Search

This study investigated the flow patterns and wall stresses in a two-dimensional moving granular filter bed with a curved and symmetric louvered wall. A digital camcorder was used to record the flow pattern histories of the granular solids in the moving granular filter bed. The numerical predictions of the flow patterns were implemented using the discrete element method and compared with the experimental results. The variations in the dynamic wall stresses with time were obtained using a pressure gauge, which could simultaneously measure the normal and shear stresses of the granular solids. The effects of the curvature of the louvered wall and the exit width upon the flow patterns and the wall stresses were also elucidated. For a fixed value of radius of curvature, the ratio of the quasi-stagnant zone (Q-SZ) area in the third stage to the granular bed area in the third stage increases with decreasing width of the exit, but the average dynamic normal stress on the curved louvered wall increases with increasing width of the exit. For a fixed value of the exit width, both the ratio of the Q-SZ area in the third stage to the granular bed area in the third stage and the average dynamic normal stress on the curved louvered wall increase with increasing radius of curvature. Additionally, the curved louvered wall can not only shrink the area of the Q-SZ in the granular bed, but also reduce the dynamic normal stress on the louvered wall. [ABSTRACT FROM AUTHOR]

The flow patterns and stresses on the wall in a two-dimensional bin-hopper with the placement of an equilateral triangle flow corrective insert were investigated. The geometric shape and location of five kinds of equilateral triangle insert used in this study were obtained using the method previously proposed by Johanson. The flow pattern histories of the quartz sand in the bin-hopper were recorded using a digital camcorder. The static wall stress distributions above the insert produced by the quartz sand were measured. The variations in the dynamic wall stresses with time were obtained. As the insert half-angle increased, the region of the quasi-stagnant zone, which sat on the equilateral triangle insert, became larger and the position of the static normal stress peak moved downward. In general, the values of the dynamic normal wall stress measured by the pressure gauge, which was not close to the insert, were smaller than those under the test without the placement of an insert during the most part of material withdrawal. Nevertheless, some of the dynamic normal wall stress peaks were measured by the pressure gauge, which was close to the insert, as the insert half-angle was between 30° and 50°. [ABSTRACT FROM AUTHOR]