PARTICLE-TRANSPORT AND THE LAMBDA BOOTIS PHENOMENON .2. AN ACCRETION DIFFUSION-MODEL

We present a large set of numerical calculations describing the surfac e and internal abundance evolution in a star accreting metal-depleted material from the interstellar medium. The transport model includes co ntributions from chemical separation mechanisms, rotationally induced meridional circulation, and accretion. Calculations are performed in a T(eff) = 8000 K main-sequence model, for a few representative chemica l species. While the main purpose of the paper is to put the accretion /diffusion model for lambda Bootis stars on a firmer quantitative foot ing, many of the conclusions drawn are directly transferable to accret ion on '' normal '' main-sequence A stars. Our computations demonstrat e the following: (1) the maintenance of the abundance signature of the accreted material is only possible for accretion rates larger than fe w 10(-14) M. yr-1, the exact value being function of the chemical elem ent under consideration. (2) Upon terMination of the accretion episode , chemical separation destroys any surface abundance pattern set up ea rlier by accretion, and does so in as little as 10(6) yr. (3) For some elements, however, a certain level of memory of the accretion phase p ersists to late epochs, in that the asymptotic overabundance levels ar e smaller than if accretion had not occurred. (4) For stars with equat orial rotational velocities of 125 km s-1 or less, and accreting at a rate of 10(-13) M. yr-1, rotationally induced meridional circulation h as no significant influence on the evolution of surface abundances, du ring the accretion episode itself. But upon termination of accretion, it further contributes in erasing the effects of the accretion episode . While the present computations lend further support to the accretion /diffusion model proposed recently for A Bootis stars, they also estab lish a number of constraints on the model. In particular, our results show that if the peculiar abundance pattern characterizing the A Booti s phenomenon is due to the differential accretion of metal-depleted ma terial, then we are not seeing the residual signature of a long-gone a ccretion episode, such as during star formation itself or early on the pre-main-sequence. Instead, accretion must be ongoing or have ceased within the last 10(6) yr or so. This suggests that nearly all A Bootis stars should show observational evidence for the presence of circumst ellar material.