The heat transfer coefficient was higher in the spout
compared to the annulus. In order to characterise the bed,
there was a hypothesis that sawdust not participating in the
spout at all times would have poor thermal contact with the
steam. Thus, this would come up as a lower heat transfer in
the experiments than in the model [30]. In the KaU dryer, the
average ratio between the experimental and the model heat
transfer coefficients was approximately 0.7. If Eq. (1) was used
for a given logarithmic mean temperature difference, then the
volumetric heat transfer coefficient would be 30% lower in the
experiments than in the model. This indicates that only about
70% of the surface area of the material was in thermal contact
with the steam. This analysis supported the visual observations.
Furthermore, the visual observations showed that the
fraction of the material in the annulus increased with more
material in the dryer.