The following is a copy of the original proposal to the NASA Innovative Research Program. It gives a detailed explanation of the experimental objectives, goals, and initial ideas for hardware design.

Investigation Title: Microgravity Experiment on Low Velocity Impacts Principal Investigator: Joshua E. Colwell Institution: Laboratory for Atmospheric and Space Physics (LASP), University of Colorado


Low velocity collisions between planetary ring particles are crucial in determining the dynamics of planetary rings and also determine the characteristics of dust rings which are often the only observable component of a planetary ring. Low velocity collisions were also important in the early stages of planetesimal accretion. We propose to conduct the first experiments on low velocity collisions into regolith in a microgravity environment. Microgravity is necessary to study impacts into a dusty regolith at velocities less than meters per second. Ground-based experiments have examined the velocity range down to about 5 meters per second. We propose to extend the data set to velocities of 1 centimeter per second with a reusable experiment for the Space Shuttle Get Away Special Canisters (GASCANs). The experiment will conduct six impacts into lunar soil simulant at various impact velocities and regolith depths. We will measure the velocity of the ejecta from the impact. This will fill a void in our understanding of the collisional balance of dust production in planetary rings. This data will also provide new information that will help determine the efficiency of accretion of planetesimals.


The solar system is a collisionally evolved system. Recently, our understanding of the important role that hypervelocity collisions play in the evolution of the planets has been broadened. Galileo has returned images of main belt asteroids Ida and Gaspra that graphically demonstrate the role of catastrophic collisions in their origin and evolution. The impact of comet Shoemaker-Levy 9 into Jupiter has provided a wealth of data on the effects of comet impacts on planetary atmospheres. Hypervelocity impacts may also play a crucial role in the origin and evolution of small planetary satellites and planetary rings (Harris 1984, Esposito and Colwell 1989, Colwell and Esposito 1992, 1993). Laboratory experiments have been able to address some of the parameter space appropriate for these types of collisions. Velocities of several km/sec have been achieved in the laboratory (e.g. Davis and Ryan 1990, Hartmann 1985, Kawakami et al. 1983, Stoffler et al. 1975, Gault et al. 1963). The scale of the impacts are necessarily small (/sim10 cm). This and the high velocities mean that material strength rather than gravity control the outcome of the impact. So-called ``gravity regime'' impacts have been explored using an overburden of gas and explosives to simulate the impact (Housen et al. 1991). These have extended the parameter space covered to include catastrophic disruption of objects on the order of 100 km in size. However, at the current epoch and throughout its history, low velocity collisions have played an important role in sculpting planetary systems. Impacts typically occur at speeds exceeding the mutual escape velocity of the two bodies. Thus, impacts at speeds on the order of 1 m/sec or less involve objects meter-sized and smaller. There are two environments in solar system history that involve collisional systems with objects in this size and velocity range: planetary rings and accreting proto-planetesimals. Planetary ring particles range in size from sub-micron to several km. The bulk of the mass and surface area of Saturn's optically thick rings is made up of particles smaller than a few meters in radius (e.g. Esposito et al. 1984). The particles in these rings suffer collisions on the short orbital time scale of several hours. Dust is produced in the rings by micrometeoroid bombardment of the ring particles. This dust coats the surfaces of the ring particles and is released when two ring particles collide. The amount of dust released and the velocity distribution of the dust is unknown, however. These parameters are crucial in our understanding of planetary rings (e.g. Weidenschilling 1984). The dust component of rings is visible in high phase angle viewing geometries achieved by spacecraft, and is often the only component of a ring that is visible. This is the case with parts of the Uranian and Neptunian ring systems (Smith et al. 1986, Smith et al. 1989). The size and velocity distributions of the macroscopic ring particles have been inferred from the dust optical depth (Colwell and Esposito 1990a, 1990b). However, these inferences were based on extrapolations of impact experiments from the km/sec velocity range to the cm/sec velocity range, and further data is needed to determine the validity of these extrapolations.

The accretion of the planets proceeded from coagulation of dust grains in the solar nebula to the binary collisional accretion of km-sized planetesimals. Van der Waals forces between dust grains are sufficient to account for growth of cm-sized aggregates in a nebula with up to a moderate level of turbulence (Weidenschilling 1980). Bodies meter-sized and larger decouple from the nebular gas, and consequently have relative velocities low enough to allow for further collisional growth. The details of the growth process in the critical centimeter-to-meter range are not known, and depend greatly on the efficiency of mass loss during collisions. These collisions likely occurred at relative velocities of 1-100 cm/sec for either a laminar or turbulent nebula (Weidenschilling 1993). Velocities in this range are significantly higher than escape velocities for centimeter-to-meter size particles, and so collisions must have been highly inelastic in order for growth to proceed. Dusty regoliths covering particles may have helped to dissipate collisional energy, reducing the rate of mass loss during collisions and promoting accretional growth of larger bodies. If not, collisional growth alone may be insufficient to account for the formation of km-sized planetresimals, and a disk layer dense and quiescent enough to allow for gravitational instability might have been required.

Our proposed research will provide the first data in an unexplored regime of collisional parameter space. This data will advance our understanding of planetary accretion and planetary ring science.


Crater formation and ejecta production from hypervelocity impacts into regolith covered surfaces have been studied in the laboratory using ground-based experiments (e.g. St\"offler et al. 1975, Hartmann 1985). One set of impact experiments has been performed down to several meters per second impact velocity (Hartmann 1985). Other low velocity experiments have examined the coefficient of restitution at low velocities (Hatzes et al. 1988, Bridges et al. 1984), while Hartmann's experiments looked at impacts into a simulated regolith at speeds as low as 5.3 m/sec. Hartmann's experiments were performed in a normal Earth gravity environment. This limits not only the velocity regime attainable, but also the utility of the available low velocity data in applications to planetary science problems. We propose to extend the existing data set to 1 cm/sec impact speed.

The parameters that need to be measured include the total amount of material ejected in the impact, and the velocity distribution of the ejecta. These parameters depend in general upon the shape, mass, density, material properties, and velocity of the impactor, and upon the material properties and gravity environment of the target material (cf. Holsapple 1993). For most hypervelocity experiments the impactor can be characterized by only its mass and velocity,


where K, alpha, and beta are constants determined by material properties of the target. The strength and shape of the impactor do not significantly affect the far field consequences of a hypervelocity impact which can be approximated by a point source delivery of energy at some distance beneath the surface of the target. Depending on the size of the target, either the target strength or its gravitational field determines how the target responds to the shocks generated by the impact. For low velocity impacts, however, the shape and density of the impactor may play important roles in cratering a regolith surface. This is because the ejection of dust happens in the near field of the impact site and the total specific energy of the event is much less than in hypervelocity impacts. The ejecta velocity distribution from hypervelocity impacts is usually characterized by a power-law of the form

f(>v) = Cv^(-gamma)

where C and gamma are constants determined by the target properties, and f(>v) is the mass fraction of ejecta travelling faster than v. The value of C is related to the minimum ejecta velocity by C=v(min)^{gamma}. The value of gamma is determined by the surface properties. Housen et al. (1983) have shown based on theoretical scaling arguments that gamma lies in the range 1.2 to 2.0, where the lower values are appropriate for ``soft'' or fluid target materials, such as sand or water, and the upper limit is approached for hart targets such as basalt. These values are consistent with experimental results in the hypervelocity regime reported by Stoffler et al. (1975) and Gault et al. (1963). At the very low velocities we consider, only impacts into unconsolidated regolith are important. The coefficient of restitution at low velocities between simulated ring particles can be and has been examined with the use of pendulums in ground-based laboratories (Hatzes et al. 1988, Bridges et al. 1984, Dilley and Crawford 1994), and has been successfully modeled using a modified Hertz theory (Hertz 1886, Dilley 1993, Spahn et al. 1994).

Scaling arguments for the outcomes of cratering events are not only almost exclusively applied to hypervelocity impacts, they treat two regimes which do not inlcude the type of collision that occurs between planetary ring particles and between accreting proto-planetesimals. These regimes are separated by the ratio Y/((rho)*gR), where Y is a measure of the target strength, R is its size, rho the density, and g the gravitational acceleration. When Y/((rho)*gR)>>1 the cratering event is in the strength regime. For high velocity impacts, this has meant impacts into a target of competent rock. Any regolith on a body with negligible gravity would be easily penetrated and hence transparent to a hypervelocity impactor. However, the situation for low velocity impacts onto small, low-gravity targets, may be entirely different. Experimental data is needed to characterize these impacts in order to have any confidence in the extrapolations to this regime from the hypervelocity impact experiments and scaling relations.

Graduate students at the University of Colorado have performed drop experiments at velocities of 3 to 10 m/sec into fine powder under a vacuum. Results on crater diameter were obtained and found to be consistent with the low velocity end of Hartmann's (1985) impact experiments. However, the ejecta velocity measurements are not practical in a one gravity environment because the time scale of the ejecta trajectories is too short. These experiments used the same basic approach we propose for the Microgravity Experiment Low Velocity Impacts (MELVI). Impactors are spring launched, and the data is recorded on video. Frame by frame analysis of the video allows accurate determination of the crater diameter, impactor velocity, and estimates of the ejecta mass.


In order to obtain data on impacts in the low velocity regime, a micro-gravity environment is necessary. Parabolic airplane flights provide micro-gravity for a few tens of seconds. The cratering event at impact velocities measured in cm/sec may last tens of seconds, and a long time base is needed to accurately measure the velocities of not only the impactor but also the ejecta. In addition, the experiment must be performed in a vacuum to avoid interference of the slow-moving dusty ejecta by fluid motions of the atmosphere. The cost, short time in microgravity, and additional complexity of a vacuum system rule out parabolic airplane flights. As discussed above, the scale of low-velocity impacts is small. Targets and impactors in planetary rings or the early proto-planetary nebula that impact at speeds less than meters per seconds are meter-scale or smaller. Furthermore, the low velocity of the impactor and the ejecta allow accurate determinations of the velocities and crater dimensions possible with a small scale experiment. The Get Away Special cannisters (GASCANs) of the Shuttle Small Payload Program offer an ideal platform for performing the low velocity impact experiments we propose.

The experiment will be reusable, but to maximize the scientific return of each flight, the experiment will perform six impacts with different impact parameters. The parameters that can be varied are impact velocity, regolith depth, impactor shape, impactor mass, and impactor density. We will use cylindrical and spherical impactors and assume azimuthal symmetry of the crater and ejecta distribution. Data will be recorded by a video camera viewing the impact plane edge-on. This will allow determination of all the relevant parameters: impactor velocity, crater diameter, and ejecta velocity distribution. Impacts will be perpendicular to the plane of the target, which will consist of a tray of fine powder, simulating regolith on a small solar system body.

The GAS cannister in which the experiment is housed has an inside length of 71.755 cm (28.25 inches) and an inside diameter of 50.165 cm (19.75 inches). An exploded view of the experimental setup is shown in figure 1:

Each tray of simulated regolith is isolated in a box with plexiglass sides allowing video recording of the impact event. The impactors will be fired at speeds between 1 and 200 cm/sec onto the regolith surfaces using six spring-loaded guns. The impactors will be spheres or cylinders, depending on the experimental run, approximately 1 cm in diameter. The regolith material is lunar simulant MLS-1 mined in Minnesota and processed to match the properties of lunar soil.

Because of the low velocities, high speed film typically used to record impact experiments is not necessary. Sufficient resolution for the scale and speed of our impact experiments is obtained with the standard video framing rate of 30 sec$^{-1}$ with a shutter speed of 1/400 sec. For high spatial resolution and compact size to accomodate the GASCAN dimensions we will use a Hi-8 video system. The record time on a standard tape is two hours, and the experiment will last approximately one hour - ten minutes for each impact.

A simple timing circuit activated by a baroswitch in the GASCAN operates the experiment. The switch is activated when the GASCAN vents, a light is turned on and the video camera begins recording. (Because we require that the experiment be performed when no maneuvers of the Space Shuttle are taking place, it may be necessary for an astronaut to switch on the experiment rather than using the baroswitch.) Muscle wire (which contracts when a current is passed through it) is used to trigger the release of a spring loaded door that contains the regolith during launch and prior to the experiment. The door is not airtight, so that pore spaces in the regolith are fully evacuated before the door is released. This avoids blow out of the regolith when the target is exposed. Following the opening of the door, another muscle wire circuit releases the spring loaded impactor. After a ten minute interval, the process is repeated with the next impactor-target pair. Mirrors may be used for imaging the different impacts at the highest possible spatial resolution and optimum orientation. Also housed in the GAS cannister are electronic boxes and the battery box, which is vented to the outside through the GAS interface plate.

Upon successful development and completion of the initial flight experiment, the experiment will be flown again with different impact parameters, with funding from other programs. This experiment has the potential to provide a large and unique data set on an important regime of collisions in the solar system that cannot be studied in conventional ground-based laboratory environments.


One of the goals of the GASCAN program is access to space experimentation by students. This experiment was originally conceived by graduate students at the University of Colorado, and they have had a major role in developing the current experiment. The scientific and overall program management will be the responsibility of the Principal Investigator. Undergraduate aerospace and electronics engineering students and planetary science graduate students will have active participation in the detailed design and construction of the experimental apparatus. Because the experiment has the potential for several flights, the program offers the opportunity for active involvement in space experimentation to several classes of students once it is established.


The proposed experiment will be the first planetary science experiment to be performed in orbit. It will produce data in an area that cannot be obtained through traditional experimental techniques on the ground. This data will further our understanding of the dynamics of planetary rings, the inventory of planetary rings, and will facilitate the analysis of planetary ring data returned by the Voyager spacecraft, Galileo at Jupiter, and Cassini at Saturn in the next decade.

This data will also help fill a gap in understanding conditions in the preplanetary nebula, after sticking forces resulted in cm-sized objects and prior to the accretion of km-sized planetesimals. The experiment represents a new scientific use of the microgravity environment offered by the space shuttle and eventually the international space station. The experiment also has educational value through the active involvement of students - graduate and undergraduate - in a space experiment with rapid turnaround. Students involved in the project will be able to participate through the construction and completion of the experiment within the course of their studies.


The experimental apparatus will be constructed at the Laboratory for Atmospheric and Space Physics at the University of Colorado. Researchers at LASP have designed and built rocket experiments and cooperated with the University's Space Grant College on GASCAN experiments. The Lab has full facilities for the construction of the experiment, and thermal testing of the various components.


Bridges, F. G., A. Hatzes, and D. N. C. Lin 1984. Structure, stability and evolution of Saturn's rings. Nature 309, 333-335.

Colwell, J. E., and L. W. Esposito 1990a. A numerical model of the Uranian dust rings. Icarus 86, 530-560.

Colwell, J. E., and L. W. Esposito 1990b. A model of dust production in the Neptune ring system. Geophys. Res. Lett. 17, 1741-1744.

Colwell, J. E., and L. W. Esposito 1992. Origins of the rings of Uranus and Neptune, 1, statistics of satellite disruptions. J. Geophys. Res., , 10,227-10,241.

Colwell, J. E., and L. W. Esposito 1993. Origins of the rings of Uranus and Neptune, 2, initial conditions and ring moon populations. J. Geophys. Res. 98, 7387-7401.

Davis, D. R., and E. V. Ryan 1990. On collisional disruption: experimental results and scaling laws. Icarus 83, 156-182.

Dilley, J. 1993. Energy loss in collisions of icy spheres: loss mechanism and size-mass dependence. Icarus 105, 225-234.

Dilley, J., and D. Crawford 1994. Size/mass dependence of elasticity in collisions of icy spheres. Bull. Am. Astron. Soc. 26, 1141.

Esposito, L. W., J. N. Cuzzi, J. B. Holberg, E. A. Marouf, G. L. Tyler, and C. C. Porco 1984. Saturn's rings: structure, dynamics, and particle properties, in Saturn, (T. Gehrels and M. S. Matthews, eds.), pp. 463-545, Univ. of Arizona Press, Tucson.

Esposito, L. W., and J. E. Colwell 1989. Creation of the Uranus rings and dust bands. Nature 339, 605-607.

Gault, D. E., E. M. Shoemaker, and H. J. Moore 1963. Spray ejected from the lunar surface by meteoroid impact. NASA Tech. Note, D-1767.

Harris, A. W. 1984. The origin and evolution of planetary rings. In Planetary Rings, (R. Greenberg and A. Brahic, Eds.), pp. 641-659. Univ. of Arizona Press, Tucson.

Hartmann, W. K. 1985. Impact experiments. Icarus 63, 69-98.

Hatzes, A. P., F. G. Bridges, and D. N. C. Lin 1988. Collisional properties of ice spheres at low impact velocities. Mon. Not. R. Astron. Soc., 1091-1115.

Hertz, H. J. 1886. Reine Angew. Math. 92, 156.

Holsapple, K. A. 1993. The scaling of impact processes in planetary sciences. Annu. Rev. Earth Planet. Sci., 333-373.

Housen, K. R., R. M. Schmidt, and K. A. Holsapple 1983. Crater ejecta scaling laws: fundamental forms based on dimensional analysis. J. Geophys. Res., 2485-2499.

Housen, K. R., R. M., Schmidt, and K. A. Holsapple 1991. Laboratory simulations of large scale fragmentation events. 94, 180-190.

Kawakami, S., H. Mizutani, Y. Takagi, H. Kato, and M. Kumazawa 1983. Impact experiments on ice, J. Geophys. Res. 88, 5806-5814.

Smith, B. A., et al. 1986. Voyager 2 in the Uranian system: Imaging science results, Science 233, 43-64.

Smith, B. A., et al. 1989. Voyager 2 at Neptune: Imaging science results, Science 246, 1422-1449.

Spahn, F., N. K. Brilliantov, J.-M. Hertzch, and T. P\"oschel 1994. About collisions between granular particles: application to planetary rings, Bull. Am. Astron. Soc. 26, 1143-1142.

St\"offler, D., D. E. Gault, J. Wedekind, and G. Polkowski 1975. Experimental hypervelocity impact into quartz sand: distribution and shock metamorphism of ejecta. J. Geophys. Res. 80, 4062-4077.

Weidenschilling, S. J. 1980. Dust to planetesimals: Settling and coagulation in the solar nebula. Icarus 44, 172-189.

Weidenschilling, S. J., C. R. Chapman, D. R. Davis, and R. Greenberg 1984. Ring particles: collisional interactions and physical nature, in Planetary Rings (R. Greenberg and A. Brahic, eds.), pp. 367-416, University of Arizona Press, Tucson.

Weidenschilling, S. J., and J.N. Cuzzi 1993. Formation of Planetesimals in the Solar Nebula, in Protostars and Planets III, (E. H. Levy and J. I. Lunine, eds.), pp. 1031-1060, University of Arizona Press, Tucson.

back to COLLIDE Home Page