The paper deals with the evaluation of the distribution of the porosity of a model of the HTR-10 packed bed reactor which was generated using a Discrete Element Method (DEM) approach. The porosity of the packed bed was determined using semi-analytical methodologies. To evaluate the validity of the model, selected aspects of the DEM approach were first investigated. It was found that the appropriate packing method is to drop the spheres at a specified rate from the top of the packed bed domain. The preferable model for the contact force between the spheres was found to be the Hertz-Mindlin model. It was also established that the Young’s modulus of the spheres should be larger than 0.1 GPa. A value of less than 0.1 GPa has a marked negative effect on the porosity. The values for the porosity predicted by previously proposed correlations for the radial and axial variations in porosity were determined to be too small. From the analysis of the packed bed model, the overall porosity was found to be 0.382, whilst the average porosities of the pipe, conical and cylindrical sections were found to be 0.439, 0.378 and 0.381 respectively. Correlations are derived to describe the axial variation of the porosity in the pipe, conical and cylindrical regions. A piece-wise constant description of the distribution in the porosity is finally provided by dividing the packed bed into 34 regions and determining the average porosity of each region.
Read full abstract