Treballs Finals de Grau (TFG) - Física

URI permanent per a aquesta col·leccióhttps://hdl.handle.net/2445/59725

Treballs Finals del Grau de Física de la Facultat de Físca de la Universitat de Barcelona.

Estadístiques

Examinar

Mostrant 1 - 20 de 1054

Spin Pumping in NbRe-PY and Nb-Py Bilayers
(2026-01) Vicente Sánchez, Lídia; García Santiago, Antoni
Broadband ferromagnetic resonance measurements were performed on NbRe/Py and Nb/Py bilayers of nm thickness to characterize magnetic damping and interfacial spin transport. The extracted Gilbert damping values show no significant dependence on stacking order or on the choice of normal-state superconductor. A thickness-dependent analysis of the NbRe/Py bilayer reveals a clear increase in α with the non-magnetic layer thickness, consistent with spin-pumping-induced dissipation. These results provide a concise experimental picture of damping and spin transport in FM/NM systems involving superconducting materials in their normal state
Synthetic Cross-sequence Generation of MRI Images using Deep Learning Networks
(2026-01) Romero Díaz, Jacobo; Niñerola Baizán, Aida; Farré Melero, Arnau
Magnetic Resonance Imaging (MRI) is a non-invasive diagnostic imaging modality that employs multiple acquisition sequences to generate complementary tissue contrasts. However, in clinical practice, not all sequences are available due to time, cost, or patient-related issues. In this work, we investigate deep learning–based supervised image-to-image translation for synthetic cross-sequence MRI generation using a dataset of 148 patients. T2, FLAIR, and contrast-enhanced T1 (T1GD) images are synthesized from T1 inputs using U-Net–based encoder–decoder models and conditional GANs (pix2pix), under full-resolution and patch-based training strategies. Evaluation using Structural Similarity Index (SSIM) and Peak Signal-to-Noise Ratio (PSNR) shows that UNet models with tailored loss functions achieve results comparable to state-of-the-art approaches, outperforming pix2pix. Finally, qualitative evaluation revealed that conventional metrics fail to capture sequence-specific localized regions, highlighting the need for task-aware evaluation criteria.
Composite Models of Dynamical Dark Energy: ΛXCDM and ωXCDM
(2026-01) Torné Pujol, Carla; Solà Peracaula, Joan
Although the Standard ΛCDM Model works well, it suffers from serious theoretical problems (eg. the Cosmological Constant Problem and the Cosmic Coincidence Problem) as well as phenomenological problems, such as the cosmological tensions (Hubble Tension and growth tension). In this work, we analyze the ΛXCDM model, a composite Dark Energy (DE) model involving a variable cosmological term (Running Vacuum) and an additional dark energy field X, which can play the role of ”phantom matter” (different from phantom DE) [4, 8]. Through the analytical resolution of the cosmological equations and the study of the density evolution and the effective equation of state (ωe), we demonstrate that the model can alleviate the Coincidence Problem thanks to the existence of a local maximum in the density ratio r = ρD/ρm, without need for fine-tuning, and leading the Universe towards a Turning Point of expansion instead of a Big Rip singularity. Finally, we show that the model’s dynamics exhibits a type of crossing of the phantom divide which is fully compatible with the recent releases of DESI 2024 data [7, 10, 11].
Simulation of highly focused vector beams and their 3D Stokes vector generalization
(2026-01) Sáiz Martínez, David; Maluenda Niubó, David
This work presents the development of a versatile numerical code for the simulation and polarization analysis of tightly focused vector beams in high numerical aperture optical systems. The computational framework is based on the Richards-Wolf vectorial diffraction formalism, which enables the rigorous calculation of the full three-dimensional electromagnetic field in the focal region, including non-negligible longitudinal components. The code implements a modular pipeline that defines the incident field, performs its projection onto the reference sphere, evaluates the focusing integral and characterizes the local polarization state using a three-dimensional generalization of the Stokes parameters. This approach overcomes the limitations of the conventional transverse Stokes formalism and provides a complete local description of polarization in nonparaxial fields. Numerical simulations and advanced visualization tools are used to analyze complex polarization structures, demonstrating the potential of the developed framework as a flexible platform for the study of nonparaxial optical phenomena.
Generalized Heisenberg Uncertainty Principle applied to isospin symmetry breaking operators in atomic nuclei
(2026) Roger Garcia, Àlex; Roca Maza, Xavier
Based on Sandro Stringari’s article Heisenberg Uncertainty Inequality and Breaking of Isospin Symmetry in Atomic Nuclei [1], the generalized Heisenberg’s uncertainty inequality is used to find a lower bound to the breaking of isospin symmetry. To do this, we employ the lowering isospin operator and the isovector giant monopole resonance operator (IVGMR) instead of the original position and linear momentum operators to set up the new uncertainty principle. As an original development of our thesis, we modify the IVGMR operator in order to eliminate the contribution of the isobaric analogue state (IAS). We find, within the Hartree-Fock plus Random Phase Approximation method applied to a sound nuclear effective interaction (Skyrme-type), that the IAS isolates the main contribution to the generalized Heisenberg’s uncertainty principle presented here
Surface tension of fluids with long range interactions
(2026-01) Romero Meraner, Daniel José; Reguera López, David
Surface tension is a key physical property that quantifies the energy required to maintain an interface between two distinct phases. It plays a vital role in various scientifically and technologically significant processes, such as the nucleation of liquid droplets and the formation and collapse of bubbles. Theoretical studies suggest that for systems of molecules with long-range attractions, surface tension and its curvature corrections might not be well defined. In this work, a Monte Carlo simulation has been implemented to evaluate the planar surface tension of a Mie fluid with different attractive exponents. We have validated the simulation code for the particular case of the Lennard-Jones fluid, where we find values of the planar surface tension in perfect agreement with previous results from the literature at all temperatures. In the case of the Mie fluid, we find that the planar surface tension increases significantly as the range of the attractive interactions increases, while the width of the interface becomes progressively sharper. The influence of finite-size effects has also been analyzed, finding no significant effects. The study sets a reference value for planar surface tension as a starting point to analyze curvature effects, which might have important consequences in the nucleation behavior of fluids with long-range interactions
Machine Learning for Accelerated Cosmological Inference Beyond Flat ΛCDM
(2026-01) Ramirez Nethersole, Rafael Raul; Gómez Valent, Adrià
We review the efficacy of Neural Networks (NN) for emulating cosmological observables (angular acoustic scale θ∗ and binned angular power spectra Db of the CMB temperature anisotropies across ℓ ∈ [30, 400]), and apply our emulator-based inference framework to the nonflat ΛCDM model, in the range ΩK ∈ [−0.2, 0.2]. We achieve speed up actor computation times beyond 104 with respect to standard boltzamn solvers, while maintaining all NN prediction error values below 0.7% relative to current experimental uncertainties, enabling for computationally feasible parameter inference pipelines. We also provide constraints on the curvature density parameter ΩK and other ΛCDM parameters using the NN developed in this work and the temperature maps from Planck 2018 SMICA data. A marginal posterior using our NN model gives curvature density ΩK = 0.001+0.030 −0.070, consistent with a flat Universe.
Ti and Al-Ti thin films for quasiparticle trapping in superconducting quantum circuits
(2026-01) Requena Maldonado, Pol; Gómez del Pulgar, Ariadna; Forn-Díaz, Pol; Costache, Marius V.
Quasiparticles are unwanted excitations of Cooper pairs that disrupt superconductivity and undermine the performance of superconducting qubits. To mitigate this effect, engineered regions of lower superconducting gap energy that capture broken Cooper pairs, known as quasiparticle traps, can be implemented. In this work, as a first step towards the realization of a quasiparticle trapping experiment on the shunting capacitor plates of a 2D transmon-type qubit, a recipe for the deposition of titanium and aluminium-titanium thin films of different thicknesses has been developed and successfully implemented. The room-temperature conductance characteristics of the films have been assessed, and their dependence on film thickness has been ascertained.
The Galactic potential effects on the Solar System’s minor bodies
(2026-01) Roca Acarin, Pau; Romero Gómez, Mercè
This project studies the influence of the Milky Way’s galactic gravitational potential on the orbital evolution of distant Solar System minor objects. Using orbital data from Gaia Data Release 3, minor body motion is modeled through the Circular Restricted Three-Body Problem (Sun–Jupiter–minor body), with Galactic perturbations included as an external force. Two Galactic models are considered: the axisymmetric MWPotential2014 and a non-axisymmetric Spiral Arms plus MWPotential2014, both implemented with the galpy library. Numerical integrations are used to analyze secular variations of orbital elements over long timescales. The study focuses on the dwarf planet Eris and a hypothetical Sun-bound asteroid with a semi-major axis of 530 AU. Although galactic forces are much weaker than Solar System forces near the Sun, they produce measurable effects at large heliocentric distances: spiral arm effects are negligible near 100 AU, while both galactic components significantly influence orbital stability beyond 500 AU, with the spiral arm pitch angle playing a key role
Experimental characterization of rectangular-aperture acousto-optic deflectors: efficiency, bandwidth, and wavelength dependence
(2026-01) Perrella Benítez, Daniel Eloi; Tiana Alsina, Jordi
This work consists on the characterization of a rectangular-aperture acousto-optic deflector. The characterization is carried out using two different laser beams: Gaussian and elliptical. Furthermore, different wavelengths were used (488 nm, 561 nm, 637 nm). Several parameters of the AOD were measured: relative efficiency as a function of angular deflection, optimal angles of incidence, maximum bandwidth and optimal power and alignment to achieve maximum efficiency, all for different wavelengths.
Measurement of the stopping power of Al and Au for α particles
(2026-01) Moreno Millán, Maria; Fernández Varea, José María
The objective of this work is to measure the stopping power of Al and Au for α particles and analyze the agreement with theoretical models. The measurements were performed using the transmission method with a silicon detector in vacuum, employing two radioactive sources: a sealed 241Am and an unsealed triple-nuclide source (233U, 239+240Pu, 241Am), covering an energy range from 4.3 MeV to 5.3 MeV. Special attention was paid to the target characterization, determining the foil thicknesses to reduce the large uncertainties associated with nominal manufacturer specifications. The results for both materials show excellent agreement with the ASTAR database and confirm the necessity of low-energy corrections to the Bethe formula for high-atomic-number materials
Delayed Fracture in Fiber Bundles
(2026-01) Merin Busquets, Antoni; Ruiz Franco, José Manuel
Fracture is usually associated with the application of stresses or strains exceeding a critical rupture threshold, often referred to as the rupture point. However, many materials can fail after a finite stochastic time under a constant subcritical strain, known as delayed fracture. In this work, we study this phenomenon by performing numerical simulations of a one-dimensional fiber bundle model under constant strain, with the aim of developing a model capable of reproducing delayed fracture. We observe that, within a limited strain range, delayed fracture can occur in this type of model if we introduce a continuous damage rule and a local stiffness redistribution.
Agent-based simulation with non-reciprocal interactions for Lotka-Volterra model
(2026-01) Guevara Fernández, Marc; Ruiz Franco, José Manuel
This work studies the population dynamics of a system with two classes of particles that act as prey and predator. We perform an agent-based simulation where we model the interactions between particles as non-reciprocal, add movements from stochastic origins, and include multiplication and elimination mechanisms. We conclude by making a comparison with the ideal Lotka-Volterra model for population description. We find that our model exhibits oscillations in the populations, but with solutions outside the family of curves of the Lotka-Volterra equations. It is a more realistic model that exhibits the expected dynamics of a more general model of equations than the classical Lotka-Volterra.
Chemical tagging and age estimation with the GALAH DR4 survey
(2026-01) Laguarta González, Alejandra; Anders, Friedrich; Padois, Chloé
Stellar chemical abundances encode valuable information about the formation and evolution of the Milky Way. In this work, we explore two complementary approaches to extract this information from the GALAH DR4 survey. First, we explore the use of a supervised machinelearning algorithm to estimate stellar ages for red giant stars from their chemical abundances and atmospheric parameters, using asteroseismic ages as training data. While the model is able to recover a global age trend, the predicted ages show an unexpectedly poor precision. Second, we analyse the multidimensional chemical abundance space of red clump stars using an unsupervised clustering method, identifying chemically coherent groups. These groups display distinct chemical patterns that can be associated with different components of the Galactic disc.
The Two-Fluid Model of Superfluidity
(2026-01) Molina Ruiz, Fernando; Fernández-Nieves, Alberto
Liquid 4He presents a continuous phase transition at Tλ ≈ 2.17 K from a normal fluid (He I) to a superfluid (He II). This work explores the theoretical framework explaining He II’s unique behavior. We will focus on Landau’s two-fluid model, which treats He II as a mixture of a superfluid component with zero entropy that remains in the ground state, and a normal fluid component associated to the excitations of Helium. The model successfully explains key experimental phenomena and predicts the existence of second sound, a temperature wave unique to quantum fluids.
Simulating atomic nuclei with quantum annealing algorithms: role of the time
(2026-01) Molina Cerdan, Albert; Rios Huguet, Arnau
The nuclear shell model describes atomic nuclei as interacting nucleons occupying different energy orbitals. In this framework, quantum computing raises an important interest due to its more favorable scaling with the number of nucleons compared to the exponential one of classical methods. Our work consists in performing time evolutions from an easily solvable driver Hamiltonian to the complex, target shell model Hamiltonian, known as quantum annealing. Within this approach, we simulate the ground state of light and medium-mass nuclei from 8Be to 44Ti. We perform time evolutions using linear, quadratic and cos2 time schedules and check whether the non-linear schedules improve the linear results in the same number of steps. For instance, for the 20Ne nucleus and 800 steps, we get a final relative difference in energy of 0.23% in the linear schedule and 0.01% and 0.09% for the quadratic and cos2 schedules, respectively. These results open new interesting ways of implementing quantum annealing algorithms.
Modified Gerchberg–Saxton algorithm to generate vector beams
(2026-01) Ivern Ponce de León, Júlia; Maluenda Niubó, David
This study focuses on improving a modified Gerchberg-Saxton (GS) algorithm to generate vector beams under strong focusing conditions. The main goal is to solve the instability problems that happen when the beam waist is very small, where standard methods usually fail. To achieve this, we introduced a regularization strategy using separate relaxation parameters for amplitude and phase to independently smooth the iterative process. After testing some combinations, our results show that by adjusting these parameters, either the intensity or the polarization distribution can be better reconstructed. Therefore, fidelity parameters for both magnitudes are defined to monitor their evolution during the iterative process for each combination and to evaluate the optimal trade-off.
Three-dimensional magnetic textures in iron oxide nanoflowers
(2026-01) Pérez Zegarra, José Manuel; Figueroa Garcia, Adriana Isabel; Fraile Rodríguez, Arantxa
A characterization of the three-dimensional magnetization textures in iron oxide nanoflowers is presented. Tomographic reconstructions of synchrotron-based magnetic Transmission X-ray Microscopy images were obtained using precise reconstruction software, from which the texture of 34 individual particles was identified, described and categorized using geometrical and topological parameters. Four texture families were distinguished based on the morphology of the magnetization streamlines, and their robustness was tested interpreting the topological charge Q as an effective Hopf index. Furthermore, Q was established as a strong indicator of the inner magnetization structure. An apparent tendency in magnetization orientation was identified, linking the observed textures to the preferred crystalline orientations of the domains reported in previous works.
Fault-tolerant quantum computation
(2026-01) Guasch Espinosa, Alba; Iblisdir, Sofyan ; Cervera, Alba
This work analyzes fault-tolerance in quantum computation, namely the use of quantum error correction to prevent the propagation and multiplication of errors in a quantum circuit. A simple quantum circuit for Bell-state preparation is studied and considered in order to theoretically determine the threshold probability pth using Steane code. This threshold corresponds to the maximum admissible error probability for each physical component gate of the circuit. Two error-correction schemes are compared: one applying a single correction at the end of the circuit, another implementing intermediate corrections. In both cases, a threshold of the order of pth ∼ 10−4 is obtained, corresponding to a fidelity of 99,99% per gate. But only the second scheme is suitable for large circuits. The theoretical threshold is compared with recent experimental results, demonstrating the practical feasibility of fault-tolerant quantum computation.
Double-beta decays of light and medium-mass atomic nuclei
(2026-01) Ferrer Aguilera, Malena; Menéndez, Javier
Most atomic nuclei decay via β decay, a weak interaction process. In second-order β decay, two neutrons decay into two protons, accompanied by the emission of two electrons and two antineutrinos. However, if neutrinos were Majorana particles, this would allow the existence of a hypothetical neutrino-less double-β decay (0νββ), which violates the symmetries of the Standard Model of particle physics because only two electrons are emitted. In this work, we compute the 0νββ decay nuclear matrix elements (NMEs) for light and medium-mass nuclei from 6He to 60Ca. We find larger values of the NMEs for decays where the initial and final nuclei are mirror nuclei. Finally, we find a good correlation between 0νββ NMEs and double Gamow-Teller NMEs, which can be measured in nuclear reactions. This correlation is approximately universal for all decays studied.

Examinar

Enviaments recents