Modified Application of Johanson's Model towards Roller Compaction

Modified Application of Johanson's Model Towards Roller Compaction

Jasmine Rowe, John Crison

Roller compaction (RC) is a continuous dry granulation process commonly used in the pharmaceutical industry to improve powder flow and blend uniformity via powder densification.  RC has gained substantial attention over the last couple of decades because neither moisture nor heat is required for granulation (as is the case for wet granulation and hot melt extrusion, respectively).  The process by which densification is achieved is through powder being fed between two rolls that are rotating in opposite directions of each other.  This counter-rotating motion of the rolls and friction between the powder and the roll surface draw the powder through the rolls and maximum compaction (densification) occurs when the separation between the rolls is at its minimum.  The resultant compact, or briquette, may then be milled to achieve the desired granule size.

While the RC mechanical process appears to be simple, efforts to quantitatively model this process have proven challenging because of complex material behavior in the feeding and compaction zones.  Consequently, a trial-and-error approach has typically been adopted towards RC process design, which is both time- and resource-intensive.  In response to the need for a more efficient approach towards RC processing, Johanson developed one of the most widely-used models for RC simulation, a powder mechanics model based on Jenike's steady state flow theory that predicts the physical properties of the final compact based on material and process parameters [Johanson, J. R. (1965). A rolling theory for granular solids. ASME, Journal of Applied Mechanics Series E, 32(4), 842–848].  However, practical application of Johanson's model to experimental work has been limited because one of the model input parameters, the pressure being applied to the feed powder as it enters the rolls (P₀), is difficult to obtain experimentally without sophisticated pressure sensor-instrumented equipment.

In this work, an alternative approach has been established, which expands upon Johanson's model to enable its incorporation into a daily work flow.  This new method requires only standard, routinely-measured, process parameters as inputs and is capable of solving for P₀ a priori to extensive experimentation.  An excellent correlation between simulated and experimental results has been achieved for both placebo and active blends.  Additionally, a dimensionless relationship between key process input parameters and final compact properties has been elucidated, which provides valuable information for efficient scale-up procedures.