-
Machine learning based approach to fluid dynamics
Authors:
Kirill Taradiy,
Kai Zhou,
Jan Steinheimer,
Roman V. Poberezhnyuk,
Volodymyr Vovchenko,
Horst Stoecker
Abstract:
We study the applicability of a Deep Neural Network (DNN) approach to simulate one-dimensional non-relativistic fluid dynamics. Numerical fluid dynamical calculations are used to generate training data-sets corresponding to a broad range of profiles to perform supervised learning with DNN. The performance of the DNN approach is analyzed, with a focus on its interpolation and extrapolation capabili…
▽ More
We study the applicability of a Deep Neural Network (DNN) approach to simulate one-dimensional non-relativistic fluid dynamics. Numerical fluid dynamical calculations are used to generate training data-sets corresponding to a broad range of profiles to perform supervised learning with DNN. The performance of the DNN approach is analyzed, with a focus on its interpolation and extrapolation capabilities. Issues such as inference speed, the networks capacities to interpolate and extrapolate solutions with limited training samples from both initial geometries and evolution duration aspects are studied in detail. The optimal DNN performance is achieved when its objective is set to learn the mapping between hydro profiles after a fixed value time step, which can then be applied successively to reach moments in time much beyond the duration contained in the training. The DNN has an advantage over the conventional numerical methods by not being restricted by the Courant criterion, and it shows a speedup over the conventional numerical methods by at least two orders of magnitude.
△ Less
Submitted 5 June, 2021;
originally announced June 2021.
-
Critical point fluctuations: Finite size and global charge conservation effects
Authors:
Roman V. Poberezhnyuk,
Oleh Savchuk,
Mark I. Gorenstein,
Volodymyr Vovchenko,
Kirill Taradiy,
Viktor V. Begun,
Leonid Satarov,
Jan Steinheimer,
Horst Stoecker
Abstract:
We investigate simultaneous effects of finite system size and global charge conservation on thermal fluctuations in the vicinity of a critical point. For that we consider a finite interacting system which exchanges particles with a finite reservoir (thermostat), comprising a statistical ensemble that is distinct from the common canonical and grand canonical ensembles. As a particular example the v…
▽ More
We investigate simultaneous effects of finite system size and global charge conservation on thermal fluctuations in the vicinity of a critical point. For that we consider a finite interacting system which exchanges particles with a finite reservoir (thermostat), comprising a statistical ensemble that is distinct from the common canonical and grand canonical ensembles. As a particular example the van der Waals model is used. The global charge conservation effects strongly influence the cumulants of particle number distribution when the system size is comparable to that of the reservoir. If the system size is large enough to capture all the physics associated with the interactions, the global charge conservation effects can be accurately described and corrected for analytically, within a recently developed subensemble acceptance method. The finite size effects start to play a significant role when the correlation length grows large due to proximity of the critical point or when the system is small enough to be comparable to an eigenvolume of an individual particle. We discuss our results in the context of fluctuation measurements in heavy-ion collisions.
△ Less
Submitted 14 August, 2020; v1 submitted 29 April, 2020;
originally announced April 2020.
-
The analytic structure of thermodynamic systems with repulsive interactions
Authors:
Kirill Taradiy,
Anton Motornenko,
Volodymyr Vovchenko,
Mark I. Gorenstein,
Horst Stoecker
Abstract:
Thermodynamic properties of systems with repulsive interactions, are considered in the grand canonical ensemble. The analytic structure of the excluded-volume model in the complex plane of the system chemical potential (fugacity) is elaborated, based on the fact that the pressure function can be given in terms of the Lambert W-function. Even though the excluded volume model has no phase transition…
▽ More
Thermodynamic properties of systems with repulsive interactions, are considered in the grand canonical ensemble. The analytic structure of the excluded-volume model in the complex plane of the system chemical potential (fugacity) is elaborated, based on the fact that the pressure function can be given in terms of the Lambert W-function. Even though the excluded volume model has no phase transitions at real values of the chemical potential, it does exhibit a branch cut singularity in the complex plane, thus limiting the convergence range of the Taylor expansion in the chemical potential. Close similarities to analytic properties of the other models with repulsive interactions, such as a cluster expansion model, the mean-field model, and the ideal Fermi gas model, are pointed out. As an example, repulsive baryonic interactions in a hadron gas, with a focus on the fugacity/virial and Taylor expansion methods used in lattice QCD, are presented. The asymptotic behavior of the Fourier expansion coefficients in these various models suggests that the singular part of net baryonic density can to leading order be universally expressed in terms of polylogarithms.
△ Less
Submitted 17 April, 2019;
originally announced April 2019.
