Toward precise ξ \xi gauge fixing for the lattice QCD

( χ \chi QCD Collaboration) Li-Jun Zhou, Dian-Jun Zhao, Wei-jie Fu, Chun-Jiang Shi, Ji-Hao Wang, Yi-Bo Yang · 2025

Read Toward precise ξ \xi gauge fixing for the lattice QCD by ( χ \chi QCD Collaboration) Li-Jun Zhou, Dian-Jun Zhao, Wei-jie Fu, Chun-Jiang Shi, Ji-Hao Wang, Yi-Bo Yang (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-2509-09367.txt.

Introduction

Li-Jun Zhou and 5 co-authors work toward precise gauge fixing for lattice QCD in a 2025 paper. Lattice QCD solves quantum chromodynamics from first principles on a spacetime grid. Work on off-shell partons has mostly stayed in the Landau gauge, since precise fixing in other gauges is hard.

A universal power law link between parton matrix elements and gauge fixing precision motivates an empirical extrapolation method. Proper definitions of bare coupling and effective gauge parameter support the procedure. The method reproduces gauge dependent renormalization constants for quark operators at a fraction of a percent level. That accuracy holds up to a gauge parameter near one. Wider parton studies may follow.

Catalog record

CREATOR( χ \chi QCD Collaboration) Li-Jun Zhou, Dian-Jun Zhao, Wei-jie Fu, Chun-Jiang Shi, Ji-Hao Wang, Yi-Bo Yang
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE30 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-09367

Browse the full library · About the Tower of Babel

Full text

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

Toward precise ξ \xi gauge fixing for the lattice QCD ( χ \chi QCD Collaboration) Li-Jun Zhou, Dian-Jun Zhao, Wei-jie Fu, Chun-Jiang Shi, Ji-Hao Wang, Yi-Bo Yang Abstract Lattice QCD provides a first-principles framework for solving Quantum Chromodynamics (QCD). However, its application to off-shell partons has been largely restricted to the Landau gauge, as achieving high-precision ξ \xi -gauge fixing on the lattice poses significant challenges. Motivated by a universal power-law dependence of off-shell parton matrix elements on gauge-fixing precision in the Landau gauge, we propose an empirical precision extrapolation method to approximate high-precision ξ \xi -gauge fixing. By properly defining the bare gauge coupling and then the effective ξ \xi , we validate our ξ \xi -gauge fixing procedure by successfully reproducing the ξ \xi -dependent RI/MOM renormalization constants for local quark bilinear operators at 0.2% level, up to ξ ∼ 1 \xi\sim 1 . I Introduction The gauge invariance is a crucial property of gauge theories. One of the most fundamental principles in constructing a gauge theory is that the Lagrangian should remain gauge-invariant. But the usual quantization and also perturbative calculation requires to introduce an additional gauge fixing term, e.g., (\partial_{\mu}A^{\mu})^{2}/(2\xi) , while the choice of gauge can be rather arbitrary and may vary among different physicists. For example, the Feynman gauge with \xi=1 simplify the form of the gauge boson propagator and then all the perturbative calculation, while the Landau gauge with \xi=0 can minimize the loop correction of the quark field at the 1-loop level. Nevertheless, physical observables derived from different gauge fixing choices are also gauge-invariant as the experimental value is evidently unique. The scenario in lattice QCD differs slightly from the framework described above. In lattice QCD, the fundamental gauge degree of freedom–the gauge potential A_{\mu}(x) at a given spacetime point x –is replaced by the gauge link U_{\mu}(x+\hat{\mu}a/2)=e^{iag\int_{x}^{x+\hat{\mu}a}\mathrm{d}y^{\mu}A_{\mu}(y)} connecting x and x+\hat{\mu}a . Gauge invariance is then automatically ensured as long as the gauge links terminate at quark fields or form closed loops, as in hadronic correlation functions. On the other hand, gauge-dependent quantities–such as off-shell parton (quark and gluon) propagators and interaction vertices–vanish entirely unless gauge fixing is imposed. Those quantities are important ingredients of nonperturbative functional QCD, such as the Dyson-Swinger Equations (DSE) Chang et al. 2011 ; Qin et al. 2011 ; Bashir et al. 2012 ; Fischer et al. 2014 ; Gao et al. 2016 ; Aguilar et al. 2019 ; Gao and Pawlowski 2021 ; Roberts et al. 2021 ; Gunkel and Fischer 2021 ; Chang et al. 2021 and the functional renormalization group (fRG) Mitter et al. 2015 ; Braun et al. 2016 ; Cyrol et al. 2016 ; Cyrol et al. 2018 ; Fu et al. 2020 ; Fu et al. 2023 ; Fu et al. 2024 ; Ihssen et al. 2024 ; Fu et al. 2025 ; Zhang et al. 2025 ; Dupuis et al. 2021 ; Fu 2022 . Usually the gauge-dependent propagators and vertices are computed in functional QCD with the Landau gauge \xi=0 , from which benchmark comparison between the functional QCD and lattice QCD can be made, see e.g. Fu et al. 2025 for a recent study. Lattice QCD simulations at low energy scales with the Landau gauge have shown kinds of highly-nontrivial features at hadron scale, such as the emergent masses of the quark and gluon Bowman et al. 2005 ; Boucaud et al. 2018 ; Chang et al. 2021 , non-degenerate gauge coupling from the gluon-ghost Zafeiropoulos et al. 2019 and triple gluon vertices Aguilar et al. 2020 , and so on, which are consistent with the results of functional QCD, cf. e.g., Mitter et al. 2015 ; Braun et al. 2016 ; Cyrol et al. 2016 ; Cyrol et al. 2018 ; Fu et al. 2020 ; Ihssen et al. 2024 ; Fu et al. 2025 . However, most of those lattice calculations are restricted to the Landau gauge, despite the fact that lattice implementations of general \xi -gauge fixing Fujikawa et al. 1972 were proposed years ago Giusti 1997 ; Cucchieri et al. 2009 ; Bicudo et al. 2015 . The primary challenge lies in severe convergence issues that arise at large \xi and/or strong gauge coupling g , making the application of \xi -gauge fixing to realistic configurations numerically demanding, to understand how those inferred features of parton are sensitive to the specific Landau gauge fixing. Recently, the dependence of gauge links and non-local operators on gauge-fixing precision has been investigated at multiple lattice spacings and for varying gauge link lengths, in both Landau and \xi gauges Zhang et al. 2024 . The values of these quantities follow an empirical power law in terms of gauge-fixing precision, regardless of the gauge link length. In this work, we further validate that this power law also holds for local operators with different gamma matrices and off-shell momenta. Based on this, we propose a precision-extrapolation method to approximate high-precision \xi -gauge fixing with controllable systematic uncertainty. The paper is organized as follows: In Sec. II , we briefly review the gauge-fixing procedure, the empirical power law, and existing numerical evidence in Landau gauge. Section III presents our results on the non-perturbative \xi -gauge dependence of quark bilinear operators, including a detailed comparison with perturbative calculations. Finally, Sec. IV provides a concise summary of our findings. II Gauge fixing and its precision extrapolation II.1 Gauge fixing on the lattice In the path integral formalism, gauge fixing with the additional Lagrangian term (\partial_{\mu}A^{\mu})^{2}/(2\xi) can be equivalently implemented by introducing 3^{2}-1 random variables \Lambda^{a} which follow the distribution P(\Lambda^{a}(x))=\frac{1}{\sqrt{2\pi\xi}}\exp\left\{-\frac{1}{2\xi}[\Lambda^{a}(x)]^{2}\right\} . The gauge fixing condition is then enforced by integrating over \Lambda\equiv\Lambda^{a}t^{a} with the delta function constraint \delta\left(\partial_{\mu}A^{\mu}(x)-\Lambda(x)\right) , where t^{a} are the generators of the adjoint representation of \mathrm{SU}(3) . On the lattice, the delta function constraint is discretized into the gauge-fixing condition: \displaystyle\Delta(x) \displaystyle\equiv\sum_{\mu,\eta=\pm}\eta\left[\frac{U_{\mu}(x+\eta\frac{\hat{\mu}}{2}a)-U_{\mu}^{\dagger}(x+\eta\frac{\hat{\mu}}{2}a)}{2{\rm i}g_{0}}\right]_{\text{Traceless}} \displaystyle\quad\quad-\Lambda(x)a^{2}=0, (1) and the integration over \Lambda can be efficiently performed by averaging over different gauge configurations with independent \Lambda . The bare gauge coupling g_{0} in Eq. ( II.1 ) can be defined in multiple ways, differing at next-to-leading order in \alpha_{s} . As an example, consider the tadpole-improved tree-level Symanzik gauge action S_{g} , defined as: S_{g}=\frac{1}{3}\mathrm{Re}\sum_{x,\mu<\nu}\mathrm{Tr}\left[1-\beta\left(\frac{5}{3}\mathcal{P}^{U}_{\mu\nu}(x)-\frac{\mathcal{R}^{U}_{\mu\nu}(x)}{12u_{0}^{2}}\right)\right], where \displaystyle\mathcal{P}^{U}_{\mu\nu}(x) \displaystyle=U_{\mu}(x)U_{\nu}(x+a\hat{\mu})U^{\dagger}_{\mu}(x+a\hat{\nu})U^{\dagger}_{\nu}(x), \displaystyle\mathcal{R}^{U}_{\mu\nu}(x) \displaystyle=U_{\mu}(x)U_{\mu}(x+a\hat{\mu})U_{\nu}(x+2a\hat{\mu}) \displaystyle\quad\times U^{\dagger}_{\mu}(x+a\hat{\mu}+a\hat{\nu})U^{\dagger}_{\mu}(x+a\hat{\nu})U^{\dagger}_{\nu}(x), (2) and the tadpole improvement factor u_{0} is given by: u_{0}=\langle\mathrm{Re}\mathrm{Tr}\sum_{x,\mu<\nu}\mathcal{P}^{U}_{\mu\nu}(x)/(6N_{c}V)\rangle^{1/4} . Then we can have three definitions of g_{0} in Eq. ( II.1 ): 1) Naive definition: g^{(a)}_{0}=\sqrt{6/\beta} ; 2) Full tadpole improvement: Including u_{0} in both the action and also gauge link in the gauge fixing condition, and then Eq. ( II.1 ) should be rewritten into \displaystyle\Lambda(x)a^{2} \displaystyle=\sum_{\mu,\eta=\pm}\eta\left[\frac{\frac{U_{\mu}(x+\eta\frac{\hat{\mu}}{2}a)}{u_{0}}-\frac{U_{\mu}^{\dagger}(x+\eta\frac{\hat{\mu}}{2}a)}{u_{0}}}{2{\rm i}\sqrt{6/\beta/u_{0}^{4}}}\right]_{\text{Traceless}} \displaystyle=\sum_{\mu,\eta=\pm}\eta\left[\frac{U_{\mu}(x+\eta\frac{\hat{\mu}}{2}a)-U_{\mu}^{\dagger}(x+\eta\frac{\hat{\mu}}{2}a)}{2{\rm i}\sqrt{6/\beta}/u_{0}}\right]_{\text{Traceless}}. (3) Thus it leads to a effective gauge coupling g^{(b)}_{0}=\sqrt{6/\beta}/u_{0} ; 3) Approximation from u_{0} only: Using u_{0} only in gauge fixing while approximating \alpha_{s} via \alpha_{s}\simeq-\frac{4\mathrm{ln}u_{0}}{3.0684} Orginos and Toussaint 1999 which avoids to define g_{0} from the action, and similar procedure gives: g^{(c)}_{0}=\sqrt{-\frac{16\pi\mathrm{ln}u_{0}}{3.0684}}u_{0} . For the MILC ensemble a06m310 at a =0.0566 fm with m_{\pi} =310 MeV and \hat{\beta}=5/3\beta=6.72 , three definitions yield: g_{0}^{(a)}=1.2199,\ g_{0}^{(b)}=1.3768,\ g_{0}^{(c)}=1.2476 , respectively. In practice, we generate the random distribution \tilde{P}(\tilde{\Lambda}^{a}(x))=\frac{1}{\sqrt{2\pi\tilde{\xi}}}\exp\left\{-\frac{1}{2\tilde{\xi}}[\tilde{\Lambda}^{a}(x)]^{2}\right\} for the dimensionless quantity \tilde{\Lambda}\equiv g_{0}\Lambda a^{2} , meaning different g_{0} definitions correspond to different effective gauge-fixing parameters

Read the complete text