**4. Results and Discussions**

Figure 6 shows the relation between the normalized discharging time and the mass fraction of each size range of the sinter particles that were discharged from the funnel. It is confirmed that the smaller particles were discharged during the initial period, whereas the larger ones were discharged later. The particle size segregation occurred during charging and discharging at the storages (the surge hopper and the parallel hoppers); therefore, the larger particles tend to be discharged later. This phenomenon

affects the radial particle size distribution of the burden. This result of the time change of the particle size during discharging was used as the input particle condition of DEM.

**Figure 6.** Relation between the mass fraction of each size range of the sinter particles and the normalized discharging time.

Figure 7 shows pictures during ore charging in the experimental test. The black particles in the furnace are coke, and the dark brown ones, which are charged from the rotating chute, are sinter. The sinter particles are stacked near the wall at the beginning of charging, and subsequently, the particles flow toward the center. The sinter covers the coke layer after the 11 rotations. Figure 8 shows the surface profile of the burden after charging. The surface angle of the ore layer is approximately 32.1◦. A terrace is found around 1185 mm from the center, and the angle of the terrace is 13.7◦. The surface angle of the coke layer is also found to be 36.7◦. This value is larger than that of the sinter layer because of the particle shape and the size distribution.

**Figure 7.** Pictures during ore-charging in the experimental work. (**a**) 1st rotation; (**b**) 4th rotation; (**c**) 7th rotation; (**d**) 11th rotation.

**Figure 8.** Surface profile of the burden after the charging test.

Figure 9 shows the contour map for the sinter volume fraction in the burden, which was obtained by digging up the burden. The surface profiles subsequent to charging are also drawn in the contour. Most coke particles are situated near the wall, and the thickness of the coke layer around the center is extremely thin. Therefore, it is suggested that a collapse of the coke layer during ore-charging was not significant in this charging condition.

**Figure 9.** Contour map for the sinter volume fraction in the burden.

Figure 10 shows snapshots of the charging behavior simulated by DEM. The brown particles denote sinter and the blue ones are coke. The sinter particles are charged around the wall toward the center, and they reach the center in approximately 11 chute rotations. Their behavior is found to be very similar to that of the experimental ones, which are shown in Figure 7. Figure 11 shows a cross-section of the burden simulated using DEM. The surface angle is 33.6◦, and the position of the terrace is approximately 1250 mm from the center. The terrace angle is approximately 14.2◦. Although the profile and the thickness near the center are slightly different, the simulated burden layer is quite similar to the experimental one, especially the shape of the terrace.

**Figure 10.** Snapshots of ore charging simulated by DEM. (**a**) 1st rotation; (**b**) 4th rotation; (**c**) 7th rotation; (**d**) 11th rotation.

**Figure 11.** Cross-section of the burden simulated by DEM.

Figure 12 shows the relation between the ore to coke mass ratio (O/C) and the radial distance. The experimental results were obtained by digging up the burden. The value of O/C around the center is enormous because the thickness of the coke layer is thin, and it decreases with the increase in the radial distance because the coke layer becomes thicker. Trends of the O/C for both the experiment and the simulation are excellently correlated. Figure 13 shows the relation between the normalized mean particle diameter of the sinter and the radial distance. The particle diameter at each position was normalized using the mean value of all particles because the mean particle diameter in the experimental work became smaller than that of the initial condition. Some sinter particles had fragmented during the continuous charging into hoppers and discharging. The particle diameter at the terrace is smaller, and it increases while approaching the center due to the particle size segregation during the flow toward

the center. A good agreement between the experimental results and the simulated ones was obtained. Therefore, the burden distribution simulated in this work was validated, and it has a high potential to predict the particle behavior during charging. In the simulation, the burden layer formation can be clarified. Figure 14 shows the cross-section of the ore layer, which is color-coded by the chute tilting angle. It is found that the thickness of each layer is thin, and it reaches the center. This is good information for considering the mixing particles in the ore layer. The vertical layer structure for each ring is clarified. Therefore, the discussion of the vertical position of mixed particles is possible. For example, if some particles should be mixed in the purple layer, they should be charged in the blast furnace during the rotation of 39◦. It should be noted that this simulation did not consider the effect of the mixing particle and gas flow, but they will be investigated in the future work.

**Figure 12.** Relation between the ore-coke mass ratio (O/C) and the radial distance.

**Figure 13.** Relation between the normalized mean particle diameter of sinter and the radial distance.

**Figure 14.** Cross section of the ore layer, which is color-coded by the chute tilting angle.
