Phase transitions and finite-size effects in integrable virial statistical models

Xin An, Francesco Giglio, Giulio Landolfi · 2025

Read Phase transitions and finite-size effects in integrable virial statistical models by Xin An, Francesco Giglio, Giulio Landolfi (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-15719.txt.

Introduction

Xin An, Francesco Giglio, and Giulio Landolfi study fluid models in a 2025 paper that links heat facts to wave math. Their models build the stored energy of a fluid step by step from how packed it is. The team shows the models can be solved in full for any set size and any step count.

When the set grows vast, sharp shifts in the fluid show up as shock waves in the heat space. Near the tipping point, the packing follows a scaling law the team ties to a broad claim about such flows. They then apply the frame to dense star stuff and quark stuff. Physicists gain a rare model that stays exact yet still shows real phase shifts.

Catalog record

CREATORXin An, Francesco Giglio, Giulio Landolfi
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE26 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-15719

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Phase transitions and finite-size effects in integrable virial statistical models Xin An, Francesco Giglio, Giulio Landolfi Abstract We analyze thermodynamic models for fluid systems in equilibrium based on a virial expansion of the internal energy in terms of the volume density. We prove that the models, formulated for finite-size systems with N N particles, are exactly solvable to any expansion order, as expectation values of physical observables (e.g., volume density) are determined from solutions to nonlinear C-integrable partial differential equations (PDEs) of hydrodynamic type. In the limit N → ∞ N\to\infty , phase transitions emerge as classical shock waves in the space of thermodynamic variables. Near critical points, we argue that the volume density exhibits a scaling behavior consistent with the Universality Conjecture in viscous transport PDEs. As an application, we employ our framework to nuclear and quark matter, constructing a global quantum chromodynamics (QCD) phase diagram that reveals critical points for the nuclear liquid-gas transition and the hadron gas–quark-gluon plasma transition. We demonstrate how finite-size effects smear critical signatures, implying their potential impact on the search for the QCD critical point. Introduction. Understanding phase transitions in matter is a fundamental subject in physics. The presence of interparticle interactions, as well as the involvement of several thermodynamic functions, can result in effects at different scales, with possible multiple critical points and distinct ground states. Variation of a thermodynamic function, such as temperature or pressure, can suppress one phase while promoting another, unveiling a rich interplay between competing phases [ 1 , 2 , 3 ] . Such phenomena are ubiquitous in semiconductors [ 4 ] , crystal field [ 5 ] , spin systems [ 6 ] , polymorphic fluids [ 7 ] , and matter in quantum chromodynamics (QCD) [ 8 ] . Despite the numerous advances in extensive theoretical and experimental studies of these systems, there remains however a need for viable phenomenological approaches and global exact results beyond the regime of the critical point and the thermodynamic limit (TL), which largely rely on costly lattice simulations [ 9 ] . In this Letter, we analyze the thermodynamics of fluid systems where the nontrivial interactions among particles result in an effective potential energy in the form of a virial-like expansion in the order parameter. Our approach, inspired by papers [ 10 , 11 ] , will enable us to identify partial differential equations (PDEs) for the associated statistical models, and to predict behavior of their solutions globally and locally even beyond TL, when finite-size effects are foreseen [ 12 , 13 , 14 , 15 , 16 ] . Such effective statistical models provide a simple yet powerful description for van der Waals (vdW) or more imperfect fluid systems across various areas of physics, including classical liquid-gas (L-G) systems [ 17 ] , the rotating or charged anti-de Sitter black holes [ 18 ] , astrophysical objects [ 19 ] , ultracold quantum gases [ 20 ] and last but not least, dense nuclear and quark matter in QCD [ 21 ] where the finite-size effects are expected to be significant—the major motivation for this work. Virial-like statistical models. Phase transitions are conveniently studied within the framework of statistical mechanics following the maximum entropy principle [ 17 ] . This principle turns into an extremity problem for the proper entropic potential \Psi for a specific statistical ensemble, which is related to the system’s entropy S —the relevant potential in the microcanonical ensemble whose natural variables are a set of conserved quantities denoted by vector \Lambda —by Legendre transform \Psi(\tau)=\sup_{\Lambda}\,[-S(\Lambda)+\tau\Lambda] , where \tau\Lambda=\sum_{j=1}^{J}\tau_{j}\Lambda_{j} is a product of J conjugate pairs \{\Lambda_{j},\tau_{j}\} . The probability of any available microscopic state labeled by i is weighed as p_{i}=e^{-\tau\Lambda^{(i)}}/{\mathcal{Z}} , normalized by the partition function associated with the potential \Psi(\tau) to ensure \sum_{i}p_{i}=1 : \mathcal{Z}(\tau)=e^{-\Psi(\tau)}=\sum_{i}e^{-\tau\Lambda^{(i)}}\propto\int d\Lambda\,e^{S(\Lambda)-\tau\Lambda}\,, where in the last step we have taken the continuum limit of the summation over all microscopic states to the integration over d\Lambda=d\Lambda_{1}\dots d\Lambda_{J} and dropped a multiplicative constant. Once the partition function for a given system is assigned, all thermodynamic properties follow [ 22 ] . For a generic hard spheres fluid system, the microcanonical entropy (up to irrelevant constants) has the well-known form S(E,V,N)=N\ln\left[\frac{V-Nb}{N}\left(\frac{E-U(V,N)}{N}\right)^{3/2}\right]\,, (1) which depends on its natural (extensive) variables such as internal energy E , volume V , and particle number N . The parameter b identifies the minimum volume density 1 1 1 Throughout this work by density we mean per particle . of the configurations in the ensemble. The total potential energy U is assumed a virial-type form U(V,N)=-\sum_{j=1}^{m}\frac{N^{j+1}a_{j+1}}{jV^{j}}\,, (2) where the virial expansion coefficients a_{j} , for j=1,\dots,m , are real constants, since for hard core (or short range) potentials, these coefficients do not (approximately) depend on temperature. 2 2 2 Virial coefficients may acquire temperature dependence via, e.g., corrections from quantum statistics [ 23 ] , which is beyond the scope of this work. The case m=1 reproduces the potential energy of the celebrated vdW model. From a mean-field perspective, the potential U can be seen to account for the dependence of the averaged particles interaction upon their average distance \bar{r}\sim v^{1/3} , where v\equiv V/N denotes the volume density. For the case considered in this Letter, where the energy E and volume V of the fluids vary while the particle number N is kept fixed, it is more convenient to consider the isothermal-isobaric ensemble where its natural (intensive) variables, t=(\partial_{E}S)_{V} and x=(\partial_{V}S)_{E} , control the fluctuations of energy and volume respectively. 3 3 3 For a one-component fluid system, two independent intensive degrees of freedom are needed according to Gibbs phase rule , implying J=2 . In terms of standard pressure P and temperature T , t=1/T and x=P/T . This choice has been shown to arise naturally in the treatment of vdW [ 10 , 11 ] and nematic fluids [ 24 ] . Thus we are led to consider the scaled Gibbs free energy (i.e., Gibbs free energy divided by temperature) \mathcal{G}_{N}(t,x)\equiv\mathcal{G}(t,x;N)=\sup_{E,V}\,[-S(E,V,N)+tE+xV]\,, from which the partition function is defined by 4 4 4 The boundary effects in finite systems can be described by introducing in the free energy a surface energy term suppressed by N^{-1/3} [ 25 , 26 ] . \mathcal{Z}_{N}(t,x)\equiv e^{-\mathcal{G}_{N}(t,x)}=\int dEdV\,e^{S(E,V,N)-tE-xV}\,. (3) Since U=U(V,N) , the integration over E accounts only for the kinetic part K=E-U . This contributes an irrelevant normalization factor c(t)\propto t^{-3N/2} to the partition function ( 3 ), provided that the observables O under consideration are independent of K . Consequently, the integration in Eq. ( 3 ) reduces to one over all ensemble configurations labeled by V (or v for fixed N ) alone and one can define another partition function Z_{N} via \mathcal{Z}_{N}=c(t)Z_{N} : Z_{N}(t,x)\equiv e^{-G_{N}(t,x)}\equiv e^{-Ng_{N}(t,x)}\\ =\int dv\,e^{-N\psi(t,x;v)}\,, (4) where G_{N} and g_{N} denote the associated free energy and its density respectively, while \psi(t,x;v)=-s(v)+t\epsilon(v)+xv (5) is a potential density function. The expectation value of observable O_{N}(v)=O(V,N) is thus given by 5 5 5 Generically, O=O(E,V,N) . For example, the average kinetic energy \langle{K(E,V,N)\rangle}\equiv\langle{E-U(V,N)\rangle}=3N/2t , being consistent with the equipartition theorem. \langle O_{N}\rangle=\frac{1}{Z_{N}}\int dv\,e^{-N\psi(t,x;v)}O_{N}(v)\,. (6) Accordingly, we focus on the configurational entropy and potential energy density given by s(v)=\ln(v-b)\,,\qquad\epsilon(v)=-\sum_{j=1}^{m}\frac{a_{j+1}}{jv^{j}}\,, (7) which result from Eqs. (

Read the complete text