A tensor invariant approach to energy flux in magnetohydrodynamic turbulence Preprint: APS/123-QED

Conan M. Liptrott, Sandra C. Chapman, Bogdan Hnat, Nicholas W. Watkins · 2026

Read A tensor invariant approach to energy flux in magnetohydrodynamic turbulence Preprint: APS/123-QED by Conan M. Liptrott, Sandra C. Chapman, Bogdan Hnat, Nicholas W. Watkins (2026) 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-2604-14426.txt.

Introduction

Liptrott, Chapman, Hnat, and Watkins study energy flux in magnetohydrodynamic turbulence via tensor invariants of velocity and magnetic fields. Filtered flow equations give a scale-by-scale view of the cascade, while gradient-tensor invariants mark the field shapes each transfer process needs. Invariants stand in for mechanistic fluxes under stated limits, bound the flux each process can carry, and express purely hydrodynamic flux exactly. Three-dimensional runs of freely decaying turbulence show the links between field topology and cross-scale energy transfer, with multispacecraft data as a probe.

Catalog record

CREATORConan M. Liptrott, Sandra C. Chapman, Bogdan Hnat, Nicholas W. Watkins
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE68731
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.14426

Browse the full library · About the Tower of Babel

Full text

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

A tensor invariant approach to energy flux in magnetohydrodynamic turbulence Preprint: APS/123-QED Conan M. Liptrott, Sandra C. Chapman, Bogdan Hnat, Nicholas W. Watkins Abstract A scale-by-scale analysis of energy flux in the turbulent cascade can be performed using the spatially filtered magnetohydrodynamic (MHD) equations, while the gradient tensor invariants are widely used to characterise the structure of velocity and magnetic fields. Physical mechanisms responsible for energy flux require specific field configurations whose strength is quantified by these tensor invariants. We explore this requirement, showing that the tensor invariants act as proxies for mechanistic energy fluxes under quantifiable conditions. As a special case, the purely hydrodynamic contributions to energy flux can be expressed exactly in terms of the invariants of the velocity gradient tensor. We also show that the invariants bound the available energy flux for distinct physical mechanisms, formalising the idea that each transfer mechanism requires field configurations with gradients of sufficient strength to support a given energy flux. Results are illustrated using 3D simulations of freely decaying MHD turbulence. I Introduction I.1 Overview Magnetohydrodynamics (MHD) provides an effective description of many astrophysical plasmas, which often exist in a turbulent state ( 7 ; 20 ; 10 ; 11 ; 39 ) . This turbulence is characterised by nonlinear cascades of ideal invariants across a wide range of scales ( 1 ) , giving rise to complex behaviour that remains an active area of theoretical research ( 40 ) . The existence of a turbulent cascade naturally motivates a scale-by-scale analysis of energy flux. A robust construction of energy flux requires a definition of spatial scale which is amenable to theoretical studies and real world data analysis. Spatial filtering ( 19 ) has been widely used to study hydrodynamic turbulence due to its ability to facilitate such scale-by-scale analyses ( 19 ; 26 ; 27 ; 25 ) . This approach has since been applied to MHD, enabling detailed studies of energy transfer across scales (e.g. 28 ; 3 ; 31 ; 9 ), with particular emphasis on the application to sub-grid scale modelling of MHD turbulence ( 2 ) . Extended MHD regimes ( 29 ; 8 ) and fully kinetic simulations ( 44 ) have also used spatial filtering. A key advantage of this approach is that energy flux can be expressed as field gradients, which in turn can be decomposed into contributions from distinct physical mechanisms as well as into local and non-local contributions to the flux ( 26 ; 27 ; 25 ; 12 ) . Following 12 , energy flux in MHD can be partitioned into four subfluxes (Inertial, Maxwell, Dynamo, and Advection), which can in turn be separated by scale locality and decomposed into distinct physical mechanisms, allowing identification of those dominant in driving the turbulence. Recently, an approach based on invariants of the velocity and magnetic field gradient tensors, and their spatially filtered counterparts, has been used to characterise the spatial structure, or topology, of the fields ( 13 ; 43 ; 16 ; 30 ; 15 ; 14 ) . Part of their appeal is their disambiguation of the “smallest scales” and the fact that they display a number of universal signatures across a wide variety of turbulent flows ( 43 ; 30 ) , although such universality remains an open question in MHD ( 15 ) . A further appealing aspect is that multispacecraft measurements allow estimation of field gradients ( 32 ; 18 ) , which can in turn be used to construct the invariants. This approach was first applied to in situ observations by 14 using the Cluster Ion Spectrometer (CIS) instrument aboard the Cluster mission, and has subsequently been applied to a wide range of astrophysical systems ( 36 ; 6 ; 23 ; 24 ; 45 ; 22 ; 37 ; 38 ) . Taken together, these different approaches raise the natural question of whether particular field configurations, as quantified by the invariants, preferentially contribute to energy transfer, as quantified via the filtering approach ( 44 ; 6 ) . 22 recently found an ordering of the local energy transfer rate with magnetic field invariants using in situ solar wind observations, and 8 combined the filtering and tensor invariant approaches in simulations of compressible Hall MHD, observing a clear statistical ordering of the turbulent energy flux with the invariants of the velocity gradient tensor. In this paper, we propose a framework that directly links these invariants to the scale-by-scale transfer of energy in terms of underlying physical mechanisms. We show analytically that the energy fluxes are bounded above by functions of the local invariants, with the least upper bound obtained by solving the exact optimisation problem. The second invariants therefore establish an envelope for the energy available for transfer across scales. We further show that, under quantifiable conditions, the strain-rate invariants serve as proxies for the local energy flux, with the sign of the third invariant \bar{R}_{S} being a reliable predictor of the direction of energy transfer. As a special case, the purely hydrodynamic contributions to energy flux can be expressed exactly in terms of the invariants of the velocity gradient tensor. Together these results constitute a semi-empirical framework, which we call the invariant-flux framework, that explicitly connects the tensor invariants to the energy fluxes across scales. I.2 Structure and Summary of Results The paper is organised as follows. In Section II , we establish the necessary background methodology, including several complementary results that are direct consequences of this methodology and serve as foundational elements for our framework. In Section III , we present the invariant-flux framework, with Section III.1 deriving the energy flux bounds, and Section III.2 introducing the energy flux proxy. Throughout the paper, theoretical results are illustrated using numerical simulations of freely decaying, incompressible 3D MHD turbulence, performed with a pseudospectral code ( 21 ) in a periodic cubic domain of size [0,2\pi]^{3} . Details of these simulations are provided in Appendix A . To assist the reader, we provide a concise summary of our main findings in Table 1 . This table presents the energy flux bounds, the invariant flux proxies, and the complementary results. For brevity, we present the Inertial and Maxwell terms only. The Advection and Dynamo terms either share the same bounds or have very similar bounds due to symmetries, which are discussed where relevant in the text. Invariant-Flux Bounds (Section III.1 ) [-0pt] Contribution Local Contribution Bound Mechanism Bounds (Eq. ( 31 )) Inertial \Pi^{I,\ell} \begin{aligned} |\Pi^{I,\ell}_{\mathrm{L}}|\leq\begin{cases}\frac{2}{\sqrt{3}}\left(\bar{Q}_{\Omega}+\bar{Q}_{S}\right)\left(-\bar{Q}_{S}\right)^{1/2}&\text{if }\bar{Q}_{\Omega}\geq 9|\bar{Q}_{S}|\\ \frac{2}{9}(\bar{Q}_{\Omega}-3\bar{Q}_{S})^{3/2}&\text{if }\bar{Q}_{\Omega}<9|\bar{Q}_{S}|\end{cases}\end{aligned} \begin{aligned} |\Pi^{I,\ell}_{\mathrm{L},SSS}|&\leq\ell^{2}\frac{2}{\sqrt{3}}\left(-\bar{Q}_{S}\right)^{3/2}\\ |\Pi^{I,\ell}_{\mathrm{L},\Omega S\Omega}|&\leq\ell^{2}\frac{2}{\sqrt{3}}\left(-\bar{Q}_{S}\right)^{1/2}\left(\bar{Q}_{\Omega}\right)\end{aligned} Maxwell \Pi^{M,\ell} |\Pi^{M,\ell}_{\mathrm{L}}|\leq\frac{2}{\sqrt{3}}\ell^{2}(-\bar{Q}_{S})^{1/2}[(-\bar{Q}_{\Sigma})^{2}+14(-\bar{Q}_{\Sigma})\bar{Q}_{J}+\bar{Q}_{J}^{2}]^{\frac{1}{2}} \begin{aligned} |\Pi^{M,\ell}_{\mathrm{L},S\Sigma\Sigma}|&\leq\frac{2}{\sqrt{3}}\ell^{2}\left(-\bar{Q}_{S}\right)^{\frac{1}{2}}\left(-\bar{Q}_{\Sigma}\right)\\ |\Pi^{M,\ell}_{\mathrm{L},SJJ}|&\leq\frac{2}{\sqrt{3}}\ell^{2}\left(-\bar{Q}_{S}\right)^{\frac{1}{2}}\bar{Q}_{J}\\ |\Pi^{M,\ell}_{\mathrm{L},SJ\Sigma}|&\leq 4\ell^{2}\left[\bar{Q}_{J}\left(-\bar{Q}_{S}\right)\left(-\bar{Q}_{\Sigma}\right)\right]^{1/2}\end{aligned} Invariant-Flux Proxies (Section III.2 ) [-0pt] Flux Proxy Conditions General \Pi^{X,\ell}_{\mathrm{L},SYZ} \begin{aligned} &\text{Sign follows }\bar{R}_{S}\\ &\text{Magnitude follows }(\bar{R}_{S},\bar{Q}_{S})\text{ via Vi\`{e}tes formula}\end{aligned} \begin{aligned} &\cos(\psi_{2j})\approx 1\text{ and any of:}\\ &\text{(i) }\mu_{j}\text{ constant}\\ &\text{(ii) $\langle\mu_{j}|(\bar{R}_{S},\bar{Q}_{S})\rangle=\langle\mu_{j}\rangle$}\\ &\text{(iii) $\mu_{j}=\mu_{j}(\bar{R}_{S},\bar{Q}_{S})$}\end{aligned} \Pi^{I,\ell}_{\mathrm{L},SSS} 3\ell^{2}\bar{R}_{S} Exact \Pi^{I,\ell}_{\mathrm{L},\Omega S\Omega} \ell^{2}(\bar{R}_{S}-\bar{R}_{A}) Exact \Pi^{I,\ell}_{\mathrm{L}} \ell^{2}\left(4\bar{R}_{S}-\bar{R}_{A}\right) Exact Complementary Results [-0pt] Use Formula Section Viètes formula for strain-rate eigenvalues \lambda_{k}=2\sqrt{-\frac{\bar{Q}_{S}}{3}}\cos\left[\frac{1}{3}\arccos\left(\frac{3\bar{R}_{S}}{2\bar{Q}_{S}}\sqrt{-\frac{3}{\bar{Q}_{S}}}\right)-\frac{2\pi k}{3}\right] ( II.4.4 ) Dissipation rates from second invariants \mathcal{D}_{u}=-4\nu\bar{Q}_{S},\quad\mathcal{D}_{b}=4\eta\bar{Q}_{J} ( II.4.2 ) Table 1 : Summary of energy flux bounds and proxies via the coarse-grained tensor invariants. Energy flux quantities \Pi^{X,\ell}_{Y,Z} are defined in Section II.2 and depicted for reference in Fig. 1 . The scale-dependent tensor invariants \bar{Q}_{X}

Read the complete text