Driving Spiral Arms in the Circumstellar Disks of HD 100546 and HD 141569A
Résumé
With two-dimensional hydrodynamic simulations of disks perturbed externally by stars, brown dwarfs, or planets we investigate possible scenarios that can account for the spiral structure in circumstellar disks. We consider two scenarios: spiral structure driven by an external bound planet or low-mass star, and that excited by a previous stellar close encounter or flyby. We find that both scenarios produce a morphology similar to that observed in the outer disks of HD 141569A and HD 100546: moderately open two-armed outer spiral structure. Our simulations exhibit some trends. While bound prograde objects effectively truncate a disk, a close encounter by a star instead pulls out spiral arms, spreading out the disk. Following a flyby, the morphology of the spiral structure can be used to limit the mass of the perturbing object and the time since the encounter occurred. Eccentric bound planetary perturbers tend to clear gas away from the planet more efficiently, resulting in a reduction in the excited spiral amplitude. Bound perturbers in thicker disks excite more open but lower amplitude spiral structure. When the bound object has higher mass (stellar), the disk is truncated by the Roche lobe of the planet at periapse, and each time the companion approaches periapse, spiral arms may be pulled out from the disk. Thinner disks tend to exhibit more steeply truncated disk edges. We find that the outer two-armed spiral structure at radii greater than 300 AU observed in the disk of HD 141569A is qualitatively reproduced with tidal perturbations from its companion binary HD 141569B and HD 141569C on a prograde orbit near periapse. Our simulation accounts for the outer spiral arms, but is less successful than the secular model of Augereau and Papaloizou at matching the lopsidedness or asymmetry of the disk edge at 300 AU. Our simulations suggest that the disk has been previously truncated by the tidal force from the binary and has aspect ratio or thickness h/r<~0.1. The simulated disk also exhibits an enhanced density at the disk edge. We find that a bound object (stellar or planetary) is unlikely to explain the spiral structure in the disk of HD 100546. A coeval planet or brown dwarf in the disk of sufficient mass to account for the amplitude of the spiral structure would be detectable in NICMOS and Space Telescope Imaging Spectrograph images; however, existing images reveal no such object. A previous encounter could explain the observed structure, provided that the disk is thin, the mass of the perturbing star is greater than ~0.1 Msolar, and the encounter occurred less than a few thousand years ago. This suggests that the object responsible for causing the spiral structure is currently within a few arcminutes of the star. However, the USNO-B proper motion survey reveals no candidate object associated with HD 100546 or any moving object that could have recently encountered HD 100546. Furthermore, the probability that a field star encountered HD 100546 in the past few thousand years is very low.