The QCD spectrum with three quark flavors

TL;DR

The paper investigates how dynamical quarks of three flavors (2 light + strange) modify the hadron spectrum relative to quenched QCD by using matched lattices with $a \approx 0.13$ fm and improved actions. It reports that full QCD improves agreement with experiment in several sectors, notably through changes in the static potential, the J parameter, and the behavior of the $a_0$ meson, which couples to two-meson states. The work highlights the importance of sea quarks for accurate spectroscopy and establishes a framework for continuum and chiral extrapolations with finer lattices. These results motivate further simulations at smaller lattice spacings to quantify discretization effects and to enable decay-rate and excited-state analyses in a controlled setting.

Abstract

We present results from a lattice hadron spectrum calculation using three flavors of dynamical quarks - two light and one strange, and quenched simulations for comparison. These simulations were done using a one-loop Symanzik improved gauge action and an improved Kogut-Susskind quark action. The lattice spacings, and hence also the physical volumes, were tuned to be the same in all the runs to better expose differences due to flavor number. Lattice spacings were tuned using the static quark potential, so as a byproduct we obtain updated results for the effect of sea quarks on the static quark potential. We find indications that the full QCD meson spectrum is in better agreement with experiment than the quenched spectrum. For the 0++ (a0) meson we see a coupling to two pseudoscalar mesons, or a meson decay on the lattice.

Figures (20)

• Figure 1: The effect of the conjugate gradient error used in the updating on the plaquette in a three flavor run with quark masses $0.4 m_s$ and $m_s$.
• Figure 2: The effect of the conjugate gradient error used in the updating on $\bar{\psi}\psi$ in the same three flavor run.
• Figure 3: The effect of the conjugate gradient error used in the updating on the Goldstone pion mass in the same three flavor run.
• Figure 4: $\langle\bar{\psi}\psi\rangle$ versus the squared step size. The left hand panel is an algorithm test done on a $12^4$ lattice using the one-plaquette gauge action and conventional quark action at $6/g^2=5.10$ with three quark flavors with mass $am_q=0.02$. The octagons use one pseudofermion field with a factor of $3/4$ in the force term, appropriate for three flavors, while the squares use the $2+1$ flavor code, with separate fermion force terms for one and two flavors, but with the same mass for both terms. The right hand panel shows $\langle\bar{\psi}\psi\rangle$ using improved gauge and quark actions at $a \approx 0.14$ fm. ($10/g^2=6.80$ and $am_{u,d}=am_s=0.05$).
• Figure 5: Effects of dynamical quarks on the shape of the potential. Here we plot $r_0 \sqrt{\sigma}$ as a function of the quark mass. The two quenched points are at the far right, with the octagon coming from the $10/g^2=8.0$ run and the cross from the $10/g^2=8.4$ run, which has a lattice spacing of about 0.09 fm. The remaining octagons are full QCD runs with three degenerate flavors, and the squares are full QCD runs with two light flavors and one heavy. The diamond is the two flavor run, and the burst at the left is a linear extrapolation of the $2+1$ results to the physical value of $(m_\pi/m_\rho)^2$.