Reverse ballistic experiments were used to investigate confinement, pre-damaged and intact, and rod size effects on penetration of long, gold rods into silicon carbide (SiC-N). Rod diameters were 1.0mm and 0.75mm, and lengths were 70mm and 50mm, respectively. Within data scatter, penetration velocity was the same for intact (bare or sleeved), pre-damaged (thermally shocked with non-contiguous cracks), and in situ comminuted SiC-N. Penetration velocity was independent of rod diameter within the data scatter. An expression for the penetration velocity versus impact velocity is found using linear regression. It is proposed that the reason there is no difference in the penetration rate between intact and pre-damaged (failed) SiC is because, after the first few microseconds following impact, the rod penetrates failed material in both cases.