December 5, 2007 | David F. Coppedge

Is Making Planets Child’s Play?

Are star children good at child’s play? Like making mudballs, it should be easy to roll up dust into planets.

Dave Mosher says “Making planets is child’s play” (see Space.com). He refers not to a human child, but to a star child – that is, a young star, like UX Tau A or LkCa 15. The Spitzer Space Telescope detected dust disks around these two stars that are “1 million years old, which is 10 times younger than other known planet-forming systems.”

The Spitzer-JPL press release explains that these “young” disks appear to have gaps where planets are forming. Here is an overview of what theoretically goes on in the child’s game of making planets:

Such dusty disks are where planets are thought to be born. Dust grains clump together like snowballs to form larger rocks, and then the bigger rocks collide to form the cores of planets. When rocks revolve around their central star, they act like cosmic vacuum cleaners, picking up all the gas and dust in their path and creating gaps.

(How snowballs form rocks is left as an exercise.) The same people who write press releases must not read the scientific papers. There are major problems with this scenario. Among the most difficult is getting the small dust grains to clump together, a process called accretion.  Planetary scientist Jeff Cuzzi wrote about this problem in Nature August 30:1

Making planets is tricky, and probably takes several stages. First, tiny interstellar grains must accrete into mountain-sized objects massive enough to decouple from their cocoon of nebula gas. These objects probably then combine in collisions, growing ever larger, past asteroid-sized planetesimals and lunar-sized embryos, to full-blown planets. How the first stage of this process, primary accretion, works is a fundamental unsolved problem of planetary science.

Not exactly child’s play. The ingredients for planets are speeding around the star at tens of meters per second, or even several times the speed of sound. They are more likely to collide and disrupt than accrete. Furthermore, we can’t see inside a dust disk to look at what happens, so explanations have to be inferred from models. After looking at possible contributions from turbulence or magnetic fields, Cuzzi ended,

The answer could be that some combination of processes, each selecting a different particle size, acts simultaneously or sequentially, possibly in turbulent conditions. (Of course, the mechanism by which turbulence is maintained remains uncertain.) Whatever the final answer turns out to be, the results of Johansen and colleagues indicate that future efforts devoted to developing more complex models of the interactions between particles and gas in the protoplanetary nebula will be a good investment.

The child’s play is stumping the PhDs. A recent paper in Icarus addresses this question specifically.2 Paraskov, Wurm and Krauss considered all the variables: particle size and composition, gas drag, collision velocities and more. When dust particles are small and moving at low velocities relative to one another, they can stick up to a point, but “typical collision velocities go beyond all these threshold velocities for sticking,” they noted. “Therefore, for larger particles (>10 cm) it is not possible that they continue to grow by simple hit-and-stick mechanisms.” Yet particles need to accrete up to mountain size (a kilometer or more in diameter) before gravitation can take over.

Because of problems with the simple theoretical models, “Different mechanisms to further aid growth of planetesimals have been suggested, especially to overcome unfavorable conditions where the primary collisions lead to erosion rather than growth.” Obviously, a hopeful pre-planetesimal needs to grow faster than it erodes. Some ideas they evaluated included: (1) secondary accretion of dust from collisions, (2) gas drag accretion as particles slow down in the gas of a spinning disk, (3) electrostatic attraction, (4) formation of porous bodies that can absorb energy of collisions without disrupting (inelastic collision), and (5) gravitational instabilities that form pockets of higher density. Each model has its shortcomings, they said; the bottom line is that collisional dynamics must be factored into any scenario.

The team performed drop tower experiments to see what happens to collisional fragments. “The experiments reported here are intended to give realistic upper limits for fragment velocities,” one of the key parameters for any accretion model. While accretion seems reasonable up to small particle sizes, maybe up to a centimeter (given low relative velocities), disruption and fragmentation becomes a greater concern for larger particles. “To decide if planetesimals can grow in collisions or if fragmentation dominates, it is important to know what the typical collisions are,” they said; that is why there is no substitute for actual experimentation to provide an “experimental database,” instead of “ad hoc assumptions” that plague models.

The team ran experiments in a vacuum, using a Bremen drop tower that provides microgravity for about 5 seconds. They videotaped the collisions, and measured the mass gained and lost by the target particles. They tried porous and spherical targets. They varied the impact velocities from 3.5 to 21.5 m/s.

So what happened? “The impacts into highly porous targets generally show a very destructive behavior,” they reported. “They result in crater formation—an imprint of the projectile—and an extensive erosion of the whole target surface and deeper target layers.” At 19.5 m/s, the projectile barreled completely through the target. Dust projectiles fired at compact surfaces were different; they formed pyramid-like structures on the surface.

In all the experiments, the ejecta that was blasted off was measured. In most cases, the “accretion efficiency” was negative: more mass was lost than accreted. Only in the case of a dust projectile hitting a slightly compacted target was a net gain measured; otherwise, “All impacts into highly porous targets resulted in a mass loss of the target,” with more loss at higher velocities.  The single case of mass gain was a contrived and unrealistic scenario.3

What happens to the ejecta? It moves slower than the projectile, from 3 to 120 cm/s, in directions depending on the projectile mass and velocity. They believe it is possible a target can become more porous if hit enough times. “A growing body might then consist of compact parts but also large pores,” they said. “This is important as large pores enable a larger gas flow through the body which is beneficial for reaccretion of ejecta by gas flow.” Experiments with more credible dust projectiles, however, were more difficult to characterize; the impact is spread over a larger area. The highest accretion efficiency (42%) was found with high-impact projectiles onto compact targets, but the results were difficult to quantify. They could not comment much about the role of electrostatic charges on outcomes, “an important problem in itself.”

At the end of the paper, it was time to put the pieces together and see what this means for the growth of planets. “For most experiments we simulated an extreme case of collisions in the sense that we built the targets as weak as possible by using large granules but retaining dust features, i.e., a small degree of cohesion, by using dust,” they said. Result? “We find that these collisions are strongly erosive.” No matter the speed or mass, “they all show a mass loss of the larger body upon impact if the target is only weakly bound.” Does that mean the end of the planetesimal hypothesis? Not so fast:

However, care should be taken in applying these results directly to collisions in protoplanetary disks. Only a single collision of a weak body with a particle several mm in size at about 10 m/s would be enough to change this body substantially. The given collisions will therefore not be typical ones. The consequence of the erosion found in these experiments is not that growth of larger bodies is impossible but that objects of the kind studied cannot grow larger than a few cm without being at least partially compacted.

Trouble is, it is worse for chondrules – highly compact grains characteristic in meteorites called chondrites. They are already compacted and would be more subject to erosion.

Further experiments on this are needed but if we speculate that morphologies comparable to our weak dust granule targets would need to be present for aggregates built from indestructible chondrules then growth to larger bodies in a reservoir dominated by sub-mm chondrules is not possible at velocities of a few m/s. Therefore, if such a distribution of chondrules existed, for all or only certain classes of chondrites, during any phase of protoplanetary evolution, they probably do not grow by mutual collisions. At some time the chondrules would need to be incorporated into larger objects otherwise, e.g., by being accreted by existing planetesimals.

This, of course, begs the question of how the “existing planetesimals” accreted in the first place. In their last paragraph, they launched from their experimental data into speculation: maybe a projectile hits a target and knocks off ejecta, some of which is moving slower. The projectile sometimes causes decompression of the target. The energy of other projectiles can then be absorbed without launching significant ejecta; meanwhile, the dust of the previous impact might be re-accreted after a collision by gas flow.

Assuming this convoluted scenario, a permeable body might be able to grow at least within the size range tested. So in the dusty battlefield of collisions, it is “entirely possible” that a planetesimal could continue to grow, they concluded on a note of victory, however subdued by the obdurate lab results. This is no mere child’s play, however. That was clear from their penultimate sentence: “By observing deep impact channels, considering addition of new dust layers, and finding decompaction by collisions, it is very clear that the evolution of the morphology of a growing body is highly complex.


1.  Jeff Cuzzi, “Planets: the first movement,” Nature 448, 1003 (30 August 2007) | doi:10.1038/4481003a.
2.  Paraskov, Wurm and Krauss, “Impacts into weak dust targets under microgravity and the formation of planetesimals,” Icarus, Volume 191, Issue 2, 15 November 2007, Pages 779-789.
3.  “The procedure for building this target is as follows. The target is initially built in a regular fashion as a highly porous target. The top is then covered by a plate which becomes the bottom as the target is turned upside down. The initial bottom becomes the target surface for the impact experiment. This marginal modification of the target seems sufficient to change the outcome of an impact from mass loss to mass gain for slow dust aggregate projectiles. As we have only one experiment with this behavior a more detailed study on the effect of porosity for collisions has to be carried out before further conclusions can be drawn.”

We have shown you once again a stark contrast between the bluffing that goes on in the news media and the hard realities of experimental science. Some may find it sufficient to believe that this all works somehow, because clearly planets exist, even around other stars. But to think that science understands planet formation by natural processes, and that experimental science proves it, is a bad example of glittering generalities. A step toward true understanding requires some examination of the nitty gritty details behind such claims.

Leave a Reply