Perturbations in the early stages of heavy ion collisions
Submission Date: July 2026
Abstract:
matter expected to be present microseconds after the Big Bang. By colliding heavy nuclei, matter is put under extreme conditions, creating a system far from equilibrium, before it relaxes towards a locally equilibrated state described by relativistic hydrodynamics. This thesis works towards a better understanding of the initial state at the onset of hydrodynamics as well as the preceding non-equilibrium stage. Using an effective kinetic theory of quantum chromodynamics, we investigate the interplay between high momentum perturbations (minijets) and the soft background. A novelty of our work is the examination of hard probes evolving in a non-thermal and expanding medium. By relating the evolution to both hydrodynamical and jet transport coefficients, we compute the timescales of equilibration, respectively. Furthermore, we study thermalisation in the presence of spatial gradients. Close to equilibrium, we provide a benchmark to state-of-the-art hydrodynamical simulations. In the last part of this work, we study event-by-event fluctuations in the initial state, right when hydrodynamics becomes applicable. By considering different geometries of the initial entropy density, we study the spatial distribution of fluctuations and evaluate the impact on the system size.
No-medium baseline for energy loss observables in Oxygen-Oxygen collisions
Submission Date: March 2024
Abstract:
The quark gluon plasma (QGP), a new state of hot and dense nuclear matter, is produced in ultrarelativistic heavy ion collisions. Anisotropic flow and the quenching of high-momentum probes are characteristic signatures of QGP formation. The observation of collective phenomena suggests the presence of medium-induced parton energy loss even in small collision systems. With the use of Monte Carlo Event Generators, the no-medium baseline for the jet nuclear modification factor $R_{AA}$ and semi-inclusive $I_{AA}$ in oxygen-oxygen collisions is computed at next-to-leading order. Deviations from unity demonstrate the importance of nuclear corrections to the initial state as well as the necessity of using a baseline as the reference for detecting medium effects experimentally. It is found that nPDF uncertainties constitute a major limitation in detecting potentially small energy loss effects. Yet, for jet-triggered observables, there exist kinematic regions in which errors cancel down to 2% whereas uncertainties in hadron-triggered measurements remain comparably large.
Angular Anisotropies in the Jet–Medium Response: Probing the Quark–Gluon Plasma with Effective Kinetic Theory
Submission Date: May 2026
Abstract:
Effective kinetic theory (EKT) provides a weak-coupling framework for describing the evolution of perturbations in the quark–gluon plasma produced in heavy-ion collisions. Energy and momentum deposited by energetic jets generate collective excitations in the surrounding medium that manifest as anisotropic flow. The azimuthal harmonic flow coefficients $v_n$ of the medium response are computed around a reference axis offset by an angle $\theta$ from the jet direction, motivated by the single-prong contribution to the two-prong jet response. A single-hit Monte Carlo integrator for the initial slopes $\dot v_n(t_0)$ is developed within linearised EKT. Agreement with an independent extraction from Boltzmann time-evolution data is found across the range of harmonics and 't Hooft couplings considered, establishing the single-hit framework as a reliable method for computing harmonic-flow slopes in EKT. Early-time behaviour of $v_n$ is found to deviate from a simple diffusive-Gaussian broadening picture: the observed harmonic and energy dependences differ from the predicted $n^2$ and $E^{-2}$ scalings. The deviation is found to be specific to the angular structure of the observable.
Anisotropic Viscous Hydrodynamics of Elliptic Flow in a Ten-Atom Unitary Fermi Gas
Submission Date: September 2026
Abstract:
Hydrodynamics is an emergent description, expected to hold only when the mean free path is far shorter than the system size and when there are enough particles for local equilibrium to be a meaningful statement. Elliptic flow – the inversion of an anisotropic cloud’s aspect ratio during expansion – is its clearest signature, and it has been observed in a released gas of only ten lithium atoms at unitarity, a system that satisfies neither condition. This thesis asks how far a hydrodynamic description actually reaches there. Anisotropic viscous hydrodynamics with a relaxation-time closure, which interpolates between ideal hydrodynamics and free streaming, is solved with a PPMLR code adapted here to a trapped cold-atom release: the experimental trap geometry and release protocol, Thomas–Fermi initial conditions and a quantum-pressure term. The code first reproduces published results at $N \sim 10^5$. At $N = 10$ it recovers the measured aspect-ratio inversion whenever $\tau\omega \lessapprox \mathcal{O}(1)$ near the trap centre, while the near-ballistic limit does not invert at all within the experimental window. A joint fit to the measured cloud widths constrains the relaxation-time coefficient $\alpha_n$ to a band of order unity, but leaves the quantum-pressure coefficient $\lambda_{ic}$ essentially free across its physically allowed range $[0, 1]$.
Soft pion enhancement due to resonance decays with spectral functions
Submission Date: December 2024
Abstract:
Hydrodynamical models have shown great success in describing the experimental data measured from heavy-ion collisions. Nevertheless, there still are disparities between the predictions of these models and the transverse momentum spectrum of low momentum pions, leading to the so called soft-pion puzzle. In this thesis, we investigate two approaches to address the soft-pion enhancement observed in Pb-Pb collisions at $\sqrt{s_{NN}} = 2.76\, \mathrm{TeV}$ and $\sqrt{s_{NN}} = 5.02\, \mathrm{TeV}$ using the hadron resonance gas (HRG) model combined with the blast-wave framework. The first approach examines how the number of resonances included in the HRG model affects the blast-wave model fit-parameters when accounting for resonance feed-down. By varying the decay lists used to compute the pion spectrum at 0–5\% centrality, we observe that while the inclusion of additional states improves the model’s fit to data, the reduced chi-squared values ($\chi^2/\text{dof}$) remain significantly above unity, particularly at $\sqrt{s_{NN}} = 5.02\, \mathrm{TeV}$. This suggests that increasing the number of resonances alone is insufficient to resolve the soft-pion puzzle. The second approach explores the impact of broad spectral functions on the soft-pion spectrum. We focus on $\rho$ meson, for which we consider three spectral function parametrizations, namely: Breit-Wigner, phase-shift, and thermal spectral function. Meanwhile, other resonances are described using either a Dirac-delta or a Breit-Wigner distributions. By studying the feed-down of resonances and their contribution to the final pion yield we find that in-medium modifications to the $\rho$-meson spectral function generate the biggest enhancement to the low-$p_T$ region of the pion spectrum, with the largest cumulative feed-down observed at $p_T = 0.1\, \mathrm{GeV}$, corresponding to increases of approximately $10\%$ at $\sqrt{s_{NN}} = 2.76\, \mathrm{TeV}$ and $12\%$ at $\sqrt{s_{NN}} = 5.02\, \mathrm{TeV}$. These results highlight the potential of spectral functions in addressing the soft-pion enhancement observed in heavy-ion collisions at LHC energies.
