Metamodelling of structural design parameters affecting silo discharge performance

Abstract

The work presented in this paper demonstrates the use of metamodels to capture the relations between the discharge flow and the design parameters of a silo. Based on simulation data for a discharging silo a metamodel is developed that relates the size of the discharge opening and the hopper angle to the performance of the silo. The created metamodel is a response surface model which is trained with numerical data from DEM simulations. The results show that with a full factorial DoS a reliable response surface can be created that predicts the simulation results in the points in between the training data. It can be concluded that metamodelling techniques are a promising tool for the prediction of behaviour of granular matter in silo designs and in designs of other types of bulk handling equipment.