Spectroscopy of red giants in the large magellanic cloud bar: Abundances, kinematics, and the age-metallicity relation

We report metallicities and radial velocities derived from spectra at the near-infrared calcium triplet for 373 red giants in a 200 arcmin(2) area at the optical center of the LMC bar. These are the first spectroscopic abundance measurements of intermediate-age and old field stars in the high surface brightness heart of the LMC. The metallicity distribution is sharply peaked at the median value [Fe/H] = -0.40, with a small tail of stars extending down to [Fe/H] <= -2.1; 10% of the red giants are observed to have [Fe/H] <= -0.7. The relative lack of metal-poor stars indicates that the LMC has a "G dwarf" problem, similar to the Milky Way. The abundance distribution can be closely approximated by two Gaussians containing 89% and 11% of the stars, respectively: the first component is centered at [Fe/H] = -0.37 with sigma = 0.15, and the second at [Fe/H] = -1.08 with sigma = 0.46. The dominant population has a metallicity distribution similar to that of the LMC's intermediate-age star clusters. The mean heliocentric radial velocity of the sample is 257 km s(-1), corresponding to the same center-of-mass velocity as the disk ( measured at larger radii). Because of the central location of our field, kinematic constraints are not strong, but there is no evidence that the bar deviates from the general motion of the LMC disk. The velocity dispersion of the whole sample is sigma(v) = 24.7 +/- 0: 4 km s(-1). When cut by metallicity, the most metal-poor 5% of stars ([Fe/H] < -1.15) show sigma(v) = 40.8 +/- 1.7 km s(-1), more than twice the value for the most metal-rich 5%; this suggests that an old, thicker disk or halo population is present. The age-metallicity relation (AMR) is almost flat during the period from 5 to 10 Gyr ago, with an apparent scatter of +/- 0.15 dex about the mean metallicity for a given age. Comparing this to chemical evolution models from the literature, we find that a burst of star formation 3 Gyr ago does not reproduce the observed AMR more closely than a steadily declining star formation rate. The AMR suggests that the epoch of enhanced star formation, if any, must have commenced earlier, approximate to 6 Gyr ago - the exact time is model dependent. We compare the properties of the LMC and the Milky Way, and discuss our results in the context of models that attempt to use tidal interactions with the Milky Way and Small Magellanic Cloud to explain the star and cluster formation histories of the LMC.