Partonic transport model application to heavy flavor

Abstract

Heavy-flavor particles are excellent probes of the properties of the hot and dense nuclear medium created in the relativistic heavy-ion collisions. Heavyflavor transport coefficients in the quark-gluon plasma (QGP) stage of the collisions are particularly interesting, as they contain important information on the strong interaction at finite temperatures. Studying the heavy-flavor evolution in a dynamically evolving medium requires a comprehensive multistage modeling approach of both the medium and the probes, with an accurate implementation of the physical ingredients to be tested. For this purpose, I have developed a new partonic transport model (Linear-Boltzmannplus-Diffusion-Transport-Model) LIDO and applied it to heavy quark propagation inside a QGP. The model has an improved implementation of parton in-medium bremsstrahlung and a flexible treatment of the probe-medium interactions, combining both large angle scatterings and diffusion processes. The model is then coupled to a high-energy event-generator, a hydrodynamic medium evolution and a hadronic transport model. Finally, applying a Bayesian analysis, I extract the heavy quark transport coefficients from a model-to-data comparison. The results, with uncertainty quantification, are found to be consistent with earlier extraction of the light-quark transport coefficients at high momentum and with first-principle calculations of the heavy flavor diffusion constant at low momentum.

QGP modification to single inclusive jets in a calibrated transport model

“…This suggests that induced-radiation processes alone can build up energy storage near the thermal scale p ∼ T , no matter how energetic the initial hard jets are. Such behavior has also been confirmed in recent numerical simulations including the effective kinetic theory with resummed radiation vertex [84] and the LIDO transport model in the large-medium L med τ f limit, which is discussed in section 3.5.3 of [85]. We will see immediately that this feature indeed emerges from our simulation.…”

Extracting jet transport coefficient via single hadron and dihadron productions in high-energy heavy-ion collisions

“…Although an explicit parametric form is intuitive and meaningful to humans, it is unnecessarily complicated for machine learning as it often contains strong non-linear correlations. For example, in the parametrization of a previous study [102] q/T 3 = (1 + (aT /T c ) p ) −1 , a linear variation in a or p causes a strong non-linear change in the effective value of q. Even though the physical model maps values of q to observables in a rather well-behaved fashion, the performance of the emulator can still be impaired by the non-linear response to those individual input parameters.…”