The reverse shock in the ejecta of core-collapse supernovae is potentially able to destroy newly formed dust material. In order to determine dust survival rates, we have performed a set of hydrodynamic simulations using the grid-based code AstroBEAR in order to model a shock wave interacting with clumpy supernova ejecta. Dust motions and destruction rates were computed using our newly developed external, postprocessing code Paperboats, which includes gas drag, grain charging, sputtering and grain-grain collisions. We have determined dust destruction rates for the oxygen-rich supernova remnant Cassiopeia A as a function of initial grain sizes and clump gas density. We found that up to 30 % of the carbon dust mass is able to survive the passage of the reverse shock if the initial grain size distribution is narrow with radii around ∼ 10 − 50 nm for high gas densities, or with radii around ∼ 0.5 − 1.5 µm for low and medium gas densities. Silicate grains with initial radii around 10 − 30 nm show survival rates of up to 40 % for medium and high density contrasts, while silicate material with micron sized distributions is mostly destroyed. For both materials, the surviving dust mass is rearranged into a new size distribution that can be approximated by two components: a power-law distribution of small grains and a log-normal distribution of grains having the same size range as the initial distribution. Our results show that grain-grain collisions and sputtering are synergistic and that grain-grain collisions can play a crucial role in determining the surviving dust budget in supernova remnants.