Figure 16. The distribution of the orientation of wood particles at 3 Umf, Hb0=15.24 cm, and varying wood particle counts.
Figure 15 shows the probability density function (PDF) of wood particle orientation in the fluidized bed with both shallow (Hb0= 7.62 cm) and deep bed (Hb0 = 15.24 cm) configurations and a superficial gas velocity of 2Umf and 3Umf. The orientation is quantified using the angle between the detected major axis of the wood particle and z -axis of the bed, with ranges from 0° (vertical) to 90° (horizontal). Currently, the measurements are inherently two-dimensional over the x-z plane which is incapable of measuring accurately all wood particle’s orientation relative to z-axis. As an alternative, to minimize the measurement error due to the three-dimensional projection, the orientation statistics samples only particles with an aspect ratio larger than 1.8. Namely, the particles that lie almost in the x-z plane are conditionally sampled in the orientation statistics. As the distribution shows, for all cases, the distribution has a dip around 45° and slightly favors both the vertical and horizontal orientation, respectively. This presumably corresponds to the wood particle fluidization in the dilute and dense LDPE particle regions of the bed. Namely, the wood particle tends to be horizontal when reaching the dense LDPE particle pool while staying vertical when falling after the slug disintegrates. These results are in qualitative agreement with previous experimental results by Vollmari et al. 201614 with a comparable superficial gas velocity range. For example, for the long cylindrical wood particle case (14 mm × 4 mm wood cylindrical particle) in the study, the superficial gas velocity varies from around 1.6 to 2.4 Umf, all the orientation statistics show local peaks for both vertical and horizontal particle orientation. Even though the experiment was conducted using a single non-spherical particle type. Figure 16 shows the orientation distribution of the wood particles at a fixed superficial gas velocity (3Umf) and a deep bed configuration with varying wood particle counts from 50 - 200. The results show similar trends as described above without significant variation between the tested cases.
Apart from the orientation statistics of the wood particles, the velocity has also been acquired based on the tracked wood particles. Figure 17 shows both the horizontal and vertical velocity of wood particles centroid. As expected, for horizontal velocity (Figure 17a), the distributions are well presented by a normal distribution with a mean of around zero for all cases. The standard deviation increases slightly with the increase of superficial gas velocity from 2Umf to 3Umf, and decreases with the increase of the static bed height from 7.62 cm to 15.24 cm. For the vertical velocity (Figure 17b), the distributions are non-gaussian and spread more widely compared to the horizontal velocity from around -3 m/s to 3 m/s. As the superficial gas velocity increases with a fixed static bed height, the standard deviation also increases. For the Hb0=15.24cm case, the distribution becomes bimodal. With the increase of the bed height, the distribution shifts to more negative values possibly due to the increase of mean bed height (c.f. Figure 13) and thus an increase in the downward wood particle velocity. Furthermore, the effects of wood particle counts on the wood particle velocity are shown in Figure 18. Similarly, the distributions remain a normal distribution with zero mean for all cases. The standard deviation increases slightly with the increase of wood particle counts. For the vertical velocity distribution, all three cases are bimodal. The first mode is at around zero whereas the second mode is centered at -1.5 m/s, -1.75 m/s and -1.85 m/s for the 50, 100, and 200 wood pellets cases, respectively.