Scientific Program6th Castastrophic Disruption Worskhop

Cannes, June 9-11 2003

General ScheduleMorning sessions:08:30-10:00 Session I10:00-10:30 Coffee Break10:30-12:00 Session IILunch on the beach:12:15-14:15Afternoon sessions:14:30-16:00 Session III16:00-16:30 Coffee Break16:30-18:00 Session IV

SUNDAY JUNE 8 (19:30): Welcoming Buffet on the beach

MONDAY JUNE 9 (morning)

Asteroid Families

• Cellino (IR): Asteroid families

• Binzel: A possible differentiated family

• Bus: Constraining the compositions of silicate-rich asteroid families throughnear-IR spectroscopy

• Dell’Oro, et al.: Structure and age of the asteroid families: Constraintsfrom observational data

• Morbidelli, et al.: The shallow magnitude distribution of asteroid families

MONDAY JUNE 9 (afternoon)

Impact Experiments

• Nakamura (IR): Laboratory experiments on impact disruption

• Housen and Holsapple (IR): Scaling laws of collision fragmentation

• Burchell and Lowen: Impact studies of catastrophic disruption of ice spheresat the University of Kent hypervelocity impact laboratory

• Flynn, et al.: Catastrophic Disruption of meteorites: Implications for as-teroids and interplanetary dust

1

TUESDAY JUNE 10 (morning)

Observations

• Harris and Pravec (IR): Spins and satellites: Probes of asteroid interiors

• Weissman (IR): Disruption and splitting of cometary nuclei

• Fulchignoni, et al.: Toward a taxonomy of the Edgeworth-Kuiper belt ob-jects: The multivariable approach

• Howell, et al.: Observations of 2002 NY40: An ordinary chondrite?

• Merline, et al.: Progress on the search for binary asteroids using adaptiveoptics

• Ryan and Ryan: Discovery of binary system within the Vesta family as-teroids

• Nakamura, et al.: Ejecta size-velocity relationship derived from the distri-bution of secondary craters of small, km-sized crater on Mars

TUESDAY JUNE 10 (afternoon)

Inferences of internal structure

• Consolmagno (IR): Meteoritical evidence and constraints on impacts anddisruption

• Holsapple and Housen (IR): Asteroid rotations, implications for internalstructure and disruption

• Nolan: Variety of near-Earth asteroids

• Levasseur-Regourd: Physical properties of NEOs surfaces: What can belearned from asteroid regolith observations and microgravity laboratorysimulations

20:00 Banquet in the hold harbor of Cannes

WEDNESDAY JUNE 11 (morning)

Numerical Models

• Michel (IR): Numerical models and simulations of small body collisions

• Benz (IR): Catastrophic disruption threshold and internal structure

2

• Durda, et al.: The formation of asteroid satellites in catastrophic colli-sions: Results from numerical simulations

Rubble Piles and Monoliths

• Richardson (IR): Rubble piles and monoliths

• Roig: Interacting ellipsoids: A minimal model for the dynamics of rubble-pile asteroids

WEDNESDAY JUNE 11 (afternoon)

Collisional Models and Size Distributions

• Marzari (IR): On modelling the collisional evolution of minor body popu-lations

• Durda: Wavy size distributions caused by discontinuities in size-strengthscaling laws

• O’Brien and Greenberg: Analytical and numerical modelling of asteroidcollisional evolution: Recent results

• Penco, et al.: Yarkovsky depletion and asteroid collisional evolution

• Bottke, et al.: The fossilized size distribution of the main asteroid belt

End of The CDVI Workshop (That’s all Folks!)

3

BENZ W. wbenz@phim.unibe.chBINZEL R.P. rpb@mit.eduBOTTKE W.F. bottke@boulder.swri.eduBURCHELL M.J. M.J.Burchell@ukc.ac.ukBUS S.J. sjb@ifa.hawaii.eduCAMPO BAGATIN A. adriano@disc.ua.esCELLINO A. cellino@to.astro.itCERRONI P. priscio@rm.iasf.cnr.itCONSOLMAGNO G.J. gjc@specola.vaDAVIS D. drd@psi.eduDELL’ORO A. delloro@to.astro.itDIMARTINO M. dimartino@to.astro.itDURDA D.D. durda@boulder.swri.eduFLYNN G.J. george.flynn@plattsburgh.eduFOGLIA S. Sergio Foglia@saima.itFULCHIGNONI M. Marcello.Fulchignoni@obspm.frHARRIS A.W. harrisaw@colorado.eduHOLSAPPLE K.A. holsapple@aa.washington.eduHOUSEN K.R. kevin.r.housen@boeing.comHOWELL E.S. ehowell@naic.eduLA SPINA A. laspina@dm.unipi.itLEVASSEUR-REGOURD A.C. Chantal.Levasseur@aerov.jussieu.frLEVISON H. hal@obs-nice.frMARZARI F. marzari@pd.infn.itMERLINE W.J. merline@boulder.swri.eduMICHEL P. michel@obs-nice.frMORBIDELLI A. morby@obs-nice.frNAKAMURA A. amnakamu@kobe-u.ac.jpNOLAN M. nolan@naic.eduO’BRIEN D.P. obrien@pirlserver.lpl.Arizona.EDUPAOLICCHI P. paolo@df.unipi.itPRAVEC P. ppravec@asu.cas.czRICHARDSON D.C. dcr@astro.umd.eduROIG F. froig@on.brRYAN E.V. eryan@admin.nmt.eduRYAN W.H. bryan@physics.nmhu.eduTANGA P. tanga@obs-nice.frWEISSMAN P. pweissman@lively.jpl.nasa.govWOLTERS S.D. S.D.Wolters@open.ac.uk

List of Participants

4

CATASTROPHIC DISRUPTION THRESHOLDAND INTERNAL STRUCTURE

Invited Review

Author: Willy BenzInstitute: University of Berne (Switzerland)

Recent numerical simulations of the young Karin asteroid family formation(Michel etal, 2002, Nature, 4212, 608) have pinpointed the important role playedby the internal structure of the parent body in determining the outcome of acatastrophic collision. Further studies over a wider range of impact conditionsare confirming that the size distribution and velocity distribution of fragmentsfollowing a catastrophic disruption in the gravity regime changes dramaticallybetween a parent body modeled as an elastic solid, a pre-shattered body or arubble pile. We have therefore undertaken a study in order to determine thecritical disruption threshold for rocky bodies as a function of size and internalstructure.

5

A POSSIBLE DIFFERENTIATED FAMILY

Author: R.P. BinzelInstitute: MIT, Cambrige (USA)

While numerous asteroid families are known, it is surprising that all appearto be remnants of relatively homogeneous parent bodies. Certainly the exis-tence of iron meteorites from many dozens of parent bodies tells us that manydifferentiated objects have been catastrophically disrupted - yet we do not seeany families comprised by highly differentiated members down to the size limitcurrently searched (typically 10-30 km).

The search for differentiated families has now been pushed down to the sizelimit of a few km within a very unlikely place: Mars crossers. Mars-crossingasteroids are expected to lose information on their place of origin on accountof their dynamical interactions with the planet and the resonances that drivethem into Mars-crossing orbits. Yet spectroscopic measurements of ≈ 100 Mars-crossing objects have revealed a small concentration of ≈ 5 km objects having A-and Sa-type spectra, interpreted to be indicative of olivine. (Olivine is the mostabundant constituent in the mantle of a differentiated parent body and thusa key “tracer” for finding a differentiated family). This group of four A- andSa-type asteroids at 2.20 AU is just barely within the Mars-crossing populationhaving average eccentricity values of 0.25 and inclinations near 7 degrees. Beingat the edge of the Mars-crossing population may explain why this grouping (ifreal) is recognizable - it is dynamically young and not yet dispersed.

The evidence for this possible differentiated family will be presented andopen to discussion and analysis as to whether it is likely to be real or just achance association.

6

THE FOSSILIZED SIZE DISTRIBUTIONOF THE MAIN ASTEROID BELT

Authors: W. Bottke (1), D. Durda (1), D. Nesvorny (1), R. Jedicke (2)Institutes: (1) SWRI, Boulder (USA), (2) University of Hawaii (USA)

At present, we do not understand how the main belt asteroid populationevolved into its current state. During the planet formation epoch, the primordialmain belt presumably contained several Earth masses of material, enough toallow the asteroids to accrete on relatively short timescales (e.g., Weidenschilling1977; Wetherill 1992). The present-day main belt, however, only has 5e-4 Earthmasses of material (Petit et al. 2002). The mechanism that eliminated themass is constrained by the presence of Vesta’s basaltic crust (e.g., Davis et al.2002). If the main belt were massive for too long, Vesta’s crust would have beenobliterated by collisions.

To explain this mass depletion, Petit et al. (2002) invoke a ”dynamicalexcitation event” (DEE), where one or more mechanisms (e.g., perturbationsbetween growing planetary embryos embedded in the main belt and a fully-formed Jupiter, sweeping resonances) excite primordial main belt bodies andplace them onto planet-crossing orbits, where they strike the Sun, a planet,or are ejected from the inner solar system via an encounter with Jupiter. Ifthis scenario is valid, we can postulate that Vesta is the lone survivor of a largepopulation of similar objects that quickly disappeared from the main belt. Onlya fraction of these Vesta-like objects were shattered by impacts before they werescattered out of the main belt by the DEE, leaving behind the odd assortment of”scraps” that are observed today (e.g., V-type asteroids that are not associatedwith Vesta, olivine-rich A-type asteroids, and some M-type asteroids that couldbe exposed iron cores from differentiated asteroids).

In the interim between planetesimal formation and the DEE, however, itis unclear how the primordial main belt evolved, both collisionally and dy-namically. Encounters between planetary embryos and asteroids should haveincreased asteroid collision velocities enough that fragmentation would becomea more common outcome than accretion (with velocities comparable to the em-bryos’ escape velocities) (Davis et al. 1985; Petit et al. 2002). Constraints onwhat took place during this early epoch come from fragments of differentiatedasteroids (e.g, 1459 Magnya; Lazzaro et al. 2000), numerous meteorites fromdifferentiated asteroids (e.g., Northwest Africa 011, Yamaguchi et al. 2001),(iii) the surprising paucity of asteroid families in the present-day main belt(e.g., Tanga et al. 1999), and (iv) the wavy nature of the present-day main beltsize distribution (Durda et al. 1998). For space reasons, we concentrate here onitem (iv).

As noted in Davis et al. (2002), the main belt size-frequency distributionis wavy, with ”bumps” near 100 km and 3-4 km. Durda et al. (1998) claimedthese bumps were produced over 4.5 Gyr of collisional evolution in the mainbelt, with the wave launched via a transition from strength-scaling to gravity-scaling. While the results of Durda et al. (1998) show a remarkable fit to the

7

main belt size distribution with 2 < D < 300 km, their model makes use ofseveral unrealistic assumptions (i.e., they assumed most collisional evolutionoccurs after the DEE; their fragmentation scaling law (Q*) was very differentthan those derived from SPH codes and laboratory experiments). Thus, weare left with the following problem: collisional evolution can produce the mainbelt’s wavy size distribution, but the present-day main belt is too depleted ofmaterial to generate that wave over 4.5 Gyr of evolution.

To resolve this issue, we suggest the following evolution scenario took place.Planetesimals and planetary embryos accrete (and differentiate) in the primor-dial main belt during the first few Myr of the solar system. Gravitationalperturbations dynamically stir the main belt and allow fragmentation to begin.The primordial size distribution then undergoes a relatively short but violentperiod of comminution, enough to (i) disrupt a small fraction of the Vesta-sizeobjects and (ii) reproduce the wavy shape of the size distribution observed inthe current main belt. Finally, after an interval that might be as short as 10Myr or as long as 170 Myr, the DEE removes 99% of all the bodies from theprimordial main belt, essentially ”locking-in” the main belt’s wavy size distri-bution for the next 4.5 Gyr of evolution. Therefore, according to this scenario,the shape of the observed main belt size distribution is a ”fossil” produced bothby collisional evolution in the primordial main belt and the DEE, whereas 4Vesta is the lone dynamical survivor of a population of intact Vesta-size objectsthat were scattered by the DEE.

At the workshop, we will present preliminary modeling results of our sce-nario and will discuss the implications for several open problems in the aster-oid/meteorite community.

8

IMPACT STUDIES OF CATASTROPHIC DISRUPTIONOF ICE SPHERES AT THE UNIVERSITY OF

KENT HYPERVELOCITY IMPACT LABORATORY

Authors: M.J. Burchell and D.LowenInstitute: Centre for Astrophysics and Planetary Science, Univ. of Kent, Can-terbury (UK)

Previously results from the University of Kent laboratory have been reportedfor hypervelocity impact cratering in large, effectively semi-infinite ice targets[1, 2, 3, 4] including porous ice [5]. Here this work has been extended and newresults from hypervelocity impacts into solid ice spheres are reported. The im-pacts were carried out using a two stage light gas gun at the University of Kent[6], firing mm-sized projectiles in the speed range 1 to 6 km s−1. The size of thespheres ranged from 10 to 20 cm diameter. The impact speed for the transitionfrom cratering to disruption has been found and compared to published dataand models. Once disruption has occurred, the variation in fragment numberand size with increasing impact speed has been found. In addition, impacts arealso reported into porous ice spheres and a comparison made with the resultsfor impacts into non-porous ice.

References: [1] Burchell et al., Icarus 131, 210-222, 1998. [2] Grey et al.,J. Geophys. Res. 107(E10), 5076, doi:10.1029/2001JE001525, 2002. [3] Shrineet al., Icarus 155, 475 - 485, 2002. [4] Grey et al., J. Geophys. Res. 108(E3),5019, doi:10.1029/2002JE001899, 2003. [5] Burchell et al., Proc. AsteroidsComets and Meteors (ACM 2002), ESA-SP-500, 859 - 862, 2002. [6] Burchellet al., Meas. Sci. Technol. 10, 41 - 50, 1999.

9

CONSTRAINING THE COMPOSITIONS OF SILICATE-RICHASTEROID FAMILIES THROUGH NEAR-IR SPECTROSCOPY

Author: S.J. BusInstitute: Institute for Astronomy, University of Hawaii (USA)

Over the past decade, spectroscopic studies of asteroid families have focusedon the visible wavelength region that can be measured using CCD spectro-graphs. These studies have revealed a surprising level of spectral homogeneityamong members of nearly every family observed (e.g. S. J. Bus, Ph. D. Thesis,1999, and A. Cellino et al. 2002, in Asteroids III, U. of Arizona Press). Thisis in contrast to what might be expected from the meteorite evidence, whichsuggests many of the asteroid parent bodies underwent some level of thermalprocessing and differentiation. With recent advancements in IR detector tech-nologies, new generation near-IR spectrographs are allowing us to extend ourinvestigations of asteroids to longer wavelengths. This is particularly valuablein the study of silicate-rich asteroids whose spectra contain significant absorp-tion bands, centered near 1 and 2 microns. I present some preliminary resultsobtained with SpeX, a state-of-the-art near-IR spectrograph and imager now inuse at the NASA IRTF on Mauna Kea, Hawaii (J. T. Rayner et al. 1998, Proc.SPIE 3354, 468). A program aimed at probing the mineralogy of silicate-richasteroid families has been underway for the past two years. These new observa-tions cover the wavelength interval from 0.8 to 2.5 microns, and reveal not onlyspectral similarities among asteroid families like those observed at visible wave-lengths, but suggest even more significant levels of mineralogical homogeneityamong members of several families. Most notably, some families exhibit spectrathat indicate the presence of both low- and high-calcium forms of pyroxene, butwhich contain little or no olivine (J. M. Sunshine et al. 2002, LPSC XXXIII,#1356). This composition implies a history of partial melting and silicate dif-ferentiation within the parent bodies of these families. Still, we see no evidencefor spectral variations among the family members that would suggest they werederived from the disruption of a differentiated body. I will show examples ofspectra for several asteroid families, and will discuss some of the initial analysesthat make use of both Principal Component Analysis and Modified GaussianModeling (J. M. Sunshine et al. 1990, JGR 95, 6955).

10

ASTEROID FAMILIESInvited Review

Author: Alberto CellinoInstitute: INAF - Astronomical Observatory of Torino (Italy)

Asteroid families play a crucial role in any attempt at understanding thephysics of catastrophic disruption phenomena. The obvious reason is that fam-ilies were produced by collisional disruptions of single parent bodies. Studiesof the physical properties of families have taken profit of the improvementsof family identification techniques in the early 90s. The analyses performedup to 1999 were based on the assumption that the families we see today havenot been significantly affected by evolutionary processes. In more recent yearsnumerical simulations taking into account new dynamical mechanisms like theYarkovsky effect have evidenced that the above assumption is not correct. Onone hand, this makes it more complicated to derive detailed information on theoriginal kinematical properties of family-forming events. On the other hand,there are also exciting pespectives about the possibility to evaluate family ages.Very interesting results of recent analyses are reviewed. The huge growth ofthe databases of asteroid proper elements implies that times are now maturefor a new updated search for families. This task is made difficult, however, bythe fact that now many big groupings tend to mutually overlap in the space ofproper elements. Adopting too scrict criteria for family classification, however,could lead to problems in a correct understanding of the inventory and physicalproperties of these groupings.

11

METEORITICAL EVIDENCE AND CONSTRAINTSON IMPACTS AND DISRUPTION

Invited Review

Author: Guy J. Consolmagno SJInstitute: Specola Vaticana, V-00120, Vatican City State

Most meteorites come from the asteroid belt; indeed, they are fragments ofparent body asteroids that have themselves been catastrophically disrupted, byimpact, from the surface or interior of an asteroid. Thus their fabric containsevidence of the processes of impact and disruption that they experienced in theasteroid belt.

The traditional evidence, which continues to place strong constraints onthe disruption history of the asteroid belt, includes the presence and range ofshock features within meteorites, and the cosmic ray exposure ages. The shockevidence suggests that while some materials do experience shock energies oftens of gigapascals, such high pressures are neither ubiquitous nor necessaryeither to lithify the meteorites or to launch them into Earth-crossing orbits.And the relatively young exposure ages provide strong constraints on how oftendisruption events occur, and how long it takes from meteorites to be removedfrom the shielded interiors of asteroids.

New evidence of meteorite density and porosity, coupled with a growing setof data on asteroid densities, has provided an additional set of constraints onasteroid structure which have changed much of our thinking about how asteroidsrespond to large impacts. The evidence that asteroid densities are significantlyless dense than their corresponding meteorite type appears to indicate that allbut the largest asteroids can be divided into two classes: the strongly frac-tured but still coherent bodies (mostly S-types), roughly 20macroporous, andrubble piles (mostly C-type) that may 50macroporous. The presence of signif-icant porosity even among coherent asteroids, and the high porosity inferredfor the rubble piles, means both that virtually all asteroids have experiencedsome close-to-disruptive process, and that actual disruption and dispersion ofsuch porous bodies may be much more difficult to achieve (and model) thanpreviously thought.

Finally, examining the nature and location of the porosity in the meteoritesthemselves provides additional clues to the processes that lithify and set thefabric of this material. Among ordinary chondrites, porosity is not correlated atall with metamorphic grade and only in a second-order way with shock state; theaverage microporosity varies little with shock state, but the range of porositiesaround that average narrows as shock state increases. Most meteorite microp-orosity can be identified with microcracks visible in thin section SEM imagesand detailed x-ray scans. This suggests that porosity is set by the passage ofshock waves which compress, and decompress, materials as they pass throughthe rock. The ubiquitous presence of these microcracks is evidence confirm-ing the role of large impacts in the asteroid belt on lithifying and modifyingmeteoritic material.

12

STRUCTURE AND AGE OF THE ASTEROID FAMILIES:CONSTRAINTS FROM OBSERVATIONAL DATA

Authors: A. Dell’Oro, V. Zappala’, M. Delbo’Institute: INAF – Torino Observatory (Italy)

Twenty one well studied asteroid families have been analyzed by applyingthe HCM clustering methods to a sample of about 31000 numbered objects. A”local” analysis of the background has been used to define asteroid family mem-bership. Size-proper element (SPE) plots (i.e., plots of the absolute magnitudevs the proper semimajor axis, eccentricity and inclination) have been obtained.Assuming that the diurnal Yarkovsky effect is the only source of semimajoraxis mobility, a simple method is presented and applied to estimate the relativeage of each family. Themis turns out to be the oldest, Astrid and Veritas theyoungest ones. On the other hand, the properties of the H-sinI plots, whichare not trivially explained by the Yarkovsky effect, when compared with theother SPE plots, seems to suggest that the dispersion in semimajor axis havebeen only marginally affected by a post-formation evolution. The fundamentalrole of the SPE plots for achieving a better understanding of the formation andevolution of asteroid families is discussed.

13

WAVY SIZE DISTRIBUTIONS CAUSED BYDISCONTINUITIES IN SIZE-STRENGTH SCALING LAWS

Author: D.D. DurdaInstitute: SWRI, Boulder (USA)

Recent state-of-the-art numerical collisional evolution models, which includesize-dependent removal mechanisms for smaller particles (i.e., Poynting-Robert-son drag and the Yarkovsky effect), are producing good agreement betweenmodeled population size-frequency distributions and observed main belt andNEA asteroid populations and cratering records. Notable in both the modeledand observed size distributions are fairly distinct wave-like deviations from thelinear power-law distributions that are usually assumed to describe these popu-lations. Such ’waves’ can be induced in a collisionally evolving population dueto a small-size cutoff in the particle population (resulting from the rapid re-moval of bodies by non-gravitational forces such as P-R drag or the Yarkovskyeffect) or from a distinct change in slope in a size-strength scaling law resultingfrom the transition from strength-scaling for smaller bodies to gravity-scalingfor larger objects (Durda 1993, Campo-Bagatin et al. 1994, Durda and Dermott1997, Durda et al. 1998).

Discontinuities in size-strength scaling laws, as might occur at small sizeswhen particles are no longer homogeneous over the scale of fragmentation, canalso induce significant waves in evolved size distributions (Durda 2003). At sizescales small enough that material inhomogeneities due to the presence of chon-drules or Ni-Fe metal blebs or the effects of mineral grain boundaries becomeimportant, impact strengths may be dictated by the failure of internal flawswith a narrow range of critical tensile strengths (Durda and Dermott, 1996).Particle strengths may still depend on size, but may change discontinuously atspecific sizes and/or have a different size dependence in different size ranges.

I have conducted a series of numerical experiments using a collisional evo-lution model incorporating a variety of hypothetical size-strength scaling lawswith abrupt or discontinuous changes in impact strength at centimeter sizescales. The results show that significant waves can be induced in the size dis-tribution of larger particles. The wavelength of the waves is determined by thecritical projectile-to-target size ratio set by the size-strength scaling law. Theamplitude of the waves depends on the magnitude of the discontinuity in thesize-strength scaling relation. A larger magnitude discontinuity leads to largeramplitude waves. The phase of the waves depends on the nature of the dis-continuity. An abrupt change to stronger particles at smaller sizes leads to animmediate trough in the size distribution for particles just larger than that atwhich the strength properties change, since the stronger particles survive longerand represent an overabundant source of projectiles to break up larger bodies.An abrupt change to weaker particles at smaller sizes leads to an immediatecrest in the size distribution, since the weaker particles do not survive as longand represent an underabundant source of projectiles.

Much attention has been focused on determining how impact strength varieswith size for objects larger than the 10-cm size targets typically studied in

14

laboratory impact experiments. This is understandable as impact studies andcollisional models have traditionally focused on the evolution of populations oflarger objects where telescopic data can constrain the models. However, nowthat we are beginning to understand the extent to which the details of small-particle removal mechanisms and the effects of changes in size-strength scalingrelations can affect the evolved size distribution at even the largest sizes, thetime has come for a better understanding of the impact properties of particlessmaller than 10 centimeters.

15

THE FORMATION OF ASTEROID SATELLITESIN CATASTROPHIC IMPACTS:

RESULTS FROM NUMERICAL SIMULATIONS

Authors: D.D. Durda (1), W.F. Bottke (1), B.L. Enke (1), E. Asphaug (2),D.C. Richardson (3), Z.M. Leinhardt (3)Institutes: (1) SWRI, Boulder (USA), (2) University of Santa Cruz (USA), (3)University of Maryland (USA)

We have performed new simulations of the formation of asteroid satellites bycollisions, using a combination of hydrodynamical and gravitational dynamicalcodes. This initial work shows that both small satellites and ejected, co-orbitingpairs are produced most favorably by moderate-energy collisions at more direct,rather than oblique, impact angles. Simulations so far seem to be able to producesystems qualitatively similar to known binaries.

Our work takes advantage of a state-of-the-art numerical model that com-bines results from smooth-particle hydrodynamics (SPH) codes (Benz and As-phaug 1995), which accurately model the pressures, temperatures, and energiesof asteroid-asteroid impacts, and efficient N-body codes (Leinhardt et al. 2000),which can efficiently track the trajectories of tens-of-thousands of individual col-lision fragments. This technique of combining the two codes, and preliminarysimulation results, have been described by Durda et al. (2001) and Michel etal. (2002). Simulations using SPH codes are used to model various impactsbetween colliding asteroids. Once the relevant portions of the impact phaseare complete (crater formation/ejecta flow fields established with no furtherfragmentation/damage), the outcomes of the SPH models are handed off asthe initial conditions for N-body simulations, which follow the trajectories ofthe ejecta fragments for sufficient time to search for the formation of boundsatellite systems. Collisions between debris fragments are treated as mergersresulting in a new spherical particle of appropriate combined mass and equiv-alent diameter. The N-body simulations are run to a time about 4 days afterthe impact, thus simulating only the initial formation of bound satellites. Longterm dynamical evolution of individual satellite systems may be run in futurestudies.

To date we have run 160 SPH/N-body simulations of impacts onto 100-kmdiameter target asteroids. The non-rotating targets are assumed to be sphericaland are composed of solid basalt with a density of 2.7 g/cm3. The sphericalbasalt projectiles range in diameter from 10 to 46 km, impact speeds range from2.5 to 7 km/s, and impact angles range from 15 to 75 degrees (nearly head-onto very oblique). These results show that energetic (catastrophic) collisionscreate numerous fragments whose orbits can be changed by (i) particle-particleinteractions and by (ii) the reaccretion of material onto the remaining targetbody. Together, these effects allow some impact debris to enter into orbit aroundthe remaining target body, often a gravitationally reaccreted rubble-pile. Werefer to this type of satellite as a SMATS (SMAshed Target Satellite). Wealso find that numerous smaller fragments escaping the impact site have similar

16

trajectories, such that many become bound to one another. We refer to thistype of satellite as an EEB (Escaping Ejecta Binary).

It appears that most SMATS are formed around the largest remnants ofmoderately catastrophic impacts, in which the mass of the largest remnant isless than half that of the original target. For our 100-km diameter targets thistranslates to those largest remnant diameters of 80 km or less. At lower impactenergies the events are essentially large cratering impacts and most ejected ma-terial eventually reaccretes onto our spherical remnants, leaving little orbitingmaterial. At very high impact energies targets are severely disrupted, resultingin small largest remnant sizes and less bound debris. We find that most SMATSare formed from low-angle rather than oblique impacts. It appears that mostEEBs result from moderately catastrophic impacts, although many are also pro-duced by the highly disruptive impacts resulting from large projectiles strikingat or below the average main-belt mutual impact speed of 5 km/s. Other largeimpactors striking at higher speeds evidently eject collision debris so energet-ically that few escaping fragments can remain bound to each other. Impactsof 34-km diameter projectiles at 3 km/s at impact angles of 25◦ appear veryefficient at producing relatively large satellites around the largest remnant aswell as large numbers of modest-size binaries among their escaping ejecta.

17

CATASTROPHIC DISRUPTION OF METEORITES:IMPLICATIONS FOR ASTEROIDSAND INTERPLANETARY DUST

Authors: G.J. Flynn (1), D.D. Durda (2), S. Hart (3), E. Asphaug (3), T. VanVeghten (1)Institutes: (1) SUNY-Plattsburgh (USA), (2) SWRI, Boulder (USA), (3) Uni-versity of Santa Cruz (USA)

Introduction

Interplanetary dust particles (IDPs) recovered from the Earth’s stratosphereare, on average, enriched in the moderately volatile elements by ≈ 3x over theCI meteorite composition [1]. One possibility is that the IDPs are derivedfrom a volatile-rich parent body. Alternatively, they may sample a volatile-rich subunit of the parent body. Chondritic meteorites are dominated by twochemically distinct subunits - a volatile-rich matrix and volatile-poor chondrulesand inclusions. The chondrules and inclusions are large (≈ mm) compared toIDPs, and may be underrepresented in the ≈ 10 µm collision debris. A detailedunderstanding of the response of meteorites to hypervelocity impact is requiredto understand the asteroid fragmentation process. We have performed a seriesof impact disruption experiments using as targets samples of several meteorites.

Samples

Three moderately weathered ordinary chondrite (OC) finds, a 235.8 gramsample of the L6 OC NWA79, a 248 gram sample of the unclassified OC NWA620and a 106.8 gram sample of the unclassified OC MOR001, and four unweatheredfalls, two samples, one 182.5 gm and the second 152.7 gm, of the L5/6 OC Mbale,a 105 gm sample of the highly-friable L4 OC Saratov, and a 70.9 gm sample ofthe CV3 carbonaceous meteorite Allende were used as targets. Each target wasstruck by an ≈ 5 km/sec (comparable to the mean collision velocity in the mainbelt) Al projectile fired from the NASA Ames Vertical Gun.

Some primary debris fragments were collected in aerogel blocks placed ar-ound the target. The individual particles collected in aerogel were analyzedin-situ using the X-Ray Microprobe at beamline X26A of the National Syn-chrotron Light Source at Brookhaven National Laboratory. We distinguisholivine chondrules from matrix by the significantly higher Ni content of thematrix. Twelve of the 19 fragments <25 µm in size in the first aerogel analyzedfrom the MOR001 shot had Ni/Fe count rate ratios >0.05 (high-Ni), while only4 of the 29 fragments >25 µm had Ni/Fe count rate ratios >0.05. For the firstaerogel from the NWA791 shot, 8 of 18 fragments <25 µm had Ni/Fe countrates >0.05, while only 1 of the 15 fragments >25 µm had Ni/Fe count rates>0.05. For the first aerogel from the NWA620 shot, 11 of 30 fragments <25µm had Ni/Fe count rates >0.05, while only 2 of the 12 fragments >25 µm hadNi/Fe count rates >0.05. These results indicate that high-Ni matrix material is

18

over-represented in the <25 µm size fraction of the impact debris compared tothe >25 µm fraction of these ordinary chondrite meteorites.

The results are even more striking for the Allende and Saratov shots. Twelveof the 15 fragments <35 µm in size in the first aerogel analyzed from the Allen-deshot had Ni/Fe count rate ratios >0.05, while only 4 of the 24 fragments >35µm had Ni/Fe count rate ratios >0.05. For the Saratov shot, 6 of the 7 frag-ments <10 µm in size had Ni/Fe count rate ratios >0.05 (high-Ni), while noneof the 15 fragments >10 µm in size had a Ni/Fe count rate ratio >0.05. Thedifference in the transition size from matrix fragments to olivine fragments be-tween Saratov and Allende may reflect the fact that Saratov contains very littlematrix occurring in small patches, while Allende has both a higher abundanceof matrix [CV3 meteorites like Allende have 35 to 50% matrix] and larger sizematrix regions. In addition, most of the samples show a change of slope of themass-frequency distribution at roughly the mass of the chondrules, indicatingthat the disruption cannot be modeled as a single power-law.Conclusion

These resultas indicate that many meteorites suffer preferential failure alongthe chondrule-matrix boundary, an effect that should be considered in modelingthe cratering and disruption of their asteroidal parent bodies. The results alsosuggest that fragments <25 µm in size from hypervelocity impacts onto chon-dritic asteroids significantly oversample the volatile-rich matrix material. Thesesmall particles are rapidly removed from the debris trail near the asteroid, spi-raling towards the Sun because of Poynting-Robertson drag before most of thelarger particles are broken up by collisions. Thus, IDPs <25 µm collected atEarth may be biased towards the matrix composition of the asteroid, whilelarger particles, collected as micrometeorites from the polar ices, may preferen-tially sample chondrule material.

Reference: [1] Flynn, G.J. et al. (1996) in Physics, Chemistry and Dynamics ofInterplanetary Dust, ASP Conf Ser. 104, 291-294.

19

TOWARD A TAXONOMY OF THEEDGEWORTH-KUIPER BELT OBJECTS:

THE MULTIVARIABLE APPROACH

Authors: M. Fulchignoni (1)-(2), A. Delsanti (1), M.A. Barucci (1), M. Birlan(1)Institutes: (1) LESIA, Observatoire de Paris (France), (2) Universite DenisDiderot - Paris 7 (France)

The principal component (PC) and g-mode multivariable statistics have beenused in analysing the set of 34 Edgeworth-Kuiper Belt objects (EKBOs) - 28TNOs and 6 Centaurs - for which the B, V, R, I, J homogeneous photometry(i.e. data obtained in one of B-V, V-R, V-I and V-J colour indices by the sameobserver, during the same run) were available. The results obtained show thatthe V-I and V-J colours are the key parameters in structuring the sample inhomogeneous groups. The PC1 axis (which contains 93% of the sample totalvariance) spans five times more than the PC2 (which contains 6% of the sampletotal variance). The extremes of PC1 axis contain the objects having a flatspectrum (low PC1 values) and a very red spectrum respectively. Independently,the g-mode analysis allows us to distinguish six homogeneous groups of objectswhich confirm and extend the results obtained with the PC analysis. In additionto these groups, a few objects remain not included in any group (i.e. do nothave significant similitude with other objects) and yet give an indication of amore complex compositional structure of the sample. These preliminary resultshave to be confirmed and completed when a larger sample will be available,but they provide some interesting hint in understanding the evolution, mainlycollisional, of the EKBOs.

20

SPINS AND SATELLITES:PROBES OF ASTEROID INTERIORS

Invited Review

Authors: Alan W. Harris (1) and Petr Pravec (2)Institutes: (1) Space Science Institute, La Canada (USA), (2) AstronomicalInstitute, Pragues (Czech Republic)

We will present what observational data reveal to us about asteroid inte-riors. Particularly interesting are asteroids with sizes 0.1-1 km where there isthe transition region between rubble piles and ”monoliths”. New observationshave shown that while the main transition lies in a narrow range (appears onlyseveral tens meters wide) around the diameter of 0.15 km, a small fraction ofobjects with diameters 0.2-1 km are rapid rotators and therefore coherent bod-ies. Another interesting group of asteroids that we will address are tumblingasteroids. Now we know of tumblers present in both groups: Slowly rotatinglarger asteroids that may be or even probably are rubble piles, and fast rotat-ing small asteroids that must be coherent bodies. In both cases, the dampingtimescales are consistent with their expected interior properties and collisionallifetimes.

21

ASTEROID ROTATIONS, IMPLICATIONS FORINTERNAL STRUCTURE AND DISRUPTION

Invited Review

Authors: K.A. Holsapple (1) and K.R. Housen (2)Institutes: (1) University of Washington, Seattle (USA), (2) The Boeing Co.,Seattle (USA)

Experiments and code calculations of the hypervelocity disruption of smallsolar-system bodies are generally for non-rotating bodies. Scaling theories of theauthors and others use those experiments and calculations, and other evidence,to patch together the estimates of the critical energy per unit mass Q∗ requiredin an impact to disrupt or to disperse the bodies. However, it is well known thatthe asteroids are generally spinning; and, according to current thought, someare rotating at nearly the limits at which they would self-destruct. Clearly asmall body at nearly a self-destructing state would be much easier to disruptby an impact than one that is far from such a state.

We shall review the data on spin of asteroids and show how it nicely fits thelimits derived by assuming a model of a porous Mohr-Coulomb material havingstrength of only a few Kpa (Holsapple, 2001, 2003). The comparison is oftenused to suggest that most asteroids have little strength, and therefore a porousstructure. Even the small so-called ”fast rotators” do not spin fast enough todisprove that assumption (Holsapple, 2003).

Then, we will speculate about the implications regarding the scaling of dis-ruption. When an asteroid is at its spin limit, determined by either of strengthor gravity, its disruption energy becomes zero. In the strength regime, for as-teroids less than a few tens of meters, most asteroids may be at a small fractionof their spin limit; so, on average, the effects are not dramatic. However, forthose that are near a spin limit, a small additional energy addition from a smallimpact would both disrupt and disperse the asteroid. However, in the large,gravity dominated regime, many asteroids rotate at spins up to 50% of theirlimit, which offsets about 1/4 of the gravity stress. As a consequence, onemight expect that its disruption energy, which is dictated by the gravity stress,would be reduced to perhaps 75% of that for zero spin. However, some initialexperiments in the strength regime suggest a much greater reduction in disrup-tion energy than predicted from these simple arguments, the reason is not yetknown. Further more, a few asteroids are spinning nearly at their limit, aboutthree time the average, so their disruption energy would be very small: thesewould disrupt very easily. But, at the spin limit, the spin velocities are still lessthan the escape velocity by a factor of about 2, so while any small energy inputcould disrupt the body, it would not be enough to disperse it.

Among other implications, the concept that spin would make disruption eas-ier means that the faster rotating asteroids at a given size would have a shorterlifetime than the slowly rotating ones, and that would reduce the population ofthe faster spinning ones. If the effect is more pronounced for the larger ones,a different distribution might be expected for large versus small asteroid spins.Much more analysis is required.

22

SCALING LAWS OF COLLISIONAL FRAGMENTATIONInvited Review

Authors: K.R. Housen (1) and K.A. Holsapple (2)Institutes: (1) The Boeing Co., Seattle (USA), (2) University of Washington,Seattle (USA)

Asteroid collisions are conveniently studied in the lab. The materials of thecolliding bodies are freely varied as well as their shape, speed and, to a verylimited extent, their size. Numerical simulations provide a different avenue ofstudy and have a distinct advantage over experiments in that they can addressnot only the small lab-scale collisions, but also very large events where gravita-tional effects are dominant. However, the simulation results depend entirely onthe physics they incorporate, much of which is poorly known at present.

Scaling laws take an intermediate position. They are founded on basic phys-ical insights and simple yet elegant results of similarity theory. Once verifiedby experiment, scaling laws can be used to extrapolate laboratory results tolarger sizes, as long as the extrapolation does not carry into regimes governedby physical processes not included in the scaling analysis.

An important application of scaling lwas has been to study how the frag-mentation of rocks that can be held in one’s hand differs from the shattering ofkilometer-scale asteroids. The strength of rock has long been known to dependon size scale and loading duration. The shock loading process during asteroidcollisions occurs over a vastly longer duration than in laboratory experiments,which effectively makes coherent rocky asteroids much weaker than their labora-tory counterparts. The size dependence has now been confirmed in experimentsusing granite targets, but little is known about possible size effects in othermaterials. For example, porous bodies may not experience the brittle fractureobserved in the fragmentation of rocky targets and so may exhibit a completelydifferent dependence on target size. Cratering experiments indicate very littlesize dependence in the strength of soils. If the same holds for disruption ofporous asteroids, these objects could be much more resistant to fragmentationthan rocky bodies.

Porosity has been suggested to make asteroids fragmentation-resistant foranother reason. Porous structures efficiently damp the impact shock. Experi-ments over the past several years have shown that even though a porous tar-get is mechanically weak, the specific energy of fragmentation can be as largeas for competent rocky materials. However, the specific relationship betweenstrength, porosity and the fragmentation threshold is still unclear because ofthe difficulties involved in independently varying strength and porosity. Ourcurrent knwoledge, or lack thereof, regarding target strength and porosity willbe discussed in this review.

We will also discuss another scaling question involving the possible affectof asteroid spin on the specific energy for fragmentation (see abstract by Hol-sapple and Housen). Laboratory experiments involving spinning targets will besummarized.

23

OBSERVATIONS OF 2002 NY40:AN ORDINARY CHONDRITE?

Author: E.S. Howell (1), A.S. Rivkin; M.C. Nolan, J.L. Margot, G. Black, S.J.Bus, M. Hicks, W.T. Reach, T.H. Jarret, R.P. BinzelInstitute: (1) Arecibo Observatory/ Cornell University (USA)

The near-Earth asteroid 2002 NY40 was discovered on July 14, 2002 bythe LINEAR survey. The object passed within 0.0035 AU of the Earth onAugust 18, when observations were obtained at a large range of wavelengths,from visible to radar (12.6 cm). The combination of visible and near-infraredspectroscopy gives some indication of the composition. Thermal emission inthe 3-micron region gives constraints on the visible albedo, which are confirmedindependently by the radar size and visible magnitude. The lightcurve waswell measured by a large number of observers, and the rotation period is well-determined at 19.99 hours. The spectrum and albedo are a very good match toan LL chondrite spectrum over the range measured. No appreciable reddeningis seen in the asteroid spectrum, which suggests that the surface has not beennoticeably affected by the same processes seen in many other pyroxene andolivine-rich asteroids. The radar images show that this asteroid looks like twospheroidal units joined together. Analysis and implications of these observationswill be discussed.

24

PHYSICAL PROPERTIES OF NEOs SURFACES:WHAT CAN BE LEARNED FROM ASTEROID REGOLITHOBSERVATIONS AND MICROGRAVITY LABORATORY

SIMULATIONS

Author: A.-Chantal Levasseur-RegourdInstitute: University of Paris VI / Aeronomy CNRS (France)

Observing the properties of solar light scattered by NEOs objects provides,in the absence of in-situ observations, unique information on the physical prop-erties of their surface and sub-surface. Of major significance are the phase angleand wavelength dependence of the linear polarization of the scattered light. Al-though surfaces are (at least partly) covered with a regolith build up of irregularparticles with sizes larger than visible wavelengths, significant differences existbetween the morphology of the surface layers of S-type and C-type objects. Pos-sible comparison between some C-type NEOs and defunct comets nuclei will bediscussed, with emphasis on the likelihood of gravitational aggregates amongstNEOs.

Also, the significance of ICAPS facility measurements for NEOs surface andsub-surface properties will be discussed. ICAPS, now in phase B at ESA, isexpected to take place inside the Columbus laboratory on-board the ISS. Themeasurements will provide information on light scattering and impacts on re-goliths built in microgravity conditions.

Reference: A.C. Levasseur-Regourd, E. Hadamcik, Light scattering by ir-regular dust particles in the solar system: observations and interpretation bylaboratory measurements, J. Quant. Spectros. Radiat. Transfer, 79-80, 903-910, 2003.

25

ON MODELLING THE COLLISIONAL EVOLUTIONOF MINOR BODY POPULATIONS

Invited Review

Author: Francesco MarzariInstitute: Dept. of Physics, Univ. Padova (Italy)

I will review the state of art of modelling the collisional evolution of minorbody populations and describe the most recent developments. I will concen-trate on the Main Belt and outline the constraints we have from observations,experiments, and space missions. These constraints include the size distribu-tion of asteroids, the number of dynamical families, Vesta’s basaltic crust, CREages of meteorites, and rotation rates. The numerical codes simulate the evo-lution of a putative primordial population of asteroids that dates back to theearly stage of the solar system just after the great-impact phase. These codescompute the evolution with time of the size distribution and the final outcome,after 4.5 Gyr, is compared with the present size distribution. A recent fea-tures concerns the capability of predicting also the number of families and theirsize distribution. Two major open problems related to this kind of modellingare: 1) the effects of dynamics on the collisional evolution 2) the fragmentationmodel. Dyamics is particularly relevant in connection with the behaviour of thesmall size end of the distribution. Small bodies can be affected by Yarkowskyand Poynting-Robertson drag that can deplete the small tail end independentlyfrom collisions. This effect may propagates in the form of waves up to largersizes. The fragmentation model, describing the way in which a body is disruptedduring a collision, is the core of this kind of modelling. Scaling laws and thephysical structure of the bodies are relevant input parameters but, at the sametime, they can be constrained by the simulation outcomes.

26

PROGRESS ON THE SEARCH FOR BINARY ASTEROIDSUSING ADAPTIVE OPTICS

Authors: W.J. Merline (1), L.M. Close (2), C. Dumas (3), C. R. Chapman(1), F. Menard (1), P.M. Tamblyn (1), D. Nesvorny (1) and D.D. Durda (1)Institute: (1) SWRI, Boulder (USA), (2) University of Arizona (USA), (3) JetPropulsion Lab, Pasadena (USA)

We present an update on our search for binary asteroids using adaptive optics(AO) techniques on ground-based telescopes. Since 1998, we have been perform-ing observations in search for this new class of object around main-belt, JupiterTrojan, and near-Earth asteroids. These data were acquired at Mt. Wilson,Canada-France-Hawaii (CFHT), Keck, Gemini, and VLT. We have made obser-vations of about 500 different asteroids, mostly within the Main Belt. Study ofbinary asteroids allows us to determine a bulk density for the primary asteroid,which in turn yields information on the composition and structure (e.g. bulkporosity). This is important because for some asteroid populations, such as theTrojans, we may not even have meteorite samples to guide us on composition ormicroporosity. Binary asteroids also provide a real-life laboratory for studyingthe results of collisions. The populations of binaries we observe must be the re-sult of steady-state collisional processes that create and destroy these systems.However, in some cases, for example among very young (few MY) dynamicalfamilies (e.g. those being identified by Nesvorny et al.), we may have an oppor-tunity to observe the results of binary production from catastrophic collisionsbefore enough time has passed for subsequent collisions to destroy many of thesatellites. We have recently added such objects to our ground-based observingprogram and have a large HST program for this purpose as well. So far, in ourground-based AO program, we have discovered 8 binary asteroids (co-discoveredin one case), including the only Trojan binary known. Discoveries within thelast year include (121) Hermione and (1509) Esclangona. These represent 8 ofthe 11 known main-belt or Trojan binaries. Most of the binaries are seen aroundlarger C-like primaries. The satellites occur in orbits at about 10 primary radii,and are small relative to the primary. We believe these objects were formedby the mechanism of reaccretion of ejecta, in orbit, from a large impact, as de-scribed by Weidenschilling et al. (1989, Asteroids II) and further in Merline etal. (2002, Asteroids III). Interestingly, the two smallest (about 10 km primarydiameter) main-belt asteroids known to be binary, (3749) Balam and (1509) Es-clangona, are both S-type and show loosely bound companions that are larger,relative to the primary, than those in the C-like systems. These are likely tobe the first examples of systems formed as co-orbiting ejecta fragments froma catastrophic impact, also as described by Weidenschilling et al. New modelsof both of these formation mechanisms are being made by Durda et al. (2003,LPSC 33). Finally, a third mechanism described by Weidenschilling et al. isfission of an asteroid after spin up from a large impact. One of the binaries,the only true double asteroid, (90) Antiope, seems to be a candidate for beingformed by this mechanism. Thus, at least 3 mechanisms are required to explain

27

the diversity of main-belt binaries seen. Our overall statistics remain that about2% of main-belt asteroids harbor companions, at least to our detectability lim-its. Binary frequencies among the NEAs and TNOs appear to be significantlylarger and they have different characteristics, and yet other formation mecha-nisms must be operating in these populations. Other searches on main-belt andTrojan asteroids have recently been made using AO by Margot and Brown, andby Marchis et al.; searches using HST have been made by Storrs et al.

28

NUMERICAL MODELS AND SIMULATIONSOF SMALL BODY COLLISIONS

Invited Review

Author: Patrick MichelInstitute: Observatoire de la Cote d’Azur, Nice (France)

In the past two years, numerical models of catastrophic disruption have beengreatly improved. In particular, numerical simulations of the collisional disrup-tion of large asteroids (such as parent bodies of asteroid families) have beendeveloped and which account not only for the fragmentation of the solid bodythrough crack propagation, but also for the mutual gravitational interaction ofthe resulting fragments. The numerical procedure consists in using a 3D SPHhydrocode to compute the fragmentation phase and the parallel N -body codepkdgrav to compute the subsequent gravitational reaccumulation phase. Appliedto the formation of an asteroid family as a result of the disruption of a largeparent body, it is generally found that the parent body is first completely shat-tered at the end of the fragmentation phase, and then subsequent gravitationalreaccumulations lead to the formation of an entire family of large and smallobjects with dynamical properties similar to those of the parent body. Is is alsofound that the formation of satellites around family members is a frequent andnatural outcome of collisional processes (Michel et al. 2001, 2002).

Different impact regimes, from barely disruptive to highly catastrophic, havebeen studied and represented by well-identified asteroid families, such as the Eu-nomia and Koronis families. Moreover, a recently identified very young family,Karin, has provided a unique opportunity to study a collisional outcome almostunaffected by erosion and dynamical diffusion. Numerical simulations of thisfamily formation have been performed, aimed at determining which classes ofcollisions reproduce the main family characteristics. Different models of theinternal structure of the parent body have been considered. The particular sen-sitivity of the outcome to the internal structure of the parent body has allowed toshow that this family must have originated from the breakup of a pre-fragmentedparent body. Moreover, all large family members are again gravitationally reac-cumulated smaller bodies with some of them having possibly evolved towardEarth-crossing orbits (Michel et al. 2003). Most identified families are likely tohave had a similar history and to check this suggestion, new numerical simula-tions of asteroid family formations have been performed with different modelsof parent body’s internal structure.

A review of our current understanding based on these simulations will bepresented and the current limits of the numerical models will be discussed.

29

THE SHALLOW MAGNITUDE DISTRIBUTIONOF ASTEROID FAMILIES

Authors: Morbidelli A. (1), Nesvorny D. (2), Bottke W.F. (2), Michel P. (1),Vokrouhlicky D.(3), Tanga P. (1)Institutes: (1) Observatoire de la Cote d’Azur, Nice (France), (2) SWRI, Boul-der (USA)

It is well known that asteroid families have steeper absolute magnitude (H)distributions for H <12–13 values than the background population. Beyondthis threshold, the shapes of the absolute magnitude distributions in the fam-ily/background populations are difficult to determine, primarily because bothpopulations are not yet observationally complete. Using a recently generatedcatalog containing the proper elements of 106,284 main belt asteroids and aninnovative approach, we debiased the absolute magnitude distribution of themajor asteroid families relative to the local background populations. Our re-sults indicate that the magnitude distributions of asteroid families are generallynot steeper than those of the local background populations for H > 13 (i.e.,roughly for diameters smaller than 10 km). In particular, most families haveshallower magnitude distributions than the background in the 15–17 magnituderange. Thus, we conclude that, contrary to previous speculations, the popu-lation of kilometer-size asteroids in the main belt is dominated by backgroundbodies rather than by members of the most prominent asteroid families. Webelieve this result explains why the Spacewatch, Sloan Digital Sky Survey, andSubaru asteroid surveys all derived a shallow magnitude distribution for thedimmer members of the main belt population.

We speculate on a few dynamical and collisional scenarios that can explainthis shallow distribution. One possibility is that the original magnitude distri-butions of the families (i.e., at the moment of the formation event) were veryshallow for H larger than ∼ 13, and that most families have not yet had thetime to collisionally evolve to the equilibrium magnitude distribution that pre-sumably characterizes the background population. A second possibility is thatfamily members smaller than about 10 kilometers, eroded over time by colli-sional and dynamical processes, have not yet been repopulated by the break-upof larger family members. For this same reason, the older (and possibly char-acterized by a weaker impact strength) background population shows a shallowdistribution in the 15–60 km range.

30

LABORATORY EXPERIMENTSON IMPACT DISRUPTION

Invited Review

Author: Akiko NakamuraInstitute: University of Kobe (Japan)

Laboratory impact disruption experiments have been conducted for decades,providing insights into impact disruption phenomena and processes, and refer-ence data for numerical modeling and scaling laws. The collision conditionsexamined have broadened with time and include the target and projectile mate-rials, microscopic and macroscopic structures and their sizes, collision velocitiesand geometry, and ambient temperature and pressure. One question is how thedegree of fragmentation, the size and shape distribution of fragments, and thevelocity and spin distribution of ejecta depend on the above collision conditions.While most measurement methods and techniques have changed little since thelate 1970s, advances in digital high-speed imaging and image analysis have madeit easier to observe in-situ crack-growth and ejecta motion.

We will review the parameter space of previous laboratory studies, withrespect to both collision conditions and outcomes, with a slight emphasis onrecent achievements.

31

EJECTA SIZE-VELOCITY RELATIONSHIPDERIVED FROM THE DISTRIBUTION OF SECONDARYCRATERS OF SMALL, KM-SIZED CRATERS ON MARS

Authors: A. Nakamura, Y. Hirase, and T. MichikamiInstitute: University of Kobe (Japan)

The ejecta size-velocity relationship of the impact process may be dependenton the scale of the event and the material property of the target body. Melosh(1984) proposed a model in which the ejection velocity of spall fragments isproportional to the strength of the target body and inversely proportional tothe density and velocity of sound. Vickery (1987) studied the ejecta size-velocitydistribution by analyzing the size of secondary craters and the distance betweensecondary and primary craters. The diameters of the primary craters rangedfrom a few tens to one hundred km. We studied the distribution of the secondarycraters of small, km-size primary craters using images obtained with the MarsOrbiter Camera (MOC), and added data on ejecta with a velocity of 100 m/s,to fill the gap between previous observations by Vickery and laboratory work.Our data fit the tendency shown in the previous work. We will also comparethe results for Martian craters and craters on the Moon.

32

VARIETY OF NEAR-EARTH ASTEROIDS

Author: M. NolanInstitute: Arecibo Observatory, Arecibo (USA)

The recent upgrade of the Arecibo planetary radar system, combined withthe huge increase in the near-Earth asteroid (NEA) discovery rate by largesurvey programs, has greatly increased our ability to observe these objects withradar. Radar provides size, shape, rotation, and trajectory information, andin most cases is the only ground-based technique that spatially resolves near-Earth objects. Radar observations provide rotation rates in only a few minutesof observation, almost independent of rotation rate. There is a sin(inclination)factor and a weak bias (period−1/2) towards slower rotating objects from S/N,but these effects rarely cause significant problems.

The single clearest result of these observations is the great variety of near-Earth objects, with binary systems, very fast and very slow rotations, spheres,“bifurcated” objects, and “shards”, suggesting that a similar variety of produc-tion and delivery mechanisms deliver these objects to near-Earth orbit.

As we observe a reasonable number of asteroids, we can begin to identify in-ternal structure, based on shape and rotation. For NEAs, the internal structureis determined by the “recent” collisional history, and the mechanism of deliveryfrom the asteroid belt. By examining the gamut of NEA structures, we hope toconstrain those processes.

33

ANALYTICAL AND NUMERICAL MODELING OFASTEROID COLLISIONAL EVOLUTION:

RECENT RESULTS

Authors: D.P. O’Brien and R. GreenbergInstitute: University of Arizona, Tucson AZ (USA)

Analytical StudiesWe have extended the frequently cited work of Dohnanyi (1969) and sub-

sequent authors, who analytically modeled a size-dependent collisional cascadefor the case where material strength (per unit mass or volume) is indepen-dent of size. Dohnanyi found that the steady-state size distribution of such apopulation has a slope of -3.5 (incremental). It has been shown by numerousauthors that over the size range of the asteroid belt, material strength is, in fact,a size dependent property. Small bodies decrease in strength with increasingsize, due to the presence of larger primary flaws, and larger bodies increase instrength with increasing size, due to gravitational compression and reaccumu-lation. When strength varies with size, the slope of the size distribution differsfrom the Dohnanyi value of -3.5, and the transition between the small- andlarge-body strength regimes leads to ‘waves’ in the size distribution. We havedeveloped an analytical model which includes size-dependent strength. Fromthis model, we have derived simple analytical relationships which relate theslope of the steady-state asteroid size distribution to the slope of the strengthvs. size law, and which describe the amplitude and wavelength of the wavescaused by the transition between strength regimes (O’Brien and Greenberg, inpress). We will discuss these results and their implications.Numerical Studies

In order to study the collisional evolution of the asteroid belt in full de-tail, and to take non-collisional effects into account, numerical modeling is re-quired. We have modified our numerical collisional evolution model to includethe non-collisional removal of bodies by size-dependent radiation forces such asthe Yarkovsky effect. Such a model must be able to fit a number of constraints,such as the observed main-belt size distribution and the observed NEA popu-lation (which consists of bodies removed from the main belt), the lifetimes ofmeter-scale bodies (as inferred from the cosmic ray exposure ages of meteorites),and estimates of the strength of asteroids from analytical and numerical modelsand laboratory experiments. Furthermore, the non-collisional removal rates weassume must be consistent with current estimates of the removal rate of bodiesfrom the main belt. We are able to achieve a good fit to all of these constraintswith our model. We will present our most recent results, and discuss their im-plications for the collisional and dynamical evolution of the main belt and forthe relation between the NEA and main-belt populations.

34

YARKOVSKY DEPLETION ANDASTEROID COLLISIONAL EVOLUTION

Authors: U. Penco (1), A. Dell’Oro (2), P. Paolicchi (1), A. Campo Bagatin(3), A. La Spina (1), A. Cellino (2)Institutes: (1) University of Pisa (Italy), (2) INAF, Torino Observatory (Italy),(3) University of Alicante (Spain)

Using of a recently refined numerical model we investigated the effectivenessof the Yarkovsky effect in depleting the Main Belt population and in affectingthe asteroid size distribution. We took into account various possible depletionchannels associated with the major resonances which, according to many stud-ies, are efficient in removing objects from the Belt. Under these assumptions,we find that the efficiency of depletionary effects is larger than discussed ina previous paper (Penco et al, ACM 2002, ESA-SP-500, 363). The possibil-ity to obtain major modifications of the observable size distribution depends,however, on the assumed input parameters (population, collisional parameters,Yarkovsky parameters). A systematic analysis of the space of the parametersis then required. In some favourable cases it is even possible to produce wavesin the size distribution. This effect has to be taken into account when usingthe observed size distribution to constrain the main parameters of collisionalphysics (impact strength, scaling laws and so on).

35

RUBBLE PILES AND MONOLITHSInvited Review

Author: Derek C. RichardsonInstitute: University of Maryland, College Park MD (USA)

Recent ground- and space-based observations of asteroids have revealed thatthese bodies are far more complex than once imagined. Surprisingly low bulkdensities, giant craters, unusual shapes, non-principal-axis spin states, and satel-lites are all challenging our understanding of how asteroids form and evolve.Since asteroids are the remnants of the planet building era, understanding theirnature improves our understanding of the origin of solar systems in general. Iwill review some of the more puzzling aspects of asteroid morphology, includ-ing the existence of asteroid satellites, and discuss recent theoretical advancesaimed at understanding our tiny neighbors. I will show that both theoreticaland observational evidence is pointing increasingly to asteroids being fragile as-semblages of smaller pieces, called gravitational aggregates. The consequencesof such fragmented internal structure on asteroid evolution and hazard mitiga-tion will be discussed. This work has been supported in part by the NationalAeronautics and Space Administration under Contract No. NAG511722 issuedthrough the Office of Space Science.

36

INTERACTING ELLIPSOIDS: A MINIMAL MODELFOR THE DYNAMICS OF RUBBLE-PILE ASTEROIDS

Authors: F. Roig (1), R. Duffard (1), D. Lazzaro (1), P. Penteado (1), T.Kodama (2)Institutes: (1) Observatorio Nacional/MCT – Rio di Janeiro (Brazil), (2) In-stituto de Fısica/UFRJ – Rio di Janeiro (Brazil)

Rubble-pile asteroids are strength-less bodies formed by gravitational reac-cumulation of fragments after the breakup of large asteroids during energeticcollisional events. Many collisional processes in the Solar System, like the for-mation of asteroid families, the tidal disruption of comets and NEAs and theformation of asteroid binaries and satellites may involve parent bodies with arubble pile structure. We present here the results of a simple mechanical modelrecently introduced to study the dynamics of such kind of asteroids. In thismodel, a rubble-pile consists of N interacting fragments represented by rigidellipsoids, and the equations of motion explicitly incorporate the minimal de-grees of freedom necessary to describe the attitude and rotational state of eachfragment. We perform different tests of this model and compare the results withthose from hydrodynamical models and laboratory experiments. We find that,in spite of its simplicity, the model succeeds to reproduce most of the featuresexpected from typical collisional events, and the energy and angular momentumtransfer during high velocity collisions is well behaved.

37

DISCOVERY OF BINARY SYSTEMWITHIN THE VESTA FAMILY ASTEROIDS

Authors: W. H. Ryan and E. V. RyanInstitutes: New Mexico Institute of Mining and Technology and New MexicoHighlands University (USA)

Photometric observations of the minor planet (3782) Celle, which has beenassociated both dynamically and spectroscopically with the Vesta asteroid fam-ily, were obtained using the 1.8-meter Vatican Advanced Technology Telescopeduring September 2001 and December 2002-January 2003. Analysis of thesedata reveals a normal rotational lightcurve (P=3.84 hr, amplitude = 0.10-0.15mag). During the 2002-2003 run, additional attenuation events were observedlasting for about 2.6-3.5 hrs that varied in amplitude from 0.15-0.3 mag. Theattenuations were of two distinct types that can clearly be identified as primaryand secondary occultation/eclipses similar to those that have been previouslyobserved in known minor planet binary systems (Pravec et. al. 2000, Icarus146, 190). We therefore interpret our data as clear evidence that (3782) Celle isactually an asynchronous binary system with an orbital period of 36.57 ± 0.03hrs. A preliminary model yields a primary-to-secondary diameter ratio of 0.42± 0.02 and an upper limit to the average bulk density of 2.0 ± 0.5 gm/cm3.This is indicative of a fractured or internal rubble-pile structure for at least one,if not both, of the binary components.

Since the Vesta family is believed to have been created via a cratering event,this finding has important implications for understanding possible ejecta reac-cumulation and satellite formation in subcatastrophic collisions. Further, theexistence of a stable binary system amongst such small bodies could place ageconstraints on the Vesta family-forming impact event.

This work is supported by NASA Planetary Astronomy Grant NAG5-8734and the generous observing time made available by the Vatican ObservatoryResearch Group.

38

DISRUPTION AND SPLITTINGOF COMETARY NUCLEI

Invited Review

Author: Paul WeissmanInstitute: Jet Propulsion Laboratory, Pasadena (USA)

Disruption and/or splitting events associated with cometary nuclei have beenobserved at an increasing rate in the past decade as the quality and quantityof cometary observations have increased. It is now recognized that splittingand disruption events may be a common process among comets and may be theprincipal physical loss mechanism for nuclei, far more so than outgassing and/orcollisions with other solar system bodies. Disruption and splitting events fallinto two categories: tidal and random. Tidal splittings are caused by passageof a nucleus through the Roche limit of the Sun or a planet. Classic examplesare comet Ikeya-Seki (C/1965 S1) for the Sun and comet Shoemaker-Levy 9(D/1993 F2) for Jupiter. The tidal forces break the nuclei into many pieces;these fragments may be the original, loosely bound cometesimals that cametogether to form the rubble-pile nuclei in the primordial solar nebula. Self-gravitation may then reassemble these fragments into multiple smaller nuclei,though not all fragments are incorporated in the new nuclei. Asphaug andBenz (Nature, 1994; Icarus, 1996) showed for comet Shoemaker-Levy 9 thatthis reassembly provides valuable clues to the bulk density of the cometesimals.Random disruption events occur, as the name implies, at random times in thecomets’ orbits and do not show any correlation with time of perihelion passage,perihelion distance, distance above or below the ecliptic, etc. Classic exam-ples are comet 3D/Biela (1852 III) and comet LINEAR (C/1999 S4). Randomdisruption events range from small, short-lived fragments breaking off a mainnucleus, to the creation of multiple returning comets, to total disruption anddisappearance of the comet. The physical explanation for random disruptionevents is lacking. Weissman (A & A, 1980) showed that random disruptionevents were observed about 10% of the time for dynamically new long-periodcomets, ≈ 4% of the time for returning long-period comets, and ≈ 1% of thetime for periodic comets (per orbit). The statistics suggest that the probabilitythat a comet splits or disrupts involves some intrinsic physical property of thenucleus, whereby comets that are likely to split do so early in their dynamicalevolution, and that comets resistant to random disruption survive far longerin both time and cumulative number of orbits. Disruption events are likelyrelated to cometary outbursts but again, the specific physical mechanism(s) isunknown. Some suggested mechanisms are: 1) rotational spin-up due to out-gassing torques; 2) sudden pressure release from pockets of volatile gases andices (more volatile than water); and 3) the transition of cold amorphous iceinto crystalline ice, which is an exothermic reaction. Impacts with solar systemdebris are likely too infrequent to provide a plausible explanation for disruptionevents. The current observational evidence and the suggested mechanisms willbe discussed. This work was supported by the NASA Planetary Geology &Geophysics Program and was performed at the Jet Propulsion Laboratory.

39