Research

My research activities include:

Weak-field methods for Glasma simulations

The Glasma is the precursor stage to the quark-gluon plasma that can be created in heavy-ion collisions. We have been creating a simulation framework for the Glasma, which is computationally very demanding. In the weak-field approximation, certain parts of the calculation can be handled analytically, speeding up simulations tremendously. We compare these approximate results to the full simlulations and find remarkable agreement. This work has been supported by the Austrian Science Fund FWF, project number P32446-N27 and P34764-N27.

Machine learning applied to physics simulations

We study varyious ways in which machine learning can be applied to physics simulations. A central aspect is to preserve symmetries of the underlying theory also in the neural network. For lattice gauge theory, we propose Lattice gauge equivariant Convolutional Neural Networks (L-CNNs) that preserve gauge equivariance. We demonstrate that L-CNNs can learn and generalize gauge invariant quantities that traditional convolutional neural networks are incapable of finding. This work has been supported by the Austrian Science Fund FWF, project number P32446-N27.

Jet momentum broadening from the Glasma

Jets are important probes of heavy ion collisions. They can provide information on the interaction of a highly energetic parton with the expanding quark-gluon plasma. Transverse momentum broadening quantifies the modification of a hard probe in the medium. We calculate the momentum broadening contribution from the Glasma phase, which is the initial stage right after a heavy-ion collision. We find strong time-dependence and an intrinsic anisotropy of momentum broadening in the directions transverse to the jet propagation direction.

Glasma simulation using colored particle-in-cell method

The Glasma is a precursor state to the quark-gluon plasma. It is created when two highly Lorentz-contracted nuclei collide at ultra-relativistic energies. We have created the OpenPixi open-source framework for the simulation of the initial stages of heavy ion collisions. This framework is inspired by the particle-in-cell method which is used for simulating electromagnetic plasmas. In our case, we apply this method to the strong nuclear force, also called color force, with concepts borrowed from lattice gauge theory. We could show that a finite nucleus thickness along the longitudinal collision direction leads to a Gaussian rapidity profile of observables like the Glasma energy density. This work has been supported by the Austrian Science Fund FWF, project number P26582-N27.

Yoctosecond photon emission from quark-gluon plasmas

High-energy heavy ion collisions at CERN LHC or RHIC can be a source of light flashes with very interesting properties. The quark-gluon plasma produced in such collisions is a state of matter at extremely high temperatures which consists of deconfined, but strongly interacting quarks and gluons. Due to the extremely short time of collision, photons can be emitted at yoctoseconds (10-24 s) duration. Non-trivial expansion dynamics right after the collision of two heavy ion nuclei lead to the possibility of creating double flashes at the yoctosecond scale under certain conditions. Another aspect we studied is the polarization of photons. In non-central collisions, photons may be circularly polarized.

Zeptosecond streak imaging through vacuum pair creation

We recently proposed a new concept for characterizing extremely short photon pulses down to zeptosecond (10-21 s) resolution. This is achieved by applying the widely used femto- and attosecond scheme of streak imaging to the high energy process of electron-positron pair creation in strong fields. Conventional streak imaging would not work at GeV energies, because the non-linear conversion material usually used would be simply destroyed by the high-intensity laser. The solution is not to use any material at all, but provide the non-linear conversion through electron-positron pair creation from vacuum in intense laser fields. We called this new method Streaking at High Energies with Electrons and Positrons ("SHEEP"). This method could provide unprecedented detection capabilities of extremely fast high-energy processes at future laser facilities like the upcoming Extreme Light Infrastructure (ELI).

Quark-Gluon plasma instabilitites

An important process on the yoctosecond timescale is QGP plasma instabilities that emerge in the expanding quark-gluon plasma in heavy ion collisions. These instabilities are the non-abelian analog of Weibel instabilities in ordinary electromagnetic plasmas. Such plasma instabilities could provide an explanation for the still not fully understood fast apparent thermalization of the plasma. A study of these instabilities requires extensive numerical simulations. The computations were performed on the Vienna Scientific Cluster (VSC).

QCD thermodynamics

The quark-gluon plasma is a state of matter of which the universe consisted right after the big bang. In this state, the temperatures are so high that protons and neutrons are split into their constituents, the quarks and gluons. Such a state of matter can be realized in heavy ion colliders like RHIC at Brookhaven, at the LHC at CERN near Geneva, and in a few years also at FAIR, GSI Darmstadt. An important scientific goal is to explore the phase diagram of quantum chromo-dynamics (QCD) for various temperatures and densities. We have calculated the pressure for weakly interacting, deconfined QCD up to and including forth order in a perturbative expansion in the strong coupling.

Functional renormalization group

The functional or non-perturbative renormalization group is a promising tool to describe non-perturbative phenomena, including the strong coupling limit of QCD. This method has been applied successfully to various fields of physics, for example in the description of critical exponents. We have adapted the tool for calculating thermodynamic properties like the pressure or the thermal mass of a scalar field theory and extended it to be applied to an improved truncation scheme. The improved scheme will allow to describe dynamical properties of the plasma, like decay rates or transport coefficients using the non-perturbative renormalization group.

Non-Fermi-Liquid behavior of QCD

Anomalous contributions to the specific heat beyond the classical Landau-Fermi liquid theory are caused by long-range quasi-static transverse gauge bosons. While in ultrarelativistic quantum electrodynamics (QED) the effect is tiny, in QCD it is much more pronounced. Using a formalism based on the large flavor number limit, we could complete the argument of the leading logarithm of the specific heat and go beyond leading order to find a series in the temperature involving fractional powers. These results have direct application in the calculation of the cooling of proto-neutron-stars.

Larger flavor number limit of QCD

At large number of flavors (Nf), thermodynamic quantities like the pressure or entropy can be calculated exactly for weak and strong couplings at next-to-leading order in a 1/Nf expansion. It is therefore an ideal testbed for various resummation techniques that try to overcome the poor convergence properties of strict perturbation theory. This limit turns out to be straightforwardly applicable to finite chemical potential. It was also used to successfully test a nonperturbative expression for the entropy obtained from a Phi-derivable two-loop approximation which resums the physics of hard thermal loops.