arXiv subjects

R. G. Edwards

Publications and source records attributed to R. G. Edwards.

44 records · Page 3Linked to original sources

arXiv1996

Instabilities and Non-Reversibility of Molecular Dynamics Trajectories

The theoretical justification of the Hybrid Monte Carlo algorithm depends upon the molecular dynamics trajectories within it being exactly reversible. If computations were carried out with exact arithmetic then it would be easy to ensure such reversibility, but the use of approximate floating point arithmetic inevitably introduces violations of reversibility. In the absence of evidence to the contrary, we are usually prepared to accept that such rounding errors can be made small enough to be innocuous, but in certain circumstances they are exponentially amplified and lead to blatantly erroneous results. We show that there are two types of instability of the molecular dynamics trajectories which lead to this behavior, instabilities due to insufficiently accurate numerical integration of Hamilton's equations, and intrinsic chaos in the underlying continuous fictitious time equations of motion themselves. We analyze the former for free field theory, and show that it is essentially a finite volume effect. For the latter we propose a hypothesis as to how the Liapunov exponent describing the chaotic behavior of the fictitious time equations of motion for an asymptotically free quantum field theory behaves as the system is taken to its continuum limit, and explain why this means that instabilities in molecular dynamics trajectories are not a significant problem for Hybrid Monte Carlo computations. We present data for pure $SU(3)$ gauge theory and for QCD with dynamical fermions on small lattices to illustrate and confirm some of our results.

hep-lat↗

arXiv1996

QCD $β$ Function with Two Flavors of Dynamical Wilson Fermions

We test the asymptotic scaling behavior of state-of-the-art simulations of QCD with two flavors of light Wilson fermions. This is done by matching $π$ and $ρ$ masses on lattices of size $16^3\times32$ and $8^3\times16$. We find that at $β=6/g^2=5.3$ matching is not possible over a range extending down to $β=3.5$. The large lattice data at $β=5.5$ matches the small lattice values at $β=4.9(1)$ leading to a shift $Δβ=0.6(1)$, considerably larger than the perturbative prediction of $0.45$. In both cases we conclude that the simulations are very far from the asymptotic scaling region.

hep-lat↗

arXiv1994

Toward the QCD $β$ Function with Dynamical Wilson Fermions

We present data for the scaling behavior of lattice QCD with two flavors of light Wilson fermions. This is done by matching $π$ n and $ρ$ masses at the two lattice sizes of $16^3\times32 $ and $8^3\times16$. We find that at $β=6/g^2\equiv5.3$ the matching does not occur over a range extending down to $β=3.5$. For $β=5.5$ matching may be achieved at $β=4.8-4.9$, leading to a $Δβ=0.6-0.7$ which is higher than the perturbative 2-loop value of 0.45. In both cases we conclude that the simulations are very far from the perturbative scaling region.

hep-lat↗

arXiv1994

The heavy quark potential in QCD with 2 flavors of dynamical quarks

We compute the heavy quark potential on configurations generated by the HEMCGC collaboration with dynamical staggered fermions at $6/g^2 = 5.6$ and with dynamical Wilson fermions at $6/g^2 = 5.3$. The computations are done on $16^3 \times 32$ lattices, corresponding to physical sizes of about 1.6 and 2.3 fm, respectively. Up to the distances probed no sign of string breaking is detectable. We also compute the recently proposed scale $r_0$ defined by $r_0^2 F(r_0) = 1.65$.

hep-lat↗

arXiv1992

Possible Failure of Asymptotic Freedom in Two-Dimensional $RP^2$ and $RP^3$ $σ$-Models

We have simulated the two-dimensional $RP^2$ and $RP^3$ $σ$-models, at correlation lengths up to about 220 (resp.\ 30), using a Wolff-type embedding algorithm. We see no evidence of asymptotic scaling. Indeed, the data rule out the conventional asymptotic scaling scenario at all correlation lengths less than about $10^{9}$ (resp.\ $10^{5}$). Moreover, they are consistent with a critical point at $β\approx 5.70$ (resp.\ 6.96), only 2\% (resp.\ 5\%) beyond the largest $β$ at which we ran. Preliminary studies of a mixed $S^{N-1}/RP^{N-1}$ model (i.e. isovector + isotensor action) show a similar behavior when $β_T \to \infty$ with $β_V$ fixed $\ltapprox 0.6$, while they are consistent with conventional asymptotic freedom along the lines $β_T/β_V$ fixed $\ltapprox 2$. Taken as a whole, the data cast doubt on (though they do not completely exclude) the idea that $RP^{N-1}$ and $S^{N-1}$ $σ$-models lie in the same universality class.

hep-lat↗

arXiv1992

Hadron Spectrum in QCD with Valence Wilson Fermions and Dynamical Staggered Fermions at $6/g^2=5.6

We present an analysis of hadronic spectroscopy for Wilson valence quarks with dynamical staggered fermions at lattice coupling $6/g^2 = β=5.6$ at sea quark mass $am_q=0.01$ and 0.025, and of Wilson valence quarks in quenched approximation at $β=5.85$ and 5.95, both on $16^3 \times 32$ lattices. We make comparisons with our previous results with dynamical staggered fermions at the same parameter values but on $16^4$ lattices doubled in the temporal direction.

hep-lat↗

arXiv1992

On the Dynamics of Light Wilson Quarks

We describe recent results obtained as part of the High Energy Monte Carlo Grand Challenge (HEMCGC) project concerning the behaviour of lattice QCD with light dynamical Wilson quarks. We show that it is possible to reach regions of parameter space with light pions $m_π\ll0.2/a$, but that the equilibration time for such a system is at least of the order of 1,000 unit-length Hybrid Monte Carlo (HMC) trajectories (about a Gigaflop/sec-year). If the Hybrid Molecular Dynamics (HMD) algorithm is used with the same parameters it gives incorrect results.

hep-lat↗

arXiv1992

Dynamic Critical Behaviour of Wolff's Algorithm for $RP^N$ $σ$-Models

We study the performance of a Wolff-type embedding algorithm for $RP^N$ $σ$-models. We find that the algorithm in which we update the embedded Ising model à la Swendsen-Wang has critical slowing-down as $z_χ\approx 1$. If instead we update the Ising spins with a perfect algorithm which at every iteration produces a new independent configuration, we obtain $z_χ\approx 0$. This shows that the Ising embedding encodes well the collective modes of the system, and that the behaviour of the first algorithm is connected to the poor performance of the Swendsen-Wang algorithm in dealing with a frustrated Ising model.

hep-lat↗