Asteroids respond to impact stresses differently from either laborator
y specimens or large planets. Gravity is typically so small that seism
ic disturbances of a few cm s(-1) can devastate unconsolidated topogra
phy. Yet the presence of regolith and the likelihood that many asteroi
ds are gravitational assemblages tell us that gravity cannot generally
be ignored. We use numerical models for impact fracture in solids to
examine the initial stage of crater formation on asteroid 243 Ida, up
to the cessation of fracture and the establishment of the cratering fl
ow; at this stage we can infer final crater diameters but not profiles
. We find that a modified strength scaling applies for craters up to a
few 100 m in diameter forming in rock subject to Ida's gravity, and t
hat gravity controls all craters larger than similar to 1 km. ''Bright
annuli'' around a number of intermediate craters may be the result of
low-velocity surface disturbances, rather than bright proximal ejecta
deposits. We also consider large impactors, to which Ida presents a c
urved, finite target surface with irregular gravity. These can excavat
e asymmetrical concavities. Stresses from large events can refocus and
cause fracture far from the crater; using the shape of Ida as a basis
for 3D hydrocode simulations, we show that impact genesis of the Vien
na Regio concavity can cause fracture in Pola Regio, where grooves are
observed in spacecraft images. Other simulations indicate that the fo
rmation of the similar to 10 km crater Azzurra might have reopened the
se fractures, which may account for their fresh appearance. This mecha
nism of groove formation requires an interior which coherently transmi
ts elastic stress. While this precludes a classic ''rubble pile'' aste
roid, it does allow well-joined fault planes, and welded blocks or por
es smaller than the stress pulse. (C) 1996 Academic Press, Inc.