SciPost Submission Page
Constraining the gaugefixed Lagrangian in minimal Landau gauge
by Axel Maas
This is not the latest submitted version.
This Submission thread is now published as
Submission summary
Authors (as registered SciPost users):  Axel Maas 
Submission information  

Preprint Link:  https://arxiv.org/abs/1907.10435v1 (pdf) 
Date submitted:  20190807 02:00 
Submitted by:  Maas, Axel 
Submitted to:  SciPost Physics 
Ontological classification  

Academic field:  Physics 
Specialties: 

Approaches:  Theoretical, Computational 
Abstract
A continuum formulation of gaugefixing resolving the GribovSinger ambiguity remains a challenge. Finding a Lagrangian formulation of operational resolutions in numerical lattice calculations, like minimal Landau gauge, would be one possibility. Such a formulation will here be constrained by reconstructing the DysonSchwinger equation for which the lattice minimalLandaugauge ghost propagator is a solution. It is found that this requires an additional term. As a byproduct new, high precision lattice results for the ghostgluon vertex in three and four dimensions are obtained.
Current status:
Reports on this Submission
Anonymous Report 1 on 20191226 (Invited Report)
 Cite as: Anonymous, Report on arXiv:1907.10435v1, delivered 20191226, doi: 10.21468/SciPost.Report.1414
Strengths
1 Well written
2 Problem addressed clearly stated
3 Study how to compare different nonperturbative techniques results for QCD
4 New way to address the comparison
Weaknesses
1 Does not include a full discussion of all issues
2 Biased discussion at the end
Report
Axel Maas
"Constraining the gaugefixed Lagrangian in minimal Landau gauge"
arXiv:1907.10435
The manuscript under appreciation discuss an important issue for the nonperturbative regime of QCD, the comparison between lattice simulations and DysonSchwinger (DS) results. The manuscript is well written and the problem is clearly stated. After a general introduction, the author writes the generating functional for the continuum formulation of the minimal Landau gauge, that is “identified” with the continuum version of the lattice formulation, and the usual FaddeevPopov (FP) functional. Then the FP is used to write the DysonSchwinger equation for the ghost propagator that involves the gluon and ghost propagators and ghostgluon vertex. This is one of the simplest DS equations and the author uses it as a benchmark to compare lattice and DS results.
(i) Although the author provides a good discussion of various issues, he forgets to mention that on the lattice, if one excludes the corresponding perturbative treatment, there are no ghost fields and the ghost propagator is “recovered” mimicking the usual continuum procedure. To the referee this is an important point that the author should mention in Sec. 2 to alert the reader.
The results of 3D and 4D lattice simulations, for various lattice spacings and volumes, are used in the ghost DS equation to check if the equations is solved exactly. The full exercise requires the computation of the gluon and ghost propagators, together with the lattice computation of the ghostgluon vertex that is performed for three kinematical configurations. In all the computations large statistical ensembles are used. The manuscript includes a nice discussion of the results obtained for the various Green functions and of the deviations between lattice and continuous calculations.
(ii) However, the manuscript does not discuss in detail the effects of discretisation effects on the ghostgluon vertex and on the propagators. In particular, the propagators are not shown at all and they should be reported. This is of at most importance as it is well known [O. Oliveira, P. J. Silva, Phys. Rev. D86 (2012) 114513] that for the propagators the lattice spacing and the lattice volume should be properly chosen. A possible explanation of the observed differences could be due, at least partially, for the choice of lattice spacings and volumes where the simulations are performed
(iii) The finite volume effects, being due to the lattice spacing and to the lattice volume, are not discussed properly. This is also important as the number of lattice spacings and volumes reported on the manuscript varies significantly. For example, looking at the 3D simulations with lattice sizes of ~2 fm or ~ 3.9 fm, or 4D simulations with lattice sizes ~4.9 fm or ~9.8 fm, one can see clear differences between the various data sets for the various kinematical configurations. The grouping together of all the simulations data on Figs. 5 and 6 only makes their reading quite difficult. My suggestion goes to add the results for the same volumes separately. This can be done either as an inset on the figs. or by adding extra plots.
(iv) The section that worries me most being the Summary. I completely disagree with the authors summary. First, given the finite volume effects I hardly agree with what is written there. At most, would say that this study suggests that the comparison should be done with care and further studies are need before doing any conclusion. I cannot agree, in any way, with the second paragraph of the Summary. I recall the author that there are other interpretations [M. Tissier, Phys. Lett. B784 (2018) 146] and although the efforts associated with the scaling solution are an important work and contribution to the understanding of QCD, there is no evidence of these type of solution coming from lattice [probably the closest being A. Sternbeck, M. MuellerPreuskker, Phys. Lett. B 726 (2013) 396].
The author should include all the above mentioned references in the revision.
In summary, I do think that the manuscript contains new and important contributions that can help in the understanding of QCD. It certainly, deserves to be published but I cannot agree with the present form of the manuscript and ask the author to review it.
Requested changes
See the report.
Author: Axel Maas on 20200127 [id 721]
(in reply to Report 1 on 20191226)Dear referee,
I am grateful for the constructive referee report, and have made the following changes:
I have added this in the last paragraph of section 2.
I have added another appendix with the results for the propagators. I have also extended the discussion concerning systematic artifacts.
I have added corresponding plots in the style of the main text figures 13 to compare fixed discretization or volume for two particular momentum configurations. I have also extended the discussion concerning artifacts in the whole manuscript.
I have toned down the corresponding statement, and emphasized the need for more systematics.
I have rewritten the entire second paragraph, and quite a bit of the whole summary, completely. All mention of the scaling solution are now clearly marked as speculation in situations not addressable in current lattice calculations, and the current situation, that a screened behavior in all current lattice calculations (in 3d and 4d) is unambiguously established, is fully appreciated. However, until we are able to do the same calculations in the continuum and the lattice, and especially the HirschfeldFujikawatypegauge has been done on the lattice, I cannot agree that this is already fully understood. Unfortunately, we still do not have yet any gaugefixing procedure, which resolves the GribovSinger ambiguity, and is simultaneously practically implementable both in continuum calculation and on the lattice, at least as far as I currently can see. E.g. the one of Tissier you mention has a sign problem (due to the factors s(i), and also the counting of unstable directions is not trivial) on the lattice (which is also true for the HirschfeldFujikawa gauge).
I still have considerably extended the whole discussions, and hopefully have made clearly distinct what is pretty certain  the situation in minimal Landau gauge on the lattice and the existence of screened solutions to the functional equations in truncations  and those which is speculative  everything concerning the scaling solution and what happens in unmodified HirschfeldFujikawa gauge. In particular that no lattice calculation provides any indication of a scaling solution and that such a result is also not expected in any kind of lattice simulations inside the first Gribov region. Still, I think the possibilities beyond that what is currently possible on the lattice deserves mention.
I apologize if the original version created the impression that this issue was settled.
As requested, I have added the provided references, alongside several further ones.
I have also found an error in figures 2 and 3 with respect to the compared lattice parameters, and fixed it. I also added a third comparison, to highlight even further systematic effects.