QFT on rotating boxes at finite temperature

Sebestyen Nagy, Daniel Nogradi · 2025

Read QFT on rotating boxes at finite temperature by Sebestyen Nagy, Daniel Nogradi (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-2509-19933.txt.

Introduction

Quantum field theory at finite temperature usually lives on a four-dimensional torus, but rotation complicates that picture. Sebestyen Nagy and Daniel Nogradi ask which space-time shapes allow a spinning thermal system. Moving frames can be faked with shifted boundary conditions on the torus. True rotation, they show, needs different topologies, still flat and without boundary.

They classify all the geometries that work. Rotated boundary conditions restore a path integral formulation on those shapes. The allowed spin rates, in temperature units, form a discrete set rather than a continuum. Finite volume effects get a detailed treatment. Rotating thermal matter appears in heavy-ion collisions and compact stars. A consistent lattice-ready formulation opens those systems to first-principles study.

Catalog record

CREATORSebestyen Nagy, Daniel Nogradi
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE34 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-19933

Browse the full library · About the Tower of Babel

Full text

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

QFT on rotating boxes at finite temperature Sebestyen Nagy, Daniel Nogradi Abstract We formulate thermal quantum field theory on a finite spatial periodic volume undergoing rotation. Traditional compactifications at finite temperature without rotations typically involve 𝕋 4 {\mathbb{T}}^{4} as the space-time manifold within a path integral formulation and also moving frames can be accommodated by shifted boundary conditions on the same space. We show that consistent descriptions of a rotating box are possible on space-time manifolds with topology different from 𝕋 4 {\mathbb{T}}^{4} but still flat and without boundary and we classify all possible geometries. The non-trivial topology may be implemented by rotated boundary conditions allowing for a path integral formulation. The purely imaginary angular velocity in temperature units cannot be arbitrary but several discrete values are possible. We also discuss finite volume effects in detail. I Introduction The most successful non-perturbative description of quantum chromodynamics (QCD) is via a path integral formulation. This approach allows the calculation of various properties of quarks and gluons at finite volume, finite temperature and coupling to external parameters such as conserved charges and background fields. Matter made up of quarks and gluons may also rotate at a constant angular velocity which may have significant effect on the phase diagram and lead to novel phenomena such as the chiral vortical effect Vilenkin:1979ui ; Erdmenger:2008rm ; Banerjee:2008th ; Son:2009tf ; Kharzeev:2010gr ; Landsteiner:2011cp ; Kharzeev:2015znc . In several experimental and observational scenarios rotation of QCD matter cannot be ignored. For instance in fast rotating neutron stars Grenier:2015pya ; Watts:2016uzu ; Paschalidis:2016vmz and in relativistic off-central heavy ion collisions Becattini:2007sr ; Jiang:2016woz . Naturally, there are many orders of magnitude differences in angular velocities between the two scenarios. In the latter case, which will be the focus of this paper, the experimental findings on \Lambda hyperon spin polarizations indicate vorticity of the plasma may reach as high as about \sim 10\,MeV STAR:2017ckg . The most natural approach to the spatial geometry would involve a finite cylinder. The radius perpendicular to the axis of rotation would necessarily be finite. This geometry, naturally, allows for arbitrary infinitesimal velocities, with an upper bound coming from the fact that in a relativistic theory the edge of the cylinder should not be moving faster than the speed of light. Nevertheless there is a well-defined infinitesimal rotational generator which commutes with the Hamiltonian. In quantum field theory, the simplest implementation would involve the introduction of a rotating frame Ambrus:2014uqa ; Chernodub:2016kxh , leading to a non-trivial space dependent metric tensor. The non-trivial metric on the cylindrical geometry presents, however, a crucial problem, namely it complicates the non-perturbative formulation on the lattice. First of all, it is not clear what the right boundary conditions are spatially, second, the lattice discretization itself is not compatible with infinitesimal rotations which feeds into the first problem of spatial boundary conditions and third, the definition of temperature in a curved background is non-trivial. The approach has nevertheless been used in lattice simulations Yamamoto:2013zwa ; Chernodub:2017ref ; Braguta:2021jgn ; Braguta:2023iyx with rather surprising results involving negative moment of inertia in a certain temperature range Braguta:2023yjn . Given the importance of phenomena related to rotations and their non-perturbative lattice studies here we propose a different approach. In the continuum the spatial box will be a periodic 3-torus just as the starting point of most lattice studies without rotation. Even though the 3-torus does break Poincare invariance, depending on its geometry a finite subgroup of the Poincare group can be preserved. Since there are not any infinitesimal rotational generators, it is not possible to include the effect of rotation in a non-trivial metric. Instead, after Wick rotation, the unitary operator corresponding to the preserved finite rotation over the full period in Euclidean time, will be inserted into the partition function directly. This will then be represented as a path integral over fields subject to novel rotated boundary conditions in the Euclidean time direction. Discretization does not then present conceptual problems, no further breaking of the preserved rotational group occurs by a hypercubic lattice. The rotated temporal boundary conditions are straightforward to implement with a well-defined continuum limit at finite volume and temperature. Technically, the continuum space-time manifolds we propose are closely related to the 4-torus, {\mathbb{T}}^{4} , most frequently used at finite volume and temperature. They are all also compact, flat, orientable and without boundary and are of the form {\mathbb{T}}^{4}/G with a finite group G that involves the aforementioned unbroken rotational subgroup of SO(3) . The topology of space-time with a non-trivial G will be different from {\mathbb{T}}^{4} though. It is perhaps useful to draw a parallel with the introduction of other external parameters, e.g. the quark chemical potential. In a path integral formulation there are at least two approaches. One, the action is modified locally by the chemical potential whereas the fields remain periodic, or two, leaving the action the same as before, but introducing the chemical potential in a boundary condition over the full period in the temporal direction. In the case of rotations the introduction of the non-trivial metric locally corresponds to the first choice and thereby the action has explicit dependence on the velocity. This is however not possible to do, even in the continuum, on a 4-torus. With a different geometry, like a cylinder, the continuum setup would be meaningful but a lattice discretization will introduce more breaking. Our approach is akin to the second setup whereby the entire effect is encoded in a temporal boundary condition and the action is the same as without rotation. In this way no mention is made of infinitesimal rotations thereby allowing for a well-defined discretized path integral formulation respecting the finite rotations of the continuum setup. II Conserved charges In order to discuss finite volume and finite temperature let us assume space-time is compactified in such a way that it remains flat, orientable and without boundary. Later we will see that the most often used manifold, the 4-torus, is not the only choice topologically. The partition function in the presence of a number of local conserved charges Q_{A} and chemical potentials \alpha_{A} can be written as, \displaystyle Z(\beta,\alpha_{A})={\rm Tr\,}e^{-\beta\left(H-\alpha_{A}Q_{A}\right)}\;, (1) where \beta=1/T is the inverse temperature and H the Hamiltonian, which is assumed to be Poincare invariant. For instance, since the total spatial momentum P_{k} is conserved, it can be chosen with velocities as parameters, inserting \exp(\beta v_{k}P_{k}) into the partition function trace. This setup corresponds to studying the system in a moving frame. In a Euclidean setup the velocity is naturally chosen purely imaginary, iv_{k}=Ty_{k} with a real spatial vector y_{k} . The partition function in this case can be computed via a path integral formulation where the fields satisfy shifted boundary conditions Giusti:2010bb ; Giusti:2011kt ; Giusti:2012yj ; Giusti:2014ila ; Giusti:2016iqr , \displaystyle\phi\left(t+L_{0},{\bm{x}}\right)=\phi(t,{\bm{x}-\bm{y}})\;, (2) where L_{0} is the temporal extent and \sqrt{L_{0}^{2}+{\bm{y}}^{2}}=1/T . It is worthwhile to point out that shifted boundary conditions are implemented on {\mathbb{T}}^{4} , topologically the same manifold on which one would formulate the theory without shifts i.e. zero velocity. The analogous setup for a rotating frame would start with ( 1 ), the role of the chemical potentials would be played by angular velocities and the charges would be the

Read the complete text