*3.2. Simulation Conditions*

The particle behavior during charging into the experimental blast furnace was simulated using DEM, and the burden distribution was compared with the experimental results to validate the simulation results. The geometry of the rotating chute and the throat of the furnace were identical to those of the experimental results, and 5500 kg of sinter particles, with a particle density of 3300 kg/m3, were charged into the furnace. The size of sinter particle was 6 to 20 mm, and its particle size distribution corresponded to Figure 3. The total number of sinter particles was 2,545,086, and the detailed particle condition is tabulated in Table 2. Young's modulus and Poisson's ratio were assumed to be 3.5 GPa and 0.25, respectively. Only the ore-charging process was executed in the simulation, i.e., the coke layer prior to ore charging was arranged at the top of the furnace corresponding to the surface profile subsequent to coke charging in the experimental test. The particle diameter of coke is 7.5 to 30 mm; its density, Young's modulus, and Poisson's ratio are 1050 kg/m3, 0.54 GPa, and 0.22, respectively; and the number of coke particles is 1,049,686. Table 3 shows the detailed conditions for the coke particles. The input sinter particles were generated at the outlet of the discharging funnel, i.e., the particle behavior from the parallel hopper and the discharging funnel was not simulated to reduce the computing time. The time changes in the mass ratio for each particle of the input sinter corresponded to the sampled results in the experimental work, which was described above, to consider the particle size segregation during conveying to the furnace. The input particles have 5.5 m/s of vertical velocity, which was also measured using high speed video camera [20]. Every particle has a different rolling friction coefficient, which was obtained by generating a random number at the beginning of the simulation. The distribution of the rolling friction coefficient is shown in Figure 5. The charging pattern of the chute tilting angle was the same as that of the experimental one, as shown in Table 1, and the rotational speed was 13.4 rpm. The discrete time was 1.5 μs and the total number of calculation steps was 50 million. The calculation was parallelized using OpenMP.


**Table 2.** Condition for sinter particle in DEM.

**Table 3.** Condition for coke particle in DEM.

