The extent of void growth and cracking due to ductile fracture occurri
ng during symmetric Taylor cylinder impact tests on leaded brass has b
een determined experimentally. Void growth occurs within these predomi
nantly compressively-loaded specimens through the development of large
tensile hydrostatic stresses along the specimen axis near the impact
face during expansion of the cylinder, termed ''mushrooming''. The mea
sured porosities have been compared to predictions using a constitutiv
e model based on the Gurson (1975, Ph.D. Thesis, Brown University) yie
ld function, implemented within the DYNA2D finite element code. The in
itiation of void coalescence and subsequent crack development was also
predicted using the approach of Tvergaard and Needleman (1984, Acta M
etall. 32, 157) based on a critical porosity criterion. The calculatio
ns were able to qualitatively predict the development of the porous zo
ne and void coalescence within the impact specimens; however, the pred
icted void growth exceeded that observed experimentally and the predic
ted extent of void coalescence was too large. It is suggested that the
primary source of error lies in excessively high predicted void growt
h rates using the Gurson yield function at high stress triaxiality lev
els.