Next Contents Previous

9.3 N-Body Experiments

N-body simulations of the dissipationless formation of halos have been used to study halo shape. Dubinski and Carlberg (1991) found that dissipationless collapse gave rise to halos that were typically quite flattened. Specifically, they found that the mean values of the axis ratios were < c/a > = 0.50 and < b/a > = 0.71, where a, b and c are the axes of a triaxial ellipsoid. Thus, not only were the halos flattened, they were also strongly triaxial. Similar numbers of oblate and prolate spheroids were produced by these simulations.

Frenk et al. (1988) also found a preponderance of flattened halos in their simulations. However, while Katz (1991) found mainly triaxial halos in his dissipationless N-body simulations, the inclusion of a dissipative gaseous component led predominantly to oblate halos (Katz and Gunn 1991). Nevertheless, flattened halos appear to be a generic feature of such simulations.

Obtaining an observed distribution of halo shapes is still a long way off, but the results of Dubinski and Carlberg (1991) may be confronted with observation quite soon. One of the results of their simulations was that dissipationless collapse does not produce halos flatter than E6. Thus detecting halos more flattened than this would represent powerful evidence in favor of baryonic DM. The main concern is that such observations are only likely for rather unusual galaxies and, as noted above, it would be dangerous to draw conclusions about the shape and nature of halos from such special cases. Moreover, Merritt (1991) has raised the possibility that a bending instability may disrupt stellar systems that are dynamically hot and flatter than about E6.

One further curious result of Dubinski and Carlberg's (1991) study is that dissipationless collapse produces fewer round systems than occur in the distribution of elliptical galaxies. It may therefore be the case that under certain circumstances dissipation can actually make the resulting systems rounder rather than flatter as one would intuitively expect. Warren et al. (1992) have found better agreement between their N-body simulations and more recent observed elliptical shapes. However, Ryden (1992) finds that ellipticals are on average rounder than clusters of galaxies, suggesting that if clusters form without dissipating then ellipticals must have undergone dissipative collapse. While this introduces an added complication, it does not alter the conclusion that very flat halos can only form if there is substantial dissipation. An unambigous detection of an E8 halo would therefore be strong evidence for baryonic DM.

Next Contents Previous