Unravelling gravel: Including stochastic behaviour of granular materials in design of bulk handling equipment

Granular materials are all around and have many secrets that still need to be unravelled. This thesis shows how the stochastic behaviour of granular materials can be identified and efficiently included in design of bulk handling equipment. To achieve this, metamodels play a key role as they are able to capture trends of physical models and give fast predictions. In this research it is shown what the opportunities and limitations of metamodels are in a hopper case study based on simulation data from a discrete element model (DEM). Next a procedure is presented in which the stochastic behaviour and metamodels are combined to calibrate a DEM model including the stochastic behaviour of the granular material. The stochastically calibrated DEM model is accurate and used as the basis for a design optimization case study in which the effect of robust optimization was evaluated and validated. These studies combined show that stochastic behaviour of granular material can be included in design of bulk handling equipment and lead to improved and robust designs.