cut-off and thus cannot be applied directly to non-constant fields (contradicts with the "derivative expansion"); the Polchinski method is based on a smooth cut-off and thus gives rise naturally to a "derivative expansion" for varying fields. However, the shape of the cut-off function (the "scheme") is not fixed a priori within Polchinski's ansatz. In this thesis, we compare the Wegner--Houghton and the Polchinski equation; we demonstrate the consistency of both methods for near-constant fields in the linearized level and obtain constraints on the regulator function that enters into Polchinski's equation. Analytic and numerical results are presented which illustrate the renormalization group flow for both methods. We also briefly discuss the relation of the momentum-space methods to real-space renormalization group approaches. For the two-dimensional Coulomb gas (which is investigated by a real-space renormalization group method using the dilute-gas approximation), we provide a systematic method for obtaining higher-order corrections to the dilute gas result.

No description available.

Three different aspects of the sinh-Gordon model are explored in this thesis. We begin, in chapter one, with a summary of the model and the necessary background. Chapter two studies the model with two boundary conditions. Two approaches are presented to investigate the reflection factors off the boundaries and the energy of the theory. In chapter three, perturbation theory is developed to study the theory with one general boundary condition. A contribution to the quantum reflection factor is obtained and compared with the result obtained for the special boundary condition. Chapters four and five investigate the supersymmetric extension of the model in the presence of a single boundary. Firstly, the classical limits of the supersymmetric reflection matrices are checked. The exact reflection factors are studied perturbatively up to the second order of the coupling constant. Secondly, the perturbation theory and the path integral formalism are employed in the supersymmetric model to study the quantum reflection factors. We conclude with a brief sixth chapter describing the outlook for further investigations.

In this thesis, we consider the massive field theories in 1+1 dimensions known as affine Toda quantum field theories. These have the special property that they possess an infinite number of conserved quantities, a feature which greatly simplifies their study, and makes extracting exact information about them a tractable problem. We consider these theories both in the full space (the bulk) and in the half space bounded by an impenetrable boundary at x = 0. In particular, we consider their fundamental objects: the scattering matrices in the bulk, and the reflection factors at the boundary, both of which can be found in a closed form. In Chapter 1, we provide a general introduction to the topic before going on, in Chapter 2, to consider the simplest ATFT—the sine-Gordon model—with a boundary. We begin by studying the classical limit, finding quite a clear picture of the boundary structure we can expect in the quantum case, which is introduced in Chapter 3. We obtain the bound-state structure for all integrable boundary conditions, as well as the corresponding reflection factors. This structure turns out to be much richer than had hitherto been imagined. We then consider more general ATFTs in the bulk. The sine-Gordon model is based on a(^(1))(_1), but there is an ATFT for any semi-simple Lie algebra. This underlying structure is known to show up in their S-matrices, but the path back to the parameters in the Lagrangian is still unclear. We investigate this, our main result being the discovery of a "generalised bootstrap" equation which explicitly encodes the Lie algebra into the S-matrix. This leads to a number of new S-matrix identities, as well as a generalisation of the idea that the conserved charges of the theory form an eigenvector of the Cartan matrix. Finally our results are summarised in Chapter 5, and possible directions for further study are highlighted.

Submitted by Izabel Franco (izabel-franco@ufscar.br) on 2016-09-23T14:20:00Z No. of bitstreams: 1 TeseMC.pdf: 7974204 bytes, checksum: 37e1ce261d2f4d80353e99c7ad41eada (MD5) / Approved for entry into archive by Marina Freitas (marinapf@ufscar.br) on 2016-09-26T20:39:24Z (GMT) No. of bitstreams: 1 TeseMC.pdf: 7974204 bytes, checksum: 37e1ce261d2f4d80353e99c7ad41eada (MD5) / Approved for entry into archive by Marina Freitas (marinapf@ufscar.br) on 2016-09-26T20:39:29Z (GMT) No. of bitstreams: 1 TeseMC.pdf: 7974204 bytes, checksum: 37e1ce261d2f4d80353e99c7ad41eada (MD5) / Made available in DSpace on 2016-09-26T20:39:36Z (GMT). No. of bitstreams: 1 TeseMC.pdf: 7974204 bytes, checksum: 37e1ce261d2f4d80353e99c7ad41eada (MD5) Previous issue date: 2015-04-30 / Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) / In order to obtain structures known as defects, it was used a systematic procedure which holds cyclic deformation chains. This cyclical procedure enables that the initial defect (used to trigger the chain) is recovered via the process of successive deformations. This technique was applied considering topological kink like defects derived from two models, 4 and sine-Gordon, described by a single real scalar eld. The results show that this procedure can generate simultaneously kink and lump like defects with topological mass satisfying closed relations. After the detailed description and analysis of this method, some of its results were applied in brane scenario, where we studied the quantum problem analogue derived from a metric perturbation. The scenarios includes thick branes results that could support 4-dimensional gravity inside. Finally, we studied the topological origin of vacuum transitions in scenarios supported by double-well potentials. It was found that the Wigner function, constructed by means of the ground and rst excited states (solutions of the normal modes potential spectrum), performs quantum tunneling moving from one minimum to another in the potential. The tunneling analysis was performed by a prescription of the Wigner's function dynamics and the time dependence of stagnation points for an analytical double well potential. / Com a finalidade de obter estruturas conhecidas como defeitos, foi utilizado um procedimento sistemático que encerra cadeias cíclicas de deformações. Esse procedimento ciclico possibilita que o defeito inicial (utilizado para acionar a cadeia) seja recuperado através do processo de deformações sucessivas. Essa técnica foi aplicada considerando-se defeitos topológicos (tipo kink) derivados de dois modelos, __ x4 e sine-Gordon, descritos por um único campo escalar e real. Os resultados encontrados revelam que esse procedimento pode gerar simultaneamente defeitos tipo kink e tipo lump (não topológico) com massas topológicas satisfazendo relações fechadas de vinculo. Após a descriçao e análise detalhadas desse método, alguns de seus resultados foram aplicados em cenário de branas, tencionando-se estudar seu problema quântico derivado de uma perturbação na métrica. O cenário inclui como resultado branas espessas que poderiam sustentar gravidade 4-dimensional em seu interior. Por fim, estudou-se a origem topológica das transições de vácuo em cenários sustentados por potenciais com fundo duplo. Verificou-se que a função de Wigner, construida por meio do estado fundamental e do primeiro estado excitado (soluções do espectro de modos normais do potencial), realiza tunelamento quântico deslocando-se de um minimo ao outro do potencial. A análise do tunelamento foi realizada através de uma prescrição da dinâmica da função de Wigner e da dependência temporal dos pontos de estagnação para um potencial de fundo duplo analitico.