Flattened loose particles from numerical simulations compared to Rosetta collected particles
Jeremie Lasue, Isabelle Maroger, Robert Botet, Philippe Garnier, Sihane Merouane, Thurid Mannel, Anny-Chantal Levasseur-Regourd, Mark Bentley
AAstronomy & Astrophysics manuscript no. Lasue_agg_Rosetta_rev2_final c (cid:13)
ESO 2019March 5, 2019
Flattened loose particles from numerical simulations compared toRosetta collected particles
J. Lasue , I. Maroger , R. Botet , Ph. Garnier , S. Merouane , Th. Mannel , , A.C. Levasseur-Regourd , and M.S.Bentley IRAP, Université de Toulouse, CNRS, CNES, UPS, Toulouse, Francee-mail: [email protected] Université Paris-Saclay / Université Paris-Sud / CNRS, UMR 8502, LPS, Orsay, France Max Planck Institute for Solar System Research, Göttingen, Germany Space Research Institute of the Austrian Academy of Sciences, Graz, Austria University of Graz, Graz, Austria Sorbonne Université, CNRS, LATMOS, Paris, France European Space Astronomy Centre, Madrid, SpainReceived September 15, 1996; accepted March 16, 1997
ABSTRACT
Context.
Cometary dust particles are remnants of the primordial accretion of refractory material that occurred during the initial stagesof the Solar System formation. Understanding their physical structure can help constrain their accretion process.
Aims.
The in situ study of dust particles collected at slow speeds by instruments on-board the Rosetta space mission, includingGIADA, MIDAS and COSIMA, can be used to infer the physical properties, size distribution, and typologies of the dust.
Methods.
We have developed a simple numerical simulation of aggregate impact flattening to interpret the properties of particlescollected by COSIMA. The aspect ratios of flattened particles from both simulations and observations are compared to di ff erentiatebetween initial families of aggregates characterized by di ff erent fractal dimensions D f . This dimension can di ff erentiate betweencertain growth modes, namely the Di ff usion Limited Cluster-cluster Aggregates (DLCA, D f ≈ . ff usion Limited Particle-clusterAggregates (DLPA, D f ≈ . D f ≈ . D f ≈ . Results.
The diversity of aspect ratios measured by COSIMA is consistent with either two families of aggregates with di ff erent initial D f (a family of compact aggregates with fractal dimensions close to 2.5-3 and some flu ffi er aggregates with fractal dimensions around2). Alternatively, the distribution of morphologies seen by COSIMA could originate from a single type of aggregation process, suchas DLPA, but to explain the range of aspect ratios observed by COSIMA a large range of dust particle cohesive strength is necessary.Furthermore, variations in cohesive strength and velocity may play a role in the higher aspect ratio range detected ( > Conclusions.
Our work allows us to explain the particle morphologies observed by COSIMA and those generated by laboratoryexperiments in a consistent framework. Taking into account all observations from the three dust instruments on-board Rosetta, wefavor an interpretation of our simulations based on two di ff erent families of dust particles with significantly distinct fractal dimensionsejected from the cometary nucleus. Key words. comets: general – comets: individual: 67P / Churyumov-Gerasimenko – protoplanetary disks – accretion: accretion disk– methods: numerical – space vehicles: instruments
1. Introduction
Comets are believed to preserve pristine dust grains and to pro-vide information about their aggregation processes in the earlySolar System (e.g. Weidenschilling 1997; Blum 2000). Analysesof data from the Giotto mission to comet 1P / Halley and of foilimpacts and aerogel tracks retrieved by the Stardust mission inthe coma of comet 81P / Wild 2 have indeed given clues to thepresence of low density dust particles built up of agglomerates,possibly with di ff erent tensile strengths and porosities (e.g. Fulleet al. 2000; Hörz et al. 2006; Burchell et al. 2008). The interpre-tation of remote polarimetric observations of bright comets, suchas 1P / Halley and C / D f (Dominik & Tielens 1997; Kempf et al. 1999;Bertini et al. 2009). Understanding the structure of cometary dustparticles can give clues to these early Solar System processes(Blum & Wurm 2008; Fulle & Blum 2017). During its 26 month long rendezvous with comet67P / Churyumov-Gerasimenko (hereafter 67P) in its 2015apparition, the Rosetta spacecraft monitored the properties ofcometary dust particles released by the nucleus in the pre-and post- perihelion phases, as well as during some outburstevents. Three instruments were specifically devoted to the studyof dust particles: i) COSIMA (the COmetary Secondary IonMass Analyzer, Kissel et al. (2007)) collected dust particlesof 10 to 100 µ m size on 1 cm targets, imaged them with amicroscope operating under grazing incidence illumination Article number, page 1 of 9 a r X i v : . [ a s t r o - ph . E P ] M a r & A proofs: manuscript no. Lasue_agg_Rosetta_rev2_final with a resolution of about 14 µ m, and then analyzed themthrough a mass spectrometer after indium ion beam ablation,ii) MIDAS (the Micro-Imaging Dust Analysis System, Riedleret al. (2007)) collected micron-sized dust particles on targets ofabout 3 . , in order to obtain 3D images of their surfacesdown to tens of nanometers pixel resolution using atomicforce microscopy, and iii) GIADA (the Grain Impact Analyzerand Dust Accumulator, Colangeli et al. (2007)) measured theoptical cross-section, speed, momentum and cumulative flux ofhundreds of sub-millimeter sized dust particles.The COSIMA and MIDAS instruments collected dust par-ticles at velocities in the 1 to 15 m s − range (Fulle et al. 2015),that is to say at relative velocities much lower than the 6 . − reached during the collection of 81P / Wild 2 samples. Theirchemical properties were thus mostly preserved, as well as partof their physical structure. Some small particles, which couldbe fragments of fragile individual particles, were neverthelessnoticed (e.g. Bentley et al. 2016; Merouane et al. 2016). Inter-estingly enough, some particles appeared to be flattened, mostlikely as a result of impact alteration (e.g. Langevin et al. 2016;Mannel et al. 2016).The Rosetta dust experiments provide complementary in-sights into the properties of dust particles thanks to their dif-ferent approaches (see, for a review, Levasseur-Regourd et al.2018). As far as images are concerned, the total number of dustparticles detected is above 30,000 for COSIMA and above 1,000for MIDAS (Levasseur-Regourd et al. 2018; Güttler et al. sub-mitted).More specifically, all images of dust particles indicate thatthey consist of more or less porous agglomerates of smallergrains (following the classification introduced in (Güttler et al.submitted)). Their overall sizes, identified by well-definedboundaries, range from about 1 micrometer to tens of microm-eters for MIDAS, and from tens of micrometers to several hun-dreds of micrometers for COSIMA. The presence of aggregatedstructures at distinct scales suggests a hierarchical aggregation(Bentley et al. 2016). Indeed, the fractal dimension of a veryporous agglomerate detected by MIDAS was determined via adensity-correlation function (Mannel et al. 2016), to be equal to1.7 ± / Churyumov-Gerasimenko can thus be estimated to be atleast equal to 90% for very porous ones, and about 75% for morecompact ones (e.g. Blum & Wurm 2008; Bertini et al. 2009).The porosity of 67P / Churyumov-Gerasimenko’s dust particleshas been estimated to be around 60% based on the density ofthe nucleus and the composition measured by COSIMA (Fulleet al. 2017). Analysis of the reflectance of porous dust particlescollected by COSIMA indicate that a high porosity ( > Fig. 1.
Diversity of crushed particle types detected by COSIMA (Nick:compact particle, C; Alexandros: rubble pile, R; Estelle: shattered clus-ter, S; and Johannes: glued cluster, G) (adapted from Langevin et al.(2016)). themselves, they formed in the solar nebula and the primordialdisk (e.g. Davidsson et al. 2016; Blum et al. 2017), and werenever processed within large objects.
COSIMA collected and analyzed cometary particles ejected by67P / Churyumov-Gerasimenko on gold black covered targets(Kissel et al. 2007). The dust particles ejected by the comet im-pacted COSIMA targets at a speed <
10 m s − according to GI-ADA measurements (Rotundi et al. 2015) with a deceleration < × m s − according to Hornung et al. (2016). These val-ues are enough to damage the initial structure of the dust parti-cles during the collision, as visually assessed from the imagesacquired by COSISCOPE after collection. With a resolution of14 microns, the microscope enabled studies of particle typologyand flux (Langevin et al. 2016). The images show particles rang-ing from a few tens to several hundreds of microns, the majorityof which appears to be built of micron-sized sub-components, asconfirmed by MIDAS (Bentley et al. 2016). Analysis of the par-ticle morphologies identified four families of particles (Langevinet al. 2016) which fall into two major classes, compact and clus-tered. These families are :1. Compact (type C) particles present well-defined bound-aries without smaller satellite particles and with an apparentheight above the collecting plane of the same order of mag-nitude as their horizontal ( x and y ) dimensions.2. Shattered cluster (type S) particles are defined by clusters offragments for which no individual fragment makes up a ma-jor fraction of the initial particle. These particles are inter-preted as rearrangement of fragments within the impactingparticle without associated disruption.3. Glued cluster (type G) particles have a well-defined shapeand a complex structure where sub-components appear to belinked by a fine-grained matrix with a smooth texture.4. Rubble piles (type R) particles comprise components muchsmaller than their apparent size. Upon collision with the Article number, page 2 of 9. Lasue et al.: Flattened loose particles compared to Rosetta
Fig. 2.
Probability density of aspect ratio for each type of particles de-tected by COSIMA (adapted from Langevin et al. (2016)). plate, the sub-components rearranged themselves in a flat-tened conical pile with many satellite components indicatingpoor cohesion.The di ff erent types of particles collected by COSIMA are illus-trated in Fig. 1.The grazing incidence illumination provided by COSIS-COPE allows both the surface area of the collected particles andtheir height (based on their projected shadow) to be determined(see Fig. 1). The area is determined from the ratio of bright pixelsbefore and after exposure to the dust flux from the comet. An as-pect ratio of the compacted particles can be obtained from height √ area .The aspect ratio density distribution for each detected particletype is shown in Fig. 2. The compact particles, C, appear unbro-ken and present the largest aspect ratios, with a first peak around0.5 and another close to 1. The other particles present typicalaspect ratios of around 0.3 with the shattered clusters being theflattest type of agglomerates. To understand the physical struc-ture of cometary nuclei, it is important to infer, as far as possible,properties of dust particles prior to their collection. COSIMAanalyses have shown a correlation between the flux of dust par-ticles at various distances from the comet nucleus and their mor-phology (Merouane et al. 2016). The fragmenting particles ap-pear to have a mechanical strength of a few 1000 Pa (Hornunget al. 2016) and their morphological diversity could result fromdi ff erent collection speeds in the range from 1 m s − to 6 m s − asinvestigated by laboratory simulations (Ellerbroek et al. 2017).In this work, we investigate if di ff erent dust particle struc-tures prior to their collection can also lead to the di ff erent mor-phologies found by the Rosetta dust instruments. We present aset of numerical simulations of fractal aggregates flattening onimpact with a plane surface, before presenting its results anddiscussing their implications for the interpretation of the Rosettameasurements.
2. Method
We expect the dust particles aggregating in the solar nebula topresent fractal structures. Fractal aggregates in the early SolarSystem form a diversity of porosities that can be representedby their fractal dimension, D f , based on their aggregation pro-cesses (Wurm & Blum 1998). Aggregation simulations consider Fig. 3.
3D representation of four aggregates representing the four di ff er-ent aggregation processes considered in this work. DLCA ( D f ≈ . D f ≈ . D f ≈ .
5) and RLPA ( D f ≈ = blue, y = red, z = green). the collisions of spherical monomers which represent individualgrains aggregating to form dust particles (Güttler et al. submit-ted). Four main aggregation processes, leading to significantlydi ff erent fractal dimensions, are considered: DLCA ( D f ≈ . D f ≈ . D f ≈ .
5) and RLPA ( D f ≈ ff usion Limited models, in which when onemonomer meets another one it sticks directly to it. The RL mod-els are Reaction Limited models, in which molecular reactionsoccur when two monomers encounter each other and result inthem sliding with respect to one another in order to maximizethe number of bonds, resulting in a more compact aggregate.CA stands for Cluster-cluster Aggregation and PA for Particle-cluster Aggregation : in PA particles, monomers are added tothe same main cluster which accretes all the mass and is rela-tively compact, whereas in CA particles, monomers form sep-arate clusters which then aggregate, thus resulting in a smallerfractal dimension of the aggregate. The PA process occurs whenthe number of monomers compared to the available volume ishigh, increasing the chances of collision amongst small aggre-gates.Depending on the physical conditions of the primordial pro-tosolar nebula, in terms of dust to gas ratio and dust compo-sition, we can expect each of these kinds of aggregates to beformed (Weidenschilling 1997; Kimura 2001). They have alsoeach been produced by computer simulations and laboratoryexperiments simulating the initial stages of planetary accretion(Meakin 1991; Blum & Wurm 2008). In a first step, 3D o ff -lattice aggregates of a number N =
10 000identical spherical particles (called monomers) are generated ac-cording to the 4 di ff erent aggregation processes described above.The resulting fractal aggregates are characterized by di ff erent Article number, page 3 of 9 & A proofs: manuscript no. Lasue_agg_Rosetta_rev2_final
Fig. 4.
Representation of the collision geometry for 2 superposedmonomers. If θ < θ , then the superposition remains stable. When θ > θ the cohesive link between the monomers is broken and the uppermonomer will bounce randomly following some of the green arrows andwill attach itself to the z = initial fractal dimensions according to the approximate relation-ships: DLCA ( D f ≈ . D f ≈ . D f ≈ . D f ≈ N m , of monomers constituting the aggregate located within asphere of radius R follows N m ∝ R D f , where R is smaller than thegyration radius of the aggregate. The gyration radius of a frac-tal aggregate is a measure of the extent of the aggregate, akin tothe standard deviation of the monomers’ distance to the centreof mass of the aggregate and can be calculated by R g = N × (cid:88) i , j ( r i − r j ) = N × (cid:88) i ( r i − r c ) (1)where N is the number of monomers in the aggregate, r i and r j are the spatial coordinates of the center of the monomers i and j , and r c corresponds to the spatial coordinates of the centerof mass of the aggregate (Jullien & Botet 1987). A representa-tion of each of the four aggregate types is given in Fig. 3. Theseaggregates may correspond to di ff erent types of cometary parti-cles as ejected from the surface of the nucleus by gas pressure.For each aggregate type, 1000 di ff erent aggregation simulationswere performed to statistically analyze the results.In a second step, simulating the particle collection and flat-tening observed by COSIMA during the Rosetta mission, theaggregates are projected monomer by monomer onto the plane z = z values. If a monomer is projected directly ontothe plane z = E VdW < E k where E VdW is the van der Waals energy and E k is the kinetic energy of the incoming particle. This conditioncan be written as in Eq (2) considering van der Waals interac- tions between the two spherical elements. E VdW = A H r d ≤ E k = π r ρ (sin θ ) V (2)(sin θ ) = A H π r ρ dV (3)where E VdW is the van der Waals energy, A H is the Hamakerconstant for the material considered, r is the radius of a sin-gle monomer, d is the diameter of a monomer, E k is the kineticenergy of the monomer, ρ is the density of the monomer, θ isthe angle between the direction linking the two centers of themonomers and the vertical direction, as illustrated in Fig. 4 and V is the velocity of the aggregate with respect to the collectingsurface z = d , is slightly larger (by0 . ff erence negligible ( < d . The Hamaker constant of two particlesinteracting corresponds to a measure of the relative strength ofthe particles material with respect to the attractive van der Waalsforces between them (Hamaker 1937).We call θ the angle θ for which Eq (2) is an equality. sin θ is a threshold above which monomers may break their bondsand bounce. Changing this parameter can either be viewed aschanging the cohesive strength between monomers or as chang-ing the collection velocity, as Eq (3) shows. Thus, with theHamaker constant of dry minerals under vacuum conditions A H ≈ × − J (Israelachvili 2011): – sin θ ≈ V = − and r = . µ m or for A H valueshigher than 1 × − J) – sin θ ≈ − corresponds to all bonds being broken, rel-atively higher collection speed, or larger monomer sizes(value typically obtained for V =
10 m s − and r = . µ m)So, if a projected monomer meets another monomer, we cancompute a collision parameter ( ν = sin θ sin θ ). If ν ≤
1, the monomersticks to the one it bumps into. If ν > z = P loss is introduced. In this case, if a monomermeets the condition ν >
1, then it will be removed from thesimulation with the mass loss probability P loss which matchesthe chance that some monomers do not stick to any others andbounce back to free space during the collision.An illustration of the e ff ect of changing the value for sin ( θ )is given in Fig. 5 where the morphology of flattened aggre-gates is clearly dependent upon the initial structure of the ag-gregates and the geometric parameters. Under conditions wheremost bonds are broken (sin θ ≈ − ), the more compact aggre-gates appear to generate a small pyramid of monomers with anangle of repose. The more porous the aggregates, the flatter ap-pears to be the projection. In the case where bonds are unbroken(sin θ ≈ Article number, page 4 of 9. Lasue et al.: Flattened loose particles compared to Rosetta F i g . . F i gu r e s u mm a r i s i ng t h ee ff ec t o f i n iti a l p a r ti c l e m o r pho l ogy ( D f ) a ndbond c oh e s i onon t h e m o r pho l ogyo f t h e fl a tt e n e dp a r ti c l e s . A s ca l e i s g i v e n i nnu m b e r o f m ono m e r s , a nd a r e f e r e n ti a l fr a m e i s i nd i ca t e dby c o l o r e d a rr o w s ( x = b l u e , y = r e d , z = g r ee n ) . and increase the relative height of the flattened aggregate. Thesecolumns of monomers appear due to the increased strength of thebonds between the monomers, forming chain-like vertical struc-tures that are not broken by the flattening geometry (as θ > θ over the monomers’ column). We therefore see that both param-eters ( D f and θ ) influence significantly the outcome of the sim-ulated projection. Fig. 6 shows the resulting projections in the case where themonomers have a non-zero probability to bounce back to spacedue to mass loss processes. In this case we only consider RLPAaggregates with di ff erent P loss values ranging from 0% to 50%.As the mass loss probability gets larger, only a flat footprint ofthe aggregate remains with a very low aspect ratio which canrepresent the results of the low speed laboratory aggregate stick- Article number, page 5 of 9 & A proofs: manuscript no. Lasue_agg_Rosetta_rev2_final
Fig. 6.
3D view of a RLPA aggregate after projection with di ff erentmass loss. A scale is given in number of monomers, and a referentialframe is indicated by colored arrows (x = blue, y = red, z = green) with anapproximate 30 ◦ viewing angle. ing experiments of Ellerbroek et al. (2017) where most of theinitial aggregate mass was lost. Such mass loss processes mayalso be at work during the COSIMA particle collection.
3. Results
The aggregate flattening simulations were run to create 1000 ag-gregates of each of the four types, using 10,000 monomers each,for the four fractal dimensions considered, with the sin( θ ) pa-rameter ranging from 1 to 10 − and with a P loss probability ofmass loss ranging from 0 to 0.5. This was done in order to ob-tain good statistics for the aspect ratio of each numerically flat-tened aggregate for comparison to the COSIMA measurements.The height, H , of the flattened aggregates is the maximum valueof z among all the sticking monomers. To compute the area, A ,we considered only the monomers visible from above (lookingtowards the − z direction) and, based on their position, we cal-culated the contour of the projected connected set of monomers(Lorensen & Cline 1987). We computed two di ff erent connectedareas: one with gaps and one without gaps as Fig. 7 shows. Thearea with gaps is always somewhat smaller than the area with-out gaps but is essentially linearly correlated with it. Therefore,we calculated the results based on the connected area withoutgaps. In this way, we can calculate a statistical distribution ofthe aspect ratio, H / √ A , for particles of each kind similar to theprocedure used with the COSIMA data and assess the e ff ect ofthe di ff erent parameters on the morphology of the flattened ag-gregates. Figure 8 represents the density distribution of aspect ratios cal-culated for the 1000 flattened aggregates of each fractal type andfor 4 di ff erent values of sin( θ ). The upper figure is calculatedfor a sin( θ ) = Fig. 7.
Illustration of the connected area calculated for a flattened ag-gregate of type RLCA. The flattened particles seen from above is shownon the left. The calculated connected areas are shown with gaps in themiddle and without gaps on the right. The parts of the aggregate that arenot connected to the largest connected aggregate are removed from theprocessing. between monomers is broken (illustrated on the right hand sideof the Figure 5). One can see that the distribution of aspect ratiosoverlaps between about 0.5 (relatively flat aggregates) and 1.3.The distribution also separates relatively well the di ff erent typesof aggregates with the more compact aggregates of type PA hav-ing a median aspect ratio value of 1.18 ( σ = .
15) and the flu ffi eraggregates of type CA having a median aspect ratio value of 0.73( σ = . θ ) decreases, the number of broken bonds increasesand the projected aggregates get flatter. The minimum aspectratio decreases and reaches 0.1 for values of sin( θ ) = . D f ≈
3) is now clearly separated from the distribution ofthe other aggregates and remains around 1, indicating that thesurface dimensions covered by the flattened aggregate in x and y are of the same order of magnitude as its vertical extent in z .With respect to the distributions of the less compact aggregates,we notice that the distributions for CA aggregates with fractal di-mensions lower than about 2 become quickly undistinguishable.Those flattened aggregates would therefore present essentiallythe same aspect ratio distributions irrespective of their initialmorphology. The DLPA aggregates that have a fractal dimen-sion around 2.5 are located in between those two extremes andclearly separated from them at low values of sin( θ ). For exam-ple, the standard deviation of the distributions for sin( θ ) = . θ ) = . θ ) lower than 0.1, the density distributions stabilize to-wards their final values. One can also notice a bimodal densitydistribution for the flattened RLPA aggregates, corresponding towhether vertical columns of monomers appear within the pyra-mid somewhat extending its height. We expect the random sizedistributions of monomers in real dust aggregates to limit the as-pect ratio to the lower values of around 0.75-1.0. Some similarlinear chain-like structures were also detected in the analysis ofCOSIMA particles, such as the 2CF Adeline particle (Hornunget al. 2016).The e ff ect of the sin( θ ) parameter is further illustrated inFigure 9 where aspect ratio distributions of DLPA aggregates are Article number, page 6 of 9. Lasue et al.: Flattened loose particles compared to Rosetta
Fig. 8.
Distribution of aspect ratio, H / √ A , for 1000 aggregates of eachtype (RLCA, DLCA, RLPA, DLPA) with sin( θ ) ranging from 1 to0.001 without mass loss. Fig. 9.
Distribution of aspect ratio, H / √ A , for DLPA aggregates withdi ff erent sin ( θ ) without mass loss. calculated for di ff erent values of sin( θ ) ranging from 0.001 to 1and are superposed. As the sin( θ ) value decreases, the aspect ra-tio decreases (due to the larger number of bonds breaking) fromapproximately 1 to 0.25. One can also notice that the standarddeviation of the density distribution also decreases, indicatingthat most aggregates of this type flatten in the same way. This isrelated to the randomization of monomer deposition after bondbreaking which reduces the range of vertical extent possible afterflattening. Based on this figure, we can see that given a relativelynarrow range of collection velocities, since equation 3 indicatesthat sin( θ ) is proportional to V , a large range of bond cohesivestrengths in the aggregates would lead to a larger range of aspectratios for the same initial structure of the aggregate. This is espe-cially true of DLPA as the aspect ratios for these particles rangefrom 0.25 to 1.2. Compact aggregate aspect ratios would rangebetween 0.8 and 1.3, while aggregates with fractal dimensionsaround 2 and lower present aspect ratios ranging from 0.1 to 1.If the cohesive strength of monomer bonds in the aggregate arerandomly distributed one can expect to detect more aggregateswith small flattened aspect ratios than large flattened aspect ra-tios.Finally, the e ff ect of the mass loss coe ffi cient is illustrated inFigure 10 top where the probability density of aspect ratios forRLPA aggregates with sin ( θ ) = θ ), one can notice thatthe end aspect ratio distribution remains relatively large (largerthan 0.5) which is due to the simultaneous loss of monomers inall directions, so that the dimensions of the flattened aggregateare reduced in all dimensions at more or less the same rate (in x , y , and z ). This parameter is also important in reducing the finalaspect ratio of the flattened aggregates.In the case of DLPA aggregates and the more flu ff y ones, theinitial aggregate is so porous that even moderate mass loss de-stroys the structure during flattening. This leads rapidly to a veryflat final projected structure as illustrated in Figure 10 bottom.It therefore appears that the initial fractal dimension of ag-gregates strongly a ff ects the morphology of flattened aggregates,and that, depending on the e ff ect of parameters such as the speedof collection, strength of bonds between the monomers and massloss fraction, it may, or may not, be possible to distinguish theinitial structure of the particle from their flattened morphologies. Article number, page 7 of 9 & A proofs: manuscript no. Lasue_agg_Rosetta_rev2_final
Fig. 10.
Distribution of aspect ratio, H / √ A , for RLPA aggregates (top)and DLPA aggregates (bottom) with sin ( θ ) = ff erent mass lossprobability coe ffi cients.
4. Discussion
The aspect ratio variation with initial D f (aggregate type), sin ( θ ) and P loss can be compared with the values observed byCOSISCOPE and presented in Fig. 2. On the one hand, only thePA aggregate types have an aspect ratio large enough to explainthe presence of the compact particles in the COSIMA aspect ra-tio distribution. This implies that a population of particles withfractal dimension between 2.5 and 3 must be present in the dis-tribution of particles ejected by 67P.On the other hand, in order to explain the presence of mor-phologies with aspect ratios as low as 0.1 to 0.3, where the dis-tributions of COSIMA particles of type G, R and S peak, othertypes of particles or processes need to be invoked. From our sim-ulations, even with a mass loss as large as 50%, RLPA aggregatesalone cannot explain the range of aspect ratio observed. How-ever, the DLPA type particles could reach aspect ratio values aslow as 0.2 either with di ff erent cohesive strengths and / or veloc-ities (sin( θ )) or with mass losses up to 50%. Finally, a fractaldimension lower than 2 would also lead to very low final as-pect ratios even when considering particles with higher cohesivestrengths. The large range of distribution observed by COSIMAcould therefore be explained by:1. two di ff erent initial groups of particles with low and highfractal dimensions (such as RLPA for the compact particlesand DLPA for the shattered clusters). 2. the flattest kind of particles observed (shattered clusters withan aspect ratio around 0.15) could be consistent with com-paction of the smallest fractal dimension RLCA and DLCAaggregates or with a very large mass loss during collection( > ff erent cohesive strengths amongst aggregates( sin ( θ ) ranging from at least 0.1 to 1). This distributionwould also present a peak around 0.3 as shown in Figure 9,which would be consistent with the peak of the COSIMAdistribution around 0.3 as shown in Figure 2.4. finally, a fourth process, described in Ellerbroek et al. (2017),may be playing a role here as well. Experiments show thatincoming aggregates may sometimes fragment upon impact,leaving some remains sticking to the target in a pyramidalshape (mass transfer property between 0 and 0.8).The diversity of aspect ratios observed appears consistentwith at least two families of aggregates with di ff erent D f , whichwould also be consistent with the GIADA and MIDAS measure-ments of two dust particles populations with very di ff erent frac-tal dimensions, one being close to 3 and the other around 1.8(Fulle & Blum 2017; Mannel et al. 2016)). Variations in boththe cohesive strength of the particles and the speed of collectionmay play a role in the continuity of the higher aspect ratio range( > In the work of Ellerbroek et al. (2017), laboratory simulations ofimpacts of aggregates simulating the particle collection proce-dure of Rosetta were presented. The aggregates were formed byaggregation of irregular polydisperse SiO particles with densityaround 2 . − and a size range of 0.1 to 10 µ m. The finalaggregates have porosities around 65% ±
5% and low compres-sive strength between 1 × Pa and 1 × Pa. The aggregateswere then accelerated by electrostatic forces towards a collectingplane where the collision was filmed and the resulting flattenedfootprint imaged and analyzed. The velocity of impact rangesfrom about 1 m s − to 6 m s − .The footprints obtained represent the diversity of morpholo-gies that were acquired by the COSIMA instrument. At very lowvelocities of around 1 m s − , the aggregates either stick directlyto the surface, similar to the compact COSIMA particle type,or they may bounce from the surface, leaving a very flat foot-print with mostly unconnected fragments, possibly morphologi-cally similar to the shattered cluster COSIMA type of particles.As velocities are increased from 2 m s − to 6 m s − , the particlesmostly stick to the surface and fragmentation occurs, leading tofootprints morphologically similar to COSIMA rubble piles orglued clusters.In this laboratory work, all morphologies were generated us-ing only a change in the impact velocity and impactor size, andsimilarities could be seen between the footprints of the parti-cles that were obtained on the collecting surface and the mor-phologies measured by COSIMA. The simulations presented inour work allow us to generate similar conditions of flattening byvarying the velocity and the particles sizes. However, in our sim-ulations, we can also modify the initial impacting particle mor-phology and study its e ff ect on the flattening of the aggregates. Article number, page 8 of 9. Lasue et al.: Flattened loose particles compared to Rosetta
This allows us to explore an extended set of parameters com-pared with the laboratory experiments, and we have shown thatit is also possible to generate the measured footprint morphologyby considering di ff erent initial fractal dimensions of the impact-ing particles, as discussed above. It would be of interest to studyin the laboratory how very porous particles behave when sub-jected to the type of collection that happened during the Rosettamission to confirm our analysis. A planned future study aims to investigate whether these resultsare also valid for MIDAS particles. The aspect ratios of dust par-ticles collected by MIDAS should be calculated and their distri-bution reviewed. It will be of great interest if the distribution fallsin di ff erent groups, and if they match those found in the simu-lation and with COSIMA particles. As MIDAS particles are oneorder of magnitude smaller than those of COSIMA, this will al-low us to understand how the initial structures of dust particles ofcomet 67P might look and if they remain similar over the 1 µ mto 100 µ m size range.
5. Conclusions
In this work, we have shown that simple numerical simulationsof aggregate flattening can be used to infer the initial proper-ties of particles collected by COSIMA on-board Rosetta. Thediversity of aspect ratios measured in COSIMA images appearsconsistent with several hypotheses on the initial properties of thecollected particles.1. It could be explained by at least two families of aggregateswith di ff erent fractal dimensions D f . A mixture of somecompact particles with fractal dimensions close to 2.5-3 to-gether with some flu ffi er ones with fractal dimensions < D f ≈ .
5) but presenting a largerange of cohesive strengths or collection velocities. This dis-tribution would be consistent with a maximum at an as-pect ratio around 0.3 as observed on the COSIMA typol-ogy (Langevin et al. 2016).Furthermore, variations in cohesive strength and velocity mayplay a role in the higher aspect ratio range detected by COSIMA( > ff erent families of dust particles with sig-nificantly distinct fractal dimensions ejected from the cometarynucleus. Acknowledgements.
The authors acknowledge two anonymous referees for theirpositive evaluation and constructive comments. The authors acknowledge sup-port from Centre National d’Etudes Spatiales (CNES) in the realization of in-struments devoted to space exploration of comets and in their scientific analysis.T.M. acknowledges funding by the Austrian Science Fund FWF P 28100-N36.
References
Bentley, M. S., Schmied, R., Mannel, T., et al. 2016, Nature, 537, 73 Bertini, I., Gutierrez, P. J., & Sabolo, W. 2009, Astronomy & Astrophysics, 504,625Blum, J. 2000, Space Science Reviews, 92, 265Blum, J., Gundlach, B., Krause, M., et al. 2017, Monthly Notices of the RoyalAstronomical Society, 469, S755Blum, J. & Wurm, G. 2008, Annu. Rev. Astron. Astrophys., 46, 21Burchell, M. J., Fairey, S. A., Wozniakiewicz, P., et al. 2008, Meteoritics & Plan-etary Science, 43, 23Colangeli, L., Lopez-Moreno, J. J., Palumbo, P., et al. 2007, Space Science Re-views, 128, 803Davidsson, B. J. R., Sierks, H., Guettler, C., et al. 2016, Astronomy & Astro-physics, 592, A63Dominik, C. & Tielens, A. 1997, The Astrophysical Journal, 480, 647Ellerbroek, L. E., Gundlach, B., Landeck, A., et al. 2017, Monthly Notices of theRoyal Astronomical Society, 469, S204Fulle, M. & Blum, J. 2017, Monthly Notices of the Royal Astronomical Society,469, S39Fulle, M., Della Corte, V., Rotundi, A., et al. 2017, Monthly Notices of the RoyalAstronomical Society, 469, S45Fulle, M., Della Corte, V., Rotundi, A., et al. 2015, The Astrophysical JournalLetters, 802, L12Fulle, M., Levasseur-Regourd, A. C., McBride, N., & Hadamcik, E. 2000, TheAstronomical Journal, 119, 1968Fulle, M., Marzari, F., Della Corte, V., et al. 2016, The Astrophysical Journal,821, 19Güttler, C., Mannel, T., Rotundi, A., et al. submitted, Astronomy & AstrophysicsHamaker, H. C. 1937, physica, 4, 1058Hornung, K., Merouane, S., Hilchenbach, M., et al. 2016, Planetary and SpaceScience, 133, 63Hörz, F., Bastien, R., Borg, J., et al. 2006, science, 314, 1716Israelachvili, J. N. 2011, Intermolecular and surface forces (Academic press)Jullien, R. & Botet, R. 1987, Ann. Telecomm., 41, 343Kempf, S., Pfalzner, S., & Henning, T. K. 1999, Icarus, 141, 388Kimura, H. 2001, Journal of Quantitative Spectroscopy and Radiative Transfer,70, 581Kissel, J., Altwegg, K., Clark, B. C., et al. 2007, Space Science Reviews, 128,823Langevin, Y., Hilchenbach, M., Ligier, N., et al. 2016, Icarus, 271, 76Langevin, Y., Hilchenbach, M., Vincendon, M., et al. 2017, Monthly Notices ofthe Royal Astronomical Society, 469, S535Lasue, J., Levasseur-Regourd, A. C., Hadamcik, E., & Alcou ff e, G. 2009, Icarus,199, 129Levasseur-Regourd, A.-C., Agarwal, J., Cottin, H., et al. 2018, Space sciencereviews, 214, 64Levasseur-Regourd, A.-C., Zolensky, M., & Lasue, J. 2008, Planetary and SpaceScience, 56, 1719Lorensen, W. E. & Cline, H. E. 1987, in ACM siggraph computer graphics,Vol. 21 (ACM), 163–169Mannel, T., Bentley, M. S., Schmied, R., et al. 2016, Monthly Notices of theRoyal Astronomical Society, stw2898Meakin, P. 1991, Reviews of Geophysics, 29, 317Merouane, S., Zaprudin, B., Stenzel, O., et al. 2016, Astronomy & Astrophysics,596, A87Riedler, W., Torkar, K., Jeszenszky, H., et al. 2007, Space Science Reviews, 128,869Rotundi, A., Sierks, H., Della Corte, V., et al. 2015, Science, 347, aaa3905Weidenschilling, S. J. 1997, Icarus, 127, 290Wurm, G. & Blum, J. 1998, Icarus, 132, 125e, G. 2009, Icarus,199, 129Levasseur-Regourd, A.-C., Agarwal, J., Cottin, H., et al. 2018, Space sciencereviews, 214, 64Levasseur-Regourd, A.-C., Zolensky, M., & Lasue, J. 2008, Planetary and SpaceScience, 56, 1719Lorensen, W. E. & Cline, H. E. 1987, in ACM siggraph computer graphics,Vol. 21 (ACM), 163–169Mannel, T., Bentley, M. S., Schmied, R., et al. 2016, Monthly Notices of theRoyal Astronomical Society, stw2898Meakin, P. 1991, Reviews of Geophysics, 29, 317Merouane, S., Zaprudin, B., Stenzel, O., et al. 2016, Astronomy & Astrophysics,596, A87Riedler, W., Torkar, K., Jeszenszky, H., et al. 2007, Space Science Reviews, 128,869Rotundi, A., Sierks, H., Della Corte, V., et al. 2015, Science, 347, aaa3905Weidenschilling, S. J. 1997, Icarus, 127, 290Wurm, G. & Blum, J. 1998, Icarus, 132, 125