Controlling fragment size distribution for modelling the breakage of multi-sphere particles

Voro-Pack, an open-source code is introduced to reconstruct real particles into multi-sphere clusters consisting of fragments with controlled size distributions, enabling experiment-informed fragment size distributions (FSD). Two types of silica sand are employed to evaluate the performance of the code, where FSD data are obtained through laser diffraction and sieving, and particle shape information is obtained through micro computed tomography. The results show that the particle sizes and shapes of the multi-sphere clusters generated by the code are in good agreement with those of real particles. It is found that the fragment sizes in the multi-sphere clusters aligned more closely with the real FSD data when laser diffraction data were used as input, compared to sieving. The volume ratio of the multi-sphere clusters to the actual particles is a key factor influencing the size distribution of fragments, with better matching to the actual FSD data when the volume ratio exceeds 0.5 and approaches 1.0. Additionally, the particle shape characteristics do not significantly affect the FSD data in the code-generated clusters. These findings suggest that the Voro-Pack code is an effective tool in reconstructing multi-sphere clusters for particles of various morphologies, providing valuable insights into modelling the breakage of granular materials. • An open-source code to control fragment size distribution based on experimental information in particle breakage is developed. • Volume ratio is a key factor to fragment size distribution of code-generated multi-sphere clusters. • Particle shapes do not significantly affect fragment size distribution in multi-sphere clusters generation. • Fragment sizes in multi-sphere clusters match real particles better with laser diffraction data than with sieving.