SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Experimental Procedure
  5. Results and Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Abstract– We present results of a set of impact experiments designed to examine the effects of impacts onto rocky blocks resting on and embedded within regoliths. The targets were approximately 500 g granodiorite blocks, struck with one-eighth inch aluminum spheres at nominal speeds of approximately 5 km s−1. The granodiorite blocks were emplaced in 20–30 grade silica sand to simulate an asteroidal or lunar regolith; block burial depths ranged from resting flush on the surface to submerged completely below the surface. We observe a trend for largest remnant mass to increase with block burial depth. Documentary still image and high-speed video of the resulting block fragments and surrounding regolith reveal new insights into the morphologies of blocks and secondary craters observed on asteroids like 433 Eros.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Experimental Procedure
  5. Results and Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Regoliths on the solid surfaces of airless solar system bodies like asteroids and the Moon are seen to be littered with blocky fragments of various sizes and levels of degradation and burial. Some of these blocks originated as ejecta fragments excavated during cratering events and ballistically emplaced in their present resting locations. Others may have been exhumed from deeper below the regolith surface as a result of seismic processes or have rolled or slid down slope from topographically higher positions because of mass wasting. Regardless of their provenance, once exposed at the surface the blocks themselves are exposed to the same impact environment as is the surrounding finer-scale regolith.

A prominent feature of the surfaces of the two asteroids imaged at the highest resolution, Eros and Itokawa, is that blocks, rocks, and related positive relief features become increasingly common at sizes ≤10 m, whereas craters become less common (Chapman et al. 2002; Saito et al. 2006). The degradation and disappearance of small craters has been attributed to several possible causes, most notably seismic shaking of the regolith (Richardson et al. 2005). Nevertheless, it remains the case that if a substantial fraction of a body’s surface is occupied by rocks, impacting projectiles must frequently strike rocks rather than fine regolith. While there have been descriptions of the evolution of rocks on the lunar surface subject to micrometeoritic bombardment (cf. Gault et al. 1972) and simulations of rock lifetimes in such a catastrophic disruptive environment (cf. Hörz et al. 1975), we lack a comprehensive understanding of how a blocky surface or regolith evolves under continuing impact conditions. For example, the blocks tend to “armor” the surface, preventing the creation of craters in the regolith that would otherwise be produced. On the other hand, impacted and fragmented blocks can produce a variety of what we term “recoil craters” and secondary craters of various types, as we will describe. A full understanding of the nature of the surfaces of atmosphereless, rocky bodies requires further research.

Figure 1 shows an example of a broken block on the surface of Eros resting within a crater-like depression. This block has previously been interpreted to be an ejected fragment broken upon landing and excavating a secondary, low-speed impact crater (Asphaug 2003). An alternative interpretation is that this might instead be a block broken in place by an impacting projectile that struck and fragmented the block, with the surrounding crater representing finer-scale regolith displaced by the impact and fragmentation of the block, a process we describe and illustrate below.

image

Figure 1.  A broken block (four ∼100 m diameter fragments) resting in an ∼700 m diameter crater on Eros. NEAR image 136819148.

Download figure to PowerPoint

Motivated by these considerations, we have begun a series of impact experiments to examine in detail the range of morphologic expression induced in regoliths because of the impact fragmentation of blocks on and embedded in the surface.

Experimental Procedure

  1. Top of page
  2. Abstract
  3. Introduction
  4. Experimental Procedure
  5. Results and Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

We conducted a preliminary, proof-of-concept set of impact experiments at the Ames Vertical Gun Range (AVGR) from February 25–27, 2009. The results of those experiments are presented here. The targets were six blocks of granodiorite cut into cubes roughly 5.75 cm on a side, with masses of approximately 500 g. Although basalt might have been a better block material to use for comparison with other impact experiments in the literature, the granodiorite cubes were readily available as “leftovers” from a previous experiment. However, neither the terrestrial basalt samples nor the granodoriate are likely to be particularly good analogs for the blocks on asteroid surfaces. Typical terrestrial basalt samples have nearly zero porosity, while the ordinary chondrite meteorites, believed to sample some S-type asteroids like Eros and Itokowa, and basaltic meteorites from presumed asteroid parents, have porosities of approximately 10% (Britt and Consolmagno 2003). Hypervelocity impact experiments indicate that it requires about twice as much energy per unit target mass to produce similar disruption results on the ordinary chondrite meteorites as terrestrial basalts, presumably because some of the impactor energy is used to compress rather than disrupt the target (Flynn and Durda 2004).

The granodiorite blocks were emplaced in 20–30 grade silica sand to simulate an asteroidal or lunar regolith and to provide a particulate background sufficiently texturally distinct from the granodiorite fragments to allow ready sorting of the block debris after the impacts. The sand filled a circular tub approximately 0.65 m in diameter and 0.15 m deep that was placed within the AVGR impact chamber. The sand was raked level, smooth, and flush with the rim of the tub prior to each impact.

The blocks were buried at various depths Dbur within the sand “regolith,” ranging from resting on the surface to submerged completely beneath the surface with some regolith overburden on top of the block. Burial depths are defined relative to the block dimension, with Dbur = 0 for a block resting on top of the regolith, Dbur = 0.5 for a block buried in the regolith to a depth half its own dimension, and Dbur = 1 for a block buried with its top face flush with the surface. In each case, the block was carefully leveled, aligned with the up-range face of the block normal to the incoming azimuth of the projectile, and the surrounding sand regolith raked level and flush to the base of the block. Figure 2 shows an example of the preshot configuration for one of the blocks.

image

Figure 2.  Preshot configuration of one the granodiorite blocks used in the impact experiments. This is the block from shot 090230, with Dbur = 0.25. The dot of light at the center of the top face of the block is laser illumination bore sighted with the barrel of the AVGR, used to assure proper targeting for the shot. The image is taken from the downrange direction.

Download figure to PowerPoint

The impact conditions were kept constant for each shot. Projectiles were one-eighth inch (0.3175 cm) aluminum spheres with masses of approximately 0.045 g, fired at nominal speeds of about 5 km s−1, approximating the mean relative impact speeds between Main Belt asteroids (Bottke et al. 1994). The impact angle was chosen to be 45°, the statistically most likely impact angle (Shoemaker 1962). The blocks were positioned for each shot so that the projectile impact point was at the center of the top face of the block.

Documentary still imagery was obtained of each block and its setting within the surrounding regolith prior to and after each shot. High-speed video of each impact was obtained by five different video cameras (two Phantom V10s, one Phantom V12.1, and two Shimadzu HPV-1s), with frame rates ranging from 1900 to 125,000 frames per second, to aid interpretation of the fragmentation mode of the blocks and the manner in which the regolith was disturbed and displaced around the blocks.

Results and Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Experimental Procedure
  5. Results and Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

The impact conditions and resulting largest fragment mass for each shot are presented in Table 1. As our focus was to begin to build a body of experimental data to study the range of morphologies of boulders fragmented in place on asteroidal and lunar surfaces, we chose the impact energy that would substantially fracture a block when resting on the regolith surface without completely disrupting and dispersing the fragments. The impacts were all subcatastrophic (i.e., largest remnant mass ratios are >0.5), with impact specific energies of approximately 1070 J kg−1. For comparison, Housen and Holsapple (1999) measured critical impact specific energies of about 1500 J kg−1 for 3.2 cm scale granite targets.

Table 1.   Impact conditions for the six impact experiments conducted in this study.
AVGR shot no.Depth of burial (Dbur)Block mass (Mb), gLargest remnant mass (Mlr), gLargest remnant mass ratio (Mlr/Mb)Impactor speed (Vimp), km s−1Impactor mass (Mimp), gSpecific energy (Q*), J kg−1
  1. Note: AVGR = Ames Vertical Gun Range.

0902270.004773570.7484.8990.04511134
0902300.254883280.6724.8320.04501077
0902280.505113950.7734.8520.04501037
0902310.754684120.8804.7360.04481074
0902291.004964630.9334.7150.04501008
0902321.255125121.0004.9640.04501083

The fragments from each block impact were collected from the impact chamber or sifted from the sand regolith and then weighed. As these were subcatastrophic impacts, the largest remnant in each case was the substantially intact “core” of the original target block. Prior to the experimental sequence, we hypothesized that there would be a straightforward relationship between the largest remnant mass and the initial block burial depth. Figure 3 shows that, as expected, there is indeed a rather strong trend for largest remnant mass to increase with block burial depth. There is an apparent “reversal” of this trend at Dbur = 0. It is likely that this is simply the result of the vagaries of statistical variations in the mode of fragmentation from shot to shot but until we can perform a more extensive series of shots the possibility remains that this may be a real burial depth-dependent effect because of some as-yet undocumented interactions between the block and the regolith as shock propagation proceeds during the impact.

image

Figure 3.  Largest remnant mass ratio as a function of burial depth, Dbur, for the six block impacts.

Download figure to PowerPoint

Figure 4 shows the mass–frequency distributions of the smaller debris resulting from each shot, with the line thickness proportional to block burial depth (thicker line corresponding to deeper burial). As the block buried beneath the surface (shot 090232) was not fragmented or cratered at all, there is no line plotted for that shot. As in the case for Fig. 3 (plot of largest remnant mass ratio), other than statistical “noise” in the shapes of the mass–frequency distributions associated with the blocks buried nearest the regolith surface, the trend from more disruptive events (i.e., second- and third-largest fragments are a larger fraction of total block mass) for shallow burial depth toward more cratering-like distributions (i.e., more smaller fragments) for deeper burial depth appears as expected.

image

Figure 4.  Mass–frequency distributions of the fragments resulting from the six block impacts. Fragment masses are normalized to the mass of the original target block for each shot. Line thickness is proportional to burial depth, with thicker lines representing deeper burial. Shot 090232 resulted in no fragmentation, so there is no mass–frequency distribution plotted for that shot.

Download figure to PowerPoint

Figure 5 shows the sand surface within the AVGR impact chamber after each impact, with the same viewing geometry as in Fig. 2 and lit to accentuate block fragments and the topography of disturbances induced in the regolith. The image subpanels in Fig. 5 are arranged by block burial depth, with burial depth increasing toward the bottom of the image. The basic trend in morphological features as a function of block burial depth appears to be as follows. Impact onto a surface block results in a substantially fractured largest remnant block, a number of surrounding “secondary craters” caused when ejected fragments on low elevation trajectories skip across the surface near the target block, and a crater-like disturbance around the largest remnant block, which itself may be displaced by the recoil of ejected fragments toward the up-range direction of the projectile trajectory. These features become more subdued as burial depth increases, transitioning to a more pronounced crater-like disturbance around the cratered block for more deeply buried blocks. In this latter case of impact into substantially buried blocks, the end results can closely resemble a simple percussion crater formed from low-speed reaccretion impacts of blocks into regolith. The greatest disturbance to the regolith and the largest surrounding crater appears to occur for blocks buried flush with the surface. The most deeply buried (Dbur = 1.25) block was not cratered or marred in any way; the overburden of sand took the full brunt of the impact, leaving a crater with a diameter of about 16 cm surrounding the block. As blocks are buried just below the regolith, it appears that the presence of the block interferes with the normal cratering flow that would be observed in pure sand. We suspect that for even deeper burial depths the presence of the block would have less interference with cratering flow and crater diameter would increase to that which would be observed in pure regolith. In June 2010 during a subsequent experiment run at the AVGR, we ran a control experiment of an impact with the same conditions (i.e., one-eighth inch aluminum projectile, impact speed 4.8192 km s−1, impact angle 45°) into sand only, producing a crater ∼17.8 cm in diameter and ∼3.7 cm deep.

image

Figure 5.  Outcomes of the six block impact experiments, showing the range of block fragmentation and displacement, and the range of morphologies of regolith disturbance. Images are arranged from left to right, top to bottom, by increasing block burial depth, Dbur. The up-range direction is at the top of each image.

Download figure to PowerPoint

Generally, interpretations of planetary surfaces invoke a series of processes and events, but the chief evidence is the appearance of the surface at an instant in time, the moment of observation. While Fig. 5 shows such still, after-the-fact views of the state of the surface after each of the experiments we did, we also have a rich collection of high-speed videos for each experiment, taken at a variety of frame rates and from five different observing vantage points (see, for example, supporting information [Videos S1–S3]). We have studied these videos and now describe several interesting insights.

The high-speed video of the impacts reveals a transition in the behavior of the largest remnant and the regolith at the edges of the blocks during the impact process. For surface and near surface blocks, the impact shock waves reflect off the free surfaces of the blocks and spall off the outer layers of the blocks in flat, chip-shaped fragments, contributing to the substantial fragmentation displayed in the mass–frequency distributions shown in Fig. 4. Individual sand grains at the block-regolith interfaces are observed to be displaced, but not substantially so, and not in any organized fashion. The largest remnant recoils up-range, as most of the momentum transfer to the target block comes from the spray of ejecta and not from the projectile directly. As block burial depth increases, the displacement of the regolith at the block base becomes more substantial, a sign of an increasing fraction of the interior shock being transmitted through the block to the regolith contacting the block’s exterior, rather than being internally reflected within the block. The deeper the block is buried in the regolith, the more impact energy is transmitted, through the block, to the surrounding regolith, causing the formation of a crater surrounding the remnant block. This corresponds with the observed decrease in the amount of fragmentation induced in the more deeply buried blocks. At the same time, the pattern of displacement of the regolith particles becomes more organized and begins to take on the appearance of ejecta curtains observed in conventional cratering experiments conducted into sand alone. In the videos, we see cases where gouge-type craters are formed quite close to the block by rapidly departing fragments, and then the much more slowly growing crater in the regolith actually partially or totally covers up this transient, early crater. So, at least in this case, secondary craters can be formed, then covered up or destroyed an instant later. Oftentimes we see in planetary images what appear to be pre-existing craters that have been partially buried by the continuous ejecta blanket from a crater. But our experiments demonstrate that at least in this special circumstance, the crater need not be pre-existing—it just formed earlier in the same cratering event.

Several attributes of the final surfaces formed in ways quite different from what seems to be true from the final states. For instance, in shot 090227, there is a chain of four or five craters stretching radially away from the impacted block, with a small fragment sitting on the surface just beyond the last crater (visible at about the 8 o’clock position in Fig. 5a). It would appear that a slow-moving fragment bounced along the surface, creating the several craters, and then came to rest. Viewing the videos, we can see that something totally different happened. The crater chain actually formed very shortly after the hypervelocity impact, when a fragment from the block flew away at a low angle, just above the simulated regolith surface; as it rotated, corners dipped into the regolith, forming the crater chain. The high-velocity fragment itself departed far away from the block. The rock that we now see lying on the surface had nothing to do with formation of the crater chain. Probably, it was a very slow-moving, late fragment, which was launched at a higher angle and came to rest where we now see it, only coincidentally aligned with the crater chain. (We cannot be sure of the behavior of the late-stage fragment as the high frame-rate video ends before it is emplaced. It is possible, despite design of the experimental set-up to minimize fragment rebound, that this fragment was actually a higher-velocity fragment that rebounded from the curtains enclosing the experiment.)

A different example is apparent in shot 090230. The final surface shows several bowl-shaped craters in the regolith with small fragments centered in their bottoms. A natural supposition is that these fragments struck at subhypervelocity speeds, forming the craters but remaining visible. Instead, the videos show that the craters were formed shortly after the impact by high-velocity, rotating fragments that gouged the surface as they departed the scene, leaving nothing within the craters. The rocks seen in the crater bottoms must have been late, slow-moving fragments that were preferentially caught inside the craters. As described above, it is possible that these fragments also rebounded from the curtains surrounding the experiment. As any golfer knows, objects can fall into depressions; it does not mean that the objects made the depressions.

We should emphasize that despite the similarity of the surrounding craters to secondary craters commonly seen (at larger spatial scales) on planetary surfaces, these craters in our experiments are of a very different nature. They originate from fragments ejected during impacts onto target blocks resting on or only slightly submerged into the surface. Within the confines of the AVGR impact chamber the observed secondary craters are formed from ejecta launched downward toward the regolith surface or from elongated fragments ejected horizontally where the rotation of the fragments causes a portion of the fragment to excavate the regolith. As block burial depth increases, the amount of block fragmentation is reduced, “unpeeling” of block fragments away from the point of impact is subdued, and more ejecta is launched upward and outward away from the block, so that few ejecta fragments excavate secondary craters near the block.

On actual planetary surfaces where the ballistic trajectories of ejected fragments are unconstrained by impact chamber walls, some fraction of these fragments will reaccrete to the object’s surface some distance away and will produce secondary craters as well, although they may not be able to be readily linked back to a source block. Figure 6 shows an elongated secondary crater and “skid marks” marking the terminal trajectory of the associated small ejecta block within one of the “ponds” of fine regolith material within a small crater on Eros. The vagaries of ejecta dynamics on elongated and irregularly shaped rapidly rotating bodies like Eros can sometimes be rather counterintuitive, making detailed numerical models of the dynamical evolution of ejecta particles particularly useful tools for interpreting surface morphological expressions of emplaced ejecta units (Geissler et al. 1996; Durda 2004). Given this block’s present location on the narrow elongated end of Eros and the azimuth of its landing trajectory, Durda (2009) has “back tracked” its landing trajectory in an effort to constrain the fragment’s source location on Eros’ surface. The model results tightly limit the potential source region for this ejecta fragment to be very near its present resting location. Given the results of our block impact experiments, we suggest that this fragment and its associated landing scars is a good candidate for being a “chip” off a nearby impacted block, out of frame to the bottom of Fig. 6.

image

Figure 6.  An oblong secondary crater (lower arrow), with an associated track and terminal boulder (upper arrow), within a ponded deposit on Eros. The pond is ∼90 m in diameter. NEAR image number 156087851.

Download figure to PowerPoint

In light of the observed outcomes of these impact experiments, we revisit Fig. 1 and reiterate the alternative interpretation given in the Introduction—i.e., it may be a block broken in place by impact, with the surrounding crater representing finer-scale regolith displaced by the impact and fragmentation of the block.

Future impact experiments will be needed to examine the full range of morphologies of block fragmentation and regolith displacement as a function of block size and shape, block burial depth, impact angle, impact specific energy, and regolith size fraction and cohesiveness. High-speed impact experiments in a low-gravity environment are needed in particular to better understand the behavior of such impacts on the surfaces of small solar system bodies.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Experimental Procedure
  5. Results and Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Acknowledgments— This work was supported by NASA Planetary Geology & Geophysics program grant NNX07AR29G, and NASA Discovery Data Analysis program grants NNG06GF92G, NNG05GB96G, and NNJ06HD82A. We thank Chuck Cornelison, Donald M. Holt, Donald B. Bowling, and Richard E. Smythe of the Ames Vertical Gun Range for their assistance in conducting the impact experiments, and J.-P. Wiens for photographic and high-speed video support.

Editorial Handling— Dr. Michael Gaffey

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Experimental Procedure
  5. Results and Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information
  • Asphaug E. 2003. Formation of impact craters on comets and asteroids: How little is known. In Impact cratering: Bridging the gap between modeling and observations, February 7–9, 2003, edited by HerrickR. and PierazzoE. LPI Contribution 1155. Houston, Texas: Lunar and Planetary Institute. pp. 1213.
  • Bottke W. F. Jr., Nolan M. C., Greenberg R., and Kolvoord R. A. 1994. Velocity distributions among colliding asteroids. Icarus 107:255268.
  • Britt D. T. and Consolmagno G. J. 2003. Stony meteorite porosities and densities: A review of the data through 2001. Meteoritics & Planetary Science 38:11611180.
  • Chapman C. R., Merline W. J., Thomas P. C., Joseph J., Cheng A. F., and Izenberg N. 2002. Impact history of Eros: Craters and boulders. Icarus 155:104118.
  • Durda D. D. 2004. Ejecta generation and redistribution on 433 Eros: Modeling ejecta launch conditions (abstract #1096). 35th Lunar and Planetary Science Conference. CD-ROM.
  • Durda D. D. 2009. Constraining source crater regions for boulder tracks and elongated secondary craters on Eros (abstract #2173). 40th Lunar and Planetary Science Conference. CD-ROM.
  • Flynn G. J. and Durda D. D. 2004. Chemical and mineralogical size segregation in the impact disruption of inhomogeneous, anhydrous meteorites. Planetary and Space Science 52:11291140.
  • Gault D. E., Hörz F., and Hartung J. B. 1972. Effects of microcratering on the lunar surface. Proceedings, 3rd Lunar Science Conference. Geochimica et Cosmochimica Acta 2(Suppl. 2):27132734.
  • Geissler P., Petit J.-M., Durda D. D., Greenberg R., Bottke W., Nolan M., and Moore J. 1996. Erosion and ejecta reaccretion on 243 Ida and its moon. Icarus 120:140157.
  • Hörz F., Schneider E., Gault D. E., Hartung J. B., and Brownlee D. 1975. Catastrophic rupture of lunar rocks: A Monte Carlo simulation. Moon 13:235258.
  • Housen K. R. and Holsapple K. A. 1999. Scale effects in strength-dominated collisions of rocky asteroids. Icarus 142:2133.
  • Richardson J. E. Jr., Melosh H. J., Greenberg R. J., and O’Brien D. P. 2005. The global effects of impact-induced seismic activity on fractured asteroid surface morphology. Icarus 179:325349.
  • Saito J., Miyamoto H., Nakamura R., Ishiguro M., Michikami T., Nakamura A. M., Demura H., Sasaki S., Hirata N., Honda C., Yamamoto A., Yokota Y., Fuse T., Yoshida F., Tholen D. J., Gaskell R. W., Hashimoto T., Kubota T., Higuchi Y., Nakamura T., Smith P., Hiraoka K., Honda T., Kobayashi S., Furuya M., Matsumoto N., Nemoto E., Yukishita A., Kitazato K., Dermawan B., Sogame A., Terazono J., Shinohara C., and Akiyama H. 2006. Detailed images of asteroid 25143 Itokawa from Hayabusa. Science 312:13411344.
  • Shoemaker E. M. 1962. Interpretation of lunar craters. In Physics and astronomy of the Moon, edited by KopalZ. New York: Academic Press. pp. 283359.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Experimental Procedure
  5. Results and Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Video S1. Run 090227 7237.avi. 1900 frames per second video of Shot 090227. View looking down on the target block through a viewport near the top of the AVGR impact chamber. Projectile impact from the bottom of the frame.

Video S2. Run 090230 7237.avi. 1900 frames per second video of Shot 090230. View looking down on the target block through a viewport near the top of the AVGR impact chamber. Projectile impact from the bottom of the frame.

Video S3. Run 090231 7238.avi. 1900 frames per second video of Shot 090231. View looking at the target block from the side through the large side window of the AVGR impact chamber. Projectile impact from the right side of the frame.  Note the up-range recoil of the target block.

FilenameFormatSizeDescription
MAPS_1163_sm_Run090227-7237.avi17093KSupporting info item
MAPS_1163_sm_Run090230-7237.avi14674KSupporting info item
MAPS_1163_sm_Run090231-7238.avi14677KSupporting info item
MAPS_1163_sm_vid-captions.doc24KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.