The aim of the GRAL project is to simulate full QCD with standard Wilson fermions at light quark masses on small to medium-sized lattices and to obtain infinite-volume results by extrapolation. In order to establish the functional form of the volume dependence we study systematically the finite-size effects in the light hadron spectrum. We give an update on the status of the GRAL project and show that our simulation data for the light hadron masses depend exponentially on the lattice size.
We present details of simulations for the light hadron spectrum in quenched QCD carried out on the CP-PACS parallel computer. Simulations are made with the Wilson quark action and the plaquette gauge action on 32^3x56 - 64^3x112 lattices at four latt
ice spacings (a approx 0.1-0.05 fm) and the spatial extent of 3 fm. Hadronic observables are calculated at five quark masses (m_{PS}/m_V approx 0.75 - 0.4), assuming the u and d quarks being degenerate but treating the s quark separately. We find that the presence of quenched chiral singularities is supported from an analysis of the pseudoscalar meson data. We take m_pi, m_rho and m_K (or m_phi) as input. After chiral and continuum extrapolations, the agreement of the calculated mass spectrum with experiment is at a 10% level. In comparison with the statistical accuracy of 1-3% and systematic errors of at most 1.7% we have achieved, this demonstrates a failure of the quenched approximation for the hadron spectrum: the meson hyperfine splitting is too small, and the octet masses and the decuplet mass splittings are both smaller than experiment. Light quark masses are calculated using two definitions: the conventional one and the one based on the axial-vector Ward identity. The two results converge toward the continuum limit, yielding m_{ud}=4.29(14)^{+0.51}_{-0.79} MeV. The s quark mass depends on the strange hadron mass chosen for input: m_s = 113.8(2.3)^{+5.8}_{-2.9} MeV from m_K and m_s = 142.3(5.8)^{+22.0}_{-0} MeV from m_phi, indicating again a failure of the quenched approximation. We obtain Lambda_{bar{MS}}^{(0)}= 219.5(5.4) MeV. An O(10%) deviation from experiment is observed in the pseudoscalar meson decay constants.
The volume dependence of the octet baryon masses and relations among them are explored with Lattice QCD. Calculations are performed with n_f=2+1 clover fermion discretization in four lattice volumes, with spatial extent L ~ 2.0, 2.5, 3.0 and 3.9 fm,
with an anisotropic lattice spacing of b_s ~ 0.123 fm in the spatial direction, and b_t = b_s/3.5 in the time direction, and at a pion mass of m_pi ~ 390 MeV. The typical precision of the ground-state baryon mass determination is ~0.2%, enabling a precise exploration of the volume dependence of the masses, the Gell-Mann--Okubo mass relation, and of other mass combinations. A comparison of the volume dependence with the predictions of heavy baryon chiral perturbation theory is performed in both the SU(2)_L X SU(2)_R and SU(3)_L X SU(3)_R expansions. Predictions of the three-flavor expansion for the hadron masses are found to describe the observed volume dependences reasonably well. Further, the Delta-N-pi axial coupling constant is extracted from the volume dependence of the nucleon mass in the two-flavor expansion, with only small modifications in the three-flavor expansion from the inclusion of kaons and etas. At a given value of m_pi L, the finite-volume contributions to the nucleon mass are predicted to be significantly smaller at m_pi ~ 140 MeV than at m_pi ~ 390 MeV due to a coefficient that scales as ~ m_pi^3. This is relevant for the design of future ensembles of lattice gauge-field configurations. Finally, the volume dependence of the pion and kaon masses are analyzed with two-flavor and three-flavor chiral perturbation theory.
We present the final results of the CP-PACS calculation of the light hadron spectrum and quark masses with two flavors of dynamical quarks. Simulations are made with a renormalization-group improved gauge action and a mean-field improved clover quark
action for sea quark masses corresponding to $m_{rm PS}/m_{rm V} approx 0.8$--0.6 and the lattice spacing $a=0.22$--0.11 fm. For the meson spectrum in the continuum limit a clearly improved agreement with experiment is observed compared to the quenched case, demonstrating the importance of sea quark effects. For light quark masses we obtain $m_{ud}^{bar{MS}}(2GeV)=3.44^{+0.14}_{-0.22}$ MeV and $m_s^{bar{MS}}(2GeV)=88^{+4}_{-6}$ MeV ($K$-input) and $m_s^{bar{MS}}(2GeV)=90^{+5}_{-11}$ MeV ($phi$-input), which are reduced by about 25% compared to the values in quenched QCD.
Progress in computing the spectrum of excited baryons and mesons in lattice QCD is described. Large sets of spatially-extended hadron operators are used. The need for multi-hadron operators in addition to single-hadron operators is emphasized, necess
itating the use of a new stochastic method of treating the low-lying modes of quark propagation which exploits Laplacian Heaviside quark-field smearing. A new glueball operator is tested, and computing the mixing of this glueball operator with a quark-antiquark operator and multiple two-pion operators is shown to be feasible.
We compute the strange quark mass $m_s$ and the average of the $u$ and $d$ quark masses $hat m$ using full lattice QCD with three dynamical quarks combined with experimental values for the pion and kaon masses. The simulations have degenerate $u$ and
$d$ quarks with masses $m_u=m_dequiv hat m$ as low as $m_s/8$, and two different values of the lattice spacing. The bare lattice quark masses obtained are converted to the $msbar$ scheme using perturbation theory at $O(alpha_s)$. Our results are: $m_s^msbar$(2 GeV) = 76(0)(3)(7)(0) MeV, $hat m^msbar$(2 GeV) = 2.8(0)(1)(3)(0) MeV and $m_s/hat m$ = 27.4(1)(4)(0)(1), where the errors are from statistics, simulation, perturbation theory, and electromagnetic effects, respectively.