Introduction

Asteroid families are groupings of minor planets identified on the basis of their proper orbital elements (Zappala' et al. 1994, 1995).  They are probably remnants of larger parent bodies that were collisionally disrupted hundreds of Myr to a few Gyr ago, with the smallest fragments traveling the furthest from the cluster's center (Cellino et al. 1999).  One problem with this scenario is that the ejection velocities inferred for those fragments from their observed dispersions in their current proper orbital elements seem to be consistently higher than those predicted by collisions in laboratory experiments (Fujiwara et al. 1989) and hydrocode models (Benz and Asphaug 1999).

Here we show results of long-term simulations on the Adeona, Gefion (old Ceres), and Dora families.  The purpose of these simulations is to gain understanding on how asteroid families can dynamically evolve over long time scales.   There are essentially three effects that can change proper elements:

  1. Yarkovsky effect.
  2. Resonances.
  3. Gravitational scattering by large asteroids like 1 Ceres, 2 Pallas, 4 Vesta, and 10 Hygiea
While resonances can account for motion in proper e, and I, only Yarkovsky effect and encounters with large asteroids can account for mobility in proper a.   The Yarkovsky effect is size dependent, and is dominating the dispersion in a of smaller asteroids (a 20 km regolith-covered body would experience a drift of da/dt~6*10-6 AU My-1).   For D< 20 km, it is the dominant factor in semimajor axis mobility of family members.   At larger sizes, a mobility is mainly provided by close encounters with larger asteroids.
This mechanism, which is the main topic of this research, has been for long considered ineffective, and very few studies have been carried out on the long-term effects of close encounters (see Flora's article by Nesvorny et al.).  In this work we investigate the effect of close encounters with 1 Ceres, 2 Pallas, 4 Vesta, and 10 Hygiea for members of the Adeona and Gefion family (we used the files from Cellino and took only well identified members [classes QC=2,3])    We also generated synthetic families which were assumed to be initially more tightly clustered in the orbital elements space than the observed families.   This was done in order to test the hypothesis that asteroid family start with a more compact initial distribution in orbital elements (i.e. lower ejection velocities) and are then scattered by the three mechanism of above.  We typically assumed an initial dispersion of a family consistent with an isotropic ejection velocity field not exceeding 100 m/s at infinity (for a description of the algorithm to generate synthetic families see David Nesvorny site).  We generated synthetic families with centers in the current center of Adeona, Gefion and Dora, and we submitted their members to the action of close encounters and Yarkovsky (SWIFT-SKEEL and SWIFT-RMVSY codes, see next section).   We used the same values of thermal parameters for C-types families (Adeona and Dora)  and S-type family (Gefion) used by Mira Broz  and David Vokrouhlicky (see yarko site). Our goal was to estimate the long term effect of close encounters on semimajor axis mobility, to compare its effectiveness with the mobility from Yarkovsky effect, and to try to obtain an estimate of the ages of the Adeona and Gefion family.   In the following table we report a list of our simulations, with their integration length and step size.  Except for the first simulation, with current
members of Adeona, for which we consider all the planets from Venus to Neptune, all simulations used the OSS planets (plus the four largest asteroids for the simulations with SWIFT-SKEEL).
 
 
Simulation: Integration length t [Myr] Step size dt [day] propagator 
-Adeona:  Real Members  500 7 SWIFT-SKEEL
-Adeona:  Original distribution 630 18 SWIFT-SKEEL
-Adeona:  Original distribution 880 20 SWIFT-RMVSY
-Gefion:    Real Members 580 18 SWIFT-SKEEL
-Gefion:    Original distribution 500 18 SWIFT-SKEEL
-Gefion:    Original distribution 770 20 SWIFT-RMVSY
-Dora:       Original distribution 220 18 SWIFT-SKEEL

Back to index

Integrators

We used two integrators for our simulations.   Close encounters are notoriously "tough" to account for:  for our simulations with large asteroids we used SWIFT-SKEEL, from H. Levison and M. Duncan, that is able to symplectically integrate close encounters between a massless particle and a massive body.  This algorithm combines a variant of the standard mixed-variable symplectic method (MVS, Wisdom and Holman 1991) with an improved version of the multiple time-step method originally developed by Skeel and Biesiadecki (1994), (see also Biesiadecki and Skeel 1993).  Further details can be found in Duncan et al. (1998).    For the integrations with the Yarkovsky effect we used SWIFT-RMVSY of Mira Broz, which is based on SWIFT-RMVS3 of Levison and Duncan and include the Yarkovsky effect (both diurnal and seasonal) as a gravitational perturbation of the velocities (for further detail, see the yarko-site).

Back to index