In industrial milling operations that use comminution and wet-comminution techniques, the reduction of the particle size is usually achieved through crushing the sample with a material harder than the product. These methods are convenient when the required median particle size is above 400 um. However, to obtain post-milling particle distributions with 85% sub-micron particles (in number) is both energy intensive, and time consuming. For conventional milling machines, to have the required output in several ton/hr of a product, having a large number of particles in the micron or sub-micron sizes at an affordable rate is cumbersome.Here, a wet-comminution machine that has shown to achieve the aforementioned milestones in the laboratory scale is studied. However, when the machine is scaled to industrial processes, it was recorded that some of the product variables are difficult to scale. In these studies, we attempt to understand the mechanisms by which this machine operates in order to achieve successful scaling. The apparatus operates completely on fluid mechanics principles, it consists of two concentric cylinders, the inner cylinder that has a smaller radius than the outer, rotates while the larger is held stationary. The inner cylinder is also shorter in length than outer, hollow in the inside and has transversal holes where the shaft attaches to the apparatus. The apparatus can operate in batch condition, where the liquid volume is much less than the volume of the apparatus, typically 0.3Vt, 0.42Vt and 0.54Vt. In addition, the apparatus can operate with throughflow, which the upper plate covering the apparatus is reduced in radius.Two component Laser Doppler Velocimetry (LDV) was used to obtain even-time averaged statistics of the azimuthal and axial velocities, in the gap and underneath the impeller. Also, Flow Visualization using Kalliroscopic particles was performed as means of observing large scale structures in the gap. Moreover, single plane Particle Image Velocimetry (PIV) was used to acquire statistics of the axial and radial velocities in the gap, and both underneath as well as above the inner cylinder.It was found that at both throughflow conditions, the topology of the apparatus creates a free spinning boundary both at the bottom and above the inner cylinder. Near the bottom, the thickness of the boundary was found to decrease with Reynolds number to a limiting value, where Re; is based on gap thickness and inner cylinder tip speed. For Re > 2546, the liquid/air interface thickness is constant for a given holding volume. In the regions above and underneath the inner cylinder, corner vortices were detected; if viewing the left-hand-side, the lower one rotating counter clockwise, while the upper rotates clockwise. The thickness of these vortices was found to be constant for various axial flows at Re = 1110 and 2230. The radial length scale of the stationary vortices was found to be ~2.5d;.The flow generated inside of the gap was characterized to have Taylor vortex signatures. It was found that the length scale of the Taylor vortices in the gap is rather insensitive to Reynolds number or holding volume ratio. The average vortex pair wavelength; was found to be 3.6d. Average flow statistics in batch condition indicate that in the gap, at Re = 1110 and 2230, the azimuthal velocity is 0.5U over much of the length. Similarly, it was found that the net axial flow through the gap is close to zero.