Discrete element simulation of the charge in the hopper of a blast furnace, calibrating the parameters through an optimization algorithm

Abstract

AbstractThe purpose of this study is to simulate the distribution of a coarse granular material discharged in a hopper via a conveyor belt. This simulation is intended to be a model calibration for an optimization that will be later performed to obtain a proper material distribution device. From the hopper, the material is discharged in a blast furnace. Hence, achieving an adequate distribution in the hopper is crucial, since that distribution is directly linked to how the material spreads in the blast furnace, and this heavily influences the efficiency of the whole steel-making process. The apparatus is modeled by online three dimensional Computer-Aided Design software Onshape. Rocky DEM, a Computer-Aided Engineering software based on Discrete Element Method, is used to simulate the charge. The parameters of the numerical model are calibrated through an optimization algorithm. This phase is realized thanks to the optimization platform modeFRONTIER, using an algorithm that exploits meta-models to reduce the computational time of the optimization. By comparing the simulated results with the visual data obtained from blast furnace plant, the goal is to validate the model and to better understand the behavior of the whole charging process.