arXiv subjects

Maarten Golterman

Publications and source records attributed to Maarten Golterman.

At least 127 records · Page 7Linked to original sources

arXiv2007

The Tunneling Hybrid Monte-Carlo algorithm

The hermitian Wilson kernel used in the construction of the domain-wall and overlap Dirac operators has exceptionally small eigenvalues that make it expensive to reach high-quality chiral symmetry for domain-wall fermions, or high precision in the case of the overlap operator. An efficient way of suppressing such eigenmodes consists of including a positive power of the determinant of the Wilson kernel in the Boltzmann weight, but doing this also suppresses tunneling between topological sectors. Here we propose a modification of the Hybrid Monte-Carlo algorithm which aims to restore tunneling between topological sectors by excluding the lowest eigenmodes of the Wilson kernel from the molecular-dynamics evolution, and correcting for this at the accept/reject step. We discuss the implications of this modification for the acceptance rate.

hep-lat↗

arXiv2007

Effective field theories for rooted staggered fermions

We extend the construction of the Symanzik effective action to include rooted staggered fermions, starting from a generalization of the renormalization-group approach to rooted staggered fermions. The Symanzik action, together with the usual construction of a partially quenched chiral effective theory from a local, partially quenched, fundamental theory, can then be used to derive the chiral effective theory. The latter reproduces rooted staggered chiral perturbation theory.

hep-lat↗

arXiv2007

Comment on "Chiral anomalies and rooted staggered fermions"

In hep-lat/0701018, Creutz claims that the rooting trick used in simulations of staggered fermions to reduce the number of tastes misses key physics whenever the desired theory has an odd number of continuum flavors, and uses this argument to call into question the rooting trick in general. Here we show that his argument fails as the continuum limit is approached, and therefore does not imply any problem for staggered simulations. We also show that the cancellations necessary to restore unitarity in physical correlators in the continuum limit are a straightforward consequence of the restoration of taste symmetry.

hep-lat↗

arXiv2006

Breakdown of staggered fermions at nonzero chemical potential

The staggered fermion determinant is complex when the quark chemical potential mu is nonzero. Its fourth root, used in simulations with dynamical fermions, will have phase ambiguities that become acute when Re mu is sufficiently large. We show how to resolve these ambiguities, but our prescription only works very close to the continuum limit. We argue that this regime is far from current capabilities. Other procedures require being even closer to the continuum limit, or fail altogether, because of unphysical discontinuities in the measure. At zero temperature the breakdown is expected when Re mu is greater than approximately half the pion mass. Estimates of the location of the breakdown at nonzero temperature are less certain.

hep-lat↗

arXiv2006

Quenched penguins, the Delta I=1/2 rule, and epsilon'/epsilon

The transformation properties of strong penguin operators under the action of the flavor group change when they are considered as operators in (partially) quenched QCD instead of the unquenched theory. An ambiguity arises, which is parameterized by new low-energy constants in the effective theory describing non-leptonic kaon decays in the (partially) quenched setting. Here we summarize results of the analysis for the complete set of three-flavor strong penguin operators, consisting of products of two left-handed flavor currents, or a left- and a right-handed current. Our results imply that (partially) quenched lattice computations of the Delta I=1/2 rule and epsilon'/epsilon are both affected by ambiguities intrinsic to the use of the quenched approximation at leading order in the chiral expansion. The only exception is the partially quenched case with three light sea quarks, consistent with general expectations. We also address the issue of quenched ambiguities in the case of an active charm, correcting and extending that in Phys. Rev. D 74, 014509 (2006).

hep-lat↗

arXiv2006

Regularizing QCD with staggered fermions and the fourth root trick

We investigate the properties of staggered-fermion lattice QCD in which the fourth root of the fermion determinant is taken. We show that this theory is non-local at non-zero lattice spacing $a$, and that the non-locality is caused by the breaking of taste symmetry at $a\ne 0$. We then present a renormalization-group based argument that the theory restores taste symmetry in the continuum limit. As a consequence the theory is local in that limit, and falls into the correct universality class. Finally, we argue that the correct effective theory for the physics of Goldstone bosons at $a\ne 0$ is given by staggered chiral perturbation theory with the replica trick.

hep-lat↗

arXiv2006

Why (staggered fermions)^{1/4} fail at finite density

Because the staggered fermion determinant is complex at nonzero mu, taking its fourth root leads to phase ambiguities. These unphysical effects cause the measure to become discontinuous; the problem becomes acute when Re mu exceeds approximately half the pion mass (when T>0 this rough bound probably moves towards larger mu). We show how to overcome the problem, but only very close to the continuum limit. This regime may be beyond reach with current resources.

hep-lat↗

arXiv2006

On the relation between low-energy constants and resonance saturation

Although there are phenomenological indications that the low-energy constants in the chiral lagrangian may be understood in terms of a finite number of hadronic resonances, it remains unclear how this follows from QCD. One of the arguments usually given is that low-energy constants are associated with chiral symmetry breaking, while QCD perturbation theory suggests that at high energy chiral symmetry is unbroken, so that only low-lying resonances contribute to the low-energy constants. We revisit this argument in the limit of large Nc, discussing its validity in particular for the low-energy constant L8, and conclude that QCD may be more subtle that what this argument suggests. We illustrate our considerations in a simple Regge-like model which also applies at finite Nc.

hep-ph↗

arXiv2006

Observations on staggered fermions at non-zero lattice spacing

We show that the use of the fourth-root trick in lattice QCD with staggered fermions corresponds to a non-local theory at non-zero lattice spacing, but argue that the non-local behavior is likely to go away in the continuum limit. We give examples of this non-local behavior in the free theory, and for the case of a fixed topologically non-trivial background gauge field. In both special cases, the non-local behavior indeed disappears in the continuum limit. Our results invalidate a recent claim that at non-zero lattice spacing an additive mass renormalization is needed because of taste-symmetry breaking.

hep-lat↗

arXiv2006

The operator product expansion does not imply parity doubling of hadrons

We examine whether the fact that QCD is chirally invariant at short distances necessarily leads to the prediction that hadrons form approximate parity doublets, as has been recently put forward by Glozman and collaborators. We show that this is not the case, and we exhibit some of the pitfalls of trying to link the operator product expansion to the hadron spectrum. We illustrate our arguments within a model for scalar and pseudo-scalar mesons used recently by Shifman to argue for parity doubling. We find that, whatever the experimental situation may be, there is no theoretical basis for parity doubling based on the use of the operator product expansion.

hep-ph↗

arXiv2005

Running couplings in equivariantly gauge-fixed SU(N) Yang--Mills theories

In equivariantly gauge-fixed SU(N) Yang--Mills theories, the gauge symmetry is only partially fixed, leaving a subgroup $H\subset SU(N)$ unfixed. Such theories avoid Neuberger's nogo theorem if the subgroup $H$ contains at least the Cartan subgroup $U(1)^{N-1}$, and they are thus non-perturbatively well defined if regulated on a finite lattice. We calculate the one-loop beta function for the coupling $\tilde{g}^2=ξg^2$, where $g$ is the gauge coupling and $ξ$ is the gauge parameter, for a class of subgroups including the cases that $H=U(1)^{N-1}$ or $H=SU(M)\times SU(N-M)\times U(1)$. The coupling $\tilde{g}$ represents the strength of the interaction of the gauge degrees of freedom associated with the coset $SU(N)/H$. We find that $\tilde{g}$, like $g$, is asymptotically free. We solve the renormalization-group equations for the running of the couplings $g$ and $\tilde{g}$, and find that dimensional transmutation takes place also for the coupling $\tilde{g}$, generating a scale $\tildeΛ$ which can be larger than or equal to the scale $Λ$ associated with the gauge coupling $g$, but not smaller. We speculate on the possible implications of these results.

hep-lat↗

arXiv2005

Double poles in Lattice QCD with mixed actions

We consider effects resulting from the use of different discretizations for the valence and the sea quarks, considering Wilson and/or Ginsparg--Wilson fermions. We assume that such effects appear through scaling violations that can be studied using effective-lagrangian techniques. We show that a double pole is present in flavor-neutral Goldstone meson propagators, even if the flavor non-diagonal Goldstone mesons made out of valence or sea quark have equal masses. We then consider some observables known to be anomalously sensitive to the presence of a double pole. We find that the double-pole enhanced scaling violations may turn out to be rather small in practice.

hep-lat↗

arXiv2005

Localization of lattice fermions: lessons for overlap

Lattice fermions in a fluctuating gauge field can show localization, much like electrons in a disordered potential. We study the spectrum of localized and extended states of supercritical Wilson fermions in gauge ensembles generated with plaquette and improved actions. When the Wilson fermion operator is used to construct the overlap kernel, the mobility edge, that is the boundary between the localized and extended states, determines the range of the kernel.

hep-lat↗

arXiv2005

Localization properties of lattice fermions with plaquette and improved gauge actions

We determine the location $λ_c$ of the mobility edge in the spectrum of the hermitian Wilson operator in pure-gauge ensembles with plaquette, Iwasaki, and DBW2 gauge actions. The results allow mapping a portion of the (quenched) Aoki phase diagram. We use Green function techniques to study the localized and extended modes. Where $λ_c>0$ we characterize the localized modes in terms of an average support length and an average localization length, the latter determined from the asymptotic decay rate of the mode density. We argue that, since the overlap operator is commonly constructed from the Wilson operator, its range is set by the value of $λ_c^{-1}$ for the Wilson operator. It follows from our numerical results that overlap simulations carried out with a cutoff of 1 GeV, even with improved gauge actions, could be afflicted by unphysical degrees of freedom as light as 250 MeV.

hep-lat↗

arXiv2005

Masslessness of ghosts in equivariantly gauge-fixed Yang--Mills theories

We show that the one-loop ghost self-energy in an equivariantly gauge-fixed Yang--Mills theory vanishes at zero momentum. A ghost mass is forbidden by equivariant BRST symmetry, and our calculation confirms this explicitly. The four-ghost self interaction which appears in the equivariantly gauge-fixed Yang--Mills theory is needed in order to obtain this result.

hep-lat↗

arXiv2005

The role of the double pole in lattice QCD with mixed actions

We investigate effects resulting from the use of different discretizations for the valence and the sea quarks in lattice QCD, considering Wilson and/or Ginsparg-Wilson fermions. We assume that such effects appear through scaling violations that can be studied using effective lagrangian techniques. We show that a double pole is present in flavor-neutral Goldstone meson propagators,even if the charged Goldstone mesons made out of valence quarks and those made out of sea quarks have equal masses. We then consider some observables known to be anomalously sensitive to the presence of a double pole. For these observables, we find that the double-pole enhanced scaling violations may turn out to be rather small in practice.

hep-lat↗

arXiv2005

Effective theory for quenched lattice QCD and the Aoki phase

We discuss the symmetries of quenched QCD with Wilson fermions, starting from its lagrangian formulation, taking into account the constraints needed for convergence of the ghost-quark functional integral. We construct the corresponding chiral effective lagrangian, including terms linear and quadratic in the lattice spacing. This allows us to study the phase structure of the quenched theory, and compare it to that in the unquenched theory. In particular we study whether there may be an Aoki phase (with parity and flavor spontaneously broken) or a first order transition line (with no symmetry breaking but meson masses proportional to the lattice spacing), which are the two possibilities in the unquenched theory. The presence of such phase structure, and the concomitant long-range correlations, has important implications for numerical studies using both quenched and dynamical overlap and domain-wall fermions. We argue that the phase structure is qualitatively the same as in the unquenched theory, with the choice between the two possibilities depending on the sign of a parameter in the low-energy effective theory.

hep-lat↗

arXiv2005

Mobility edge in lattice QCD

We determine the location $λ_c$ of the mobility edge in the spectrum of the hermitian Wilson operator on quenched ensembles. We confirm a theoretical picture of localization proposed for the Aoki phase diagram. When $λ_c>0$ we also determine some key properties of the localized eigenmodes with eigenvalues $|λ|<λ_c$. Our results lead to simple tests for the validity of simulations with overlap and domain-wall fermions.

hep-lat↗