arXiv subjects
Publications and source records attributed to UKQCD Collaboration.
We present details of our investigation of the Parallel Tempering algorithm. We consider the application of action matching technology to the selection of parameters. We then present a simple model of the autocorrelations for a particular parallel tempered system. Finally we present numerical results from applying the algorithm to lattice QCD with O(a)-improved dynamical fermions with twin ensemble systems
We present preliminary results for B_K, B_7^{3/2} and B_8^{3/2} from two high-statistics lattice computations. These calculations are performed at beta=6.0 and 6.2 in the quenched approximation, using mean-field-improved Sheikholeslami-Wohlert fermionic actions.
We present a calculation of f_B and of the form factors for the semi-leptonic decay B -> D l nu in the quenched approximation to QCD. Results are generated on lattices at beta = 6.2, using an O(a)-improved fermion action, with the clover coefficient determined non-perturbatively.
We present results from our study of the Parallel Tempering algorithm. We examine the swapping acceptance rate of a twin subensemble PT system. We use action matching technology in an attempt to maximise the swap acceptance rate. We model the autocorrelation times within Parallel Tempering ensembles in terms of autocorrelation times from Hybrid Monte Carlo. We present estimates for the autocorrelation times of the plaquette operator.
We propose a strategy for conducting lattice QCD simulations at fixed volume but variable quark mass so as to investigate the physical effects of dynamical fermions. We present details of techniques which enable this to be carried out effectively, namely the tuning in bare parameter space and efficient stochastic estimation of the fermion determinant. Preliminary results and tests of the method are presented. We discuss further possible applications of these techniques.
We argue that lattice simulations of full QCD with varying quark mass are best conducted at fixed lattice spacing rather than at fixed $β$. We present techniques which enable this to be carried out effectively, namely the tuning in bare parameter space and efficient stochastic estimation of the fermion determinant. Results and tests of the method are presented. We discuss other applications of such techniques.
We present the first results for the static quark potential and the light hadron spectrum using dynamical fermions at $β=5.2$ using an O(a) improved Wilson fermion action together with the standard Wilson plaquette action for the gauge part. Sea quark masses were chosen such that the pseudoscalar-vector mass ratio, m_PS/m_V$, varies from 0.86 to 0.67. Finite-size effects are studied by using three different volumes, 8^3\cdot 24, 12^3\cdot 24 and 16^3\cdot 24. Comparing our results to previous ones obtained using the quenched approximation, we find evidence for sea quark effects in quantities like the static quark potential and the vector-pseudoscalar hyperfine splitting.
The coupling $g_{B^\ast B π}$ is related to the form factor at zero momentum of the axial current between $B^\ast$- and $B$-states. This form factor is evaluated on the lattice using static heavy quarks and light quark propagators determined by a stochastic inversion of the fermionic bilinear. The $\gBBP$ coupling is related to the coupling $g$ between heavy mesons and low-momentum pions in the effective heavy meson chiral lagrangian. The coupling of the effective theory can therefore be computed by numerical simulations. We find the value $g = 0.42(4)(8)$. Besides its theoretical interest, the phenomenological implications of such a determination are discussed.
We study a new method -- maximal variance reduction -- for reducing the variance of stochastic estimators for quark propagators. We find that while this method is comparable to usual iterative inversion for light-light mesons, a considerable improvement is achieved for systems containing at least one infinitely heavy quark. Such systems are needed for heavy quark effective theory. As an illustration of the effectiveness of the method we present results for the masses of the ground state and excited states of $\bar{Q}q$ mesons and $\bar{Q}qq$ baryons. We compare these results with the experimental spectra involving $b$ quarks.
We present a quenched lattice calculation for the lowest lying $b \bar b g$-hybrid states in the framework of NRQCD using the leading order Hamiltonian up to ${\cal O}(mv^2)$. We demonstrate the existence of a nearly degenerate rotational band of states with an excitation energy approximately 1.6 GeV above the $Υ$ ground state. This lies around the $B \bar B_J^*$-threshold but well above the $B \bar B$-threshold. Therefore a heavy hybrid signal may well be detected if the centre-of-mass energy in B-factories is raised a few hundred MeV to coincide with other resonances above the 4S state. Our prediction is consistent with most phenomenological models and lattice calculations carried out in the static limit.
We describe the form factors for B to rho lepton neutrino and B to K* gamma decays with just two parameters and the two form factors for B to pi lepton neutrino with a further two or three parameters. The parametrizations are consistent with heavy quark symmetry, kinematic constraints and lattice results, which we use to determine the parameters. In addition, we test versions of the parametrizations consistent (or not) with light-cone sum rule scaling relations at q^2=0.
We present results of a quenched calculation of the heavy-light and quarkonium spectrum using the tadpole improved clover action. We resolve completely the triplet chi P-states in quarkonium systems, and obtain evidence for fine structure of the heavy-light P-states. Approximate scaling of the hyperfine splittings is observed, producing results that are significantly below experiment.
Computing quark propagators in lattice QCD is equivalent to solving large, sparse linear systems with multiple right-hand sides. Block algorithms attempt to accelerate the convergence of iterative Krylov-subspace methods by solving the multiple systems simultaneously. This paper compares a block generalisation of the quasi-minimal residual method (QMR), Block Conjugate Gradient on the normal equation, Block Lanczos and ($γ_5$-symmetric) Block BiConjugate Gradient.
We present the results of the first lattice study of semileptonic decays of baryons containing a b-quark. Predictions for the decay distributions are given and the Isgur-Wise functions for heavy baryons are computed, for values of the velocity transfer up to about omega=1.2. The computations are performed on a 24^3 x 48 lattice at beta=6.2 using the Sheikholeslami-Wohlert action in the quenched approximation.
We have studied the O(mv^6) effects in NRQCD on the spectrum of heavy quarkonia and compare our results for different lattices (quenched and dynamical). We also report on an investigation into hybrid states within the framework of NRQCD. This suggests that the lowest lying hybrid is around the B^* \bar B threshold and 3 standard deviations above the B \bar B.
We describe the form factors for semileptonic B to rho l nu and radiative B to K* gamma decays with just two parameters and the two form factors for semileptonic B to pi l nu decays with three parameters. The parametrizations are constrained by lattice results and are consistent with heavy quark symmetry, kinematic constraints and light cone sum rule scaling relations.
We explore the effect of higher order operators in the non-relativistic formulation of QCD (NRQCD). We calculated masses in the b \bar b spectrum using quenched gauge configurations at beta = 6.0 and two different NRQCD actions which have been corrected to order mv^4 and mv^6. The two-point functions are calculated in a gauge invariant fashion. We find the general structure of the spectrum to be the same in the two cases. Using the P - 3S1 splitting we determine the inverse lattice spacings to be 2.44(4) GeV and 2.44(5) GeV for the mv^4 and mv^6 actions, respectively. We do observe shifts in the spin splittings. The hyperfine splitting is reduced by approximately 4 MeV, while the fine splitting is down by up to 10 MeV, albeit with large statistical errors.
We compute the light hadron mass spectrum at beta=5.7 using the O(a) improved Sheikholeslami-Wohlert (SW) fermion action with two choices of the clover coefficient: the classical value, c=1, and a mean-field or tadpole-improved estimate c=1.57. We compare our results with those of the GF11 Collaboration who use the Wilson fermion action (c=0). We find that changing c from zero to 1 and 1.57 leads to significant differences in the masses of the chirally extrapolated and strange pseudoscalar and vector mesons, the nucleon, the Delta, and also in the Edinburgh plot. A number of other quantities, for example m_V^2-m_PS^2, J, am_K/am_ρand am_{K^*}/am_ρdo not appear to change significantly. We also investigate the effect of changing the lattice volume from approximately (2 fm)^3 to (2.6 fm)^3. We find that the meson masses are consistent to within one standard deviation and baryon masses are consistent to within two standard deviations.