Tumbling fragments from experiments simulating asteroidal catastrophic disruption

Recent analyses of radar and photometric observations of several asteroids including 4179 Toutatis (discussed in A, W, Harris, 1994, Icarus 107, 209-211), and 253 Mathilde (S, Mottola et al., 1995, Planet. Space Sci, 43, 1609-1613) have highlighted the fact that non-principal-axis rotation (sometimes just called ''tumbling'') is detectable and possibly wen commonplace for small slowly rotating asteroids, We report results from catastrophic impact experiments carried out in 1989 and 1992, designed to simulate the disruptions in the asteroid belt, A small fraction of the fragments emerging from the disrupted targets have been observed rotating in a tumbling mode, which is not in itself surprising but has never before been studied in impact experiments of this kind. Although the experimental data, in the form of digital video sequences, did not allow us to determine in detail the rotational state of the fragments, we have been able to identify in a reliable way 13 tumbling fragments, alongside the simple rotators that constitute the majority of ejected fragments, A key point regarding the reduction of data from these experiments is that not all tumbling fragments are visibly detectable as such: To estimate the true proportion of tumbling fragments, we develop a simple theoretical model that assumes initially that all fragments are tumbling, then apply realistic selection effects to allow comparison with the laboratory results. We find that the experimentally observed fraction of fragments that are visibly tumbling-approximately 6% of measured rotators-is not consistent with an isotropic distribution of angles between the spin vector of a given fragment and its short principal axis for an oblate ellipsoidal model. (C) 1997 Accademic Press.