A validation of discrete-element model simulations for predicting tablet coating variability

Area of Science:

• Pharmaceutical engineering
• Computational fluid dynamics
• Process modeling

Background:

• Ensuring uniform tablet coating distribution and quantifying variability are significant challenges in pharmaceutical manufacturing.
• Computer simulations, specifically the Discrete Element Method (DEM), offer a promising approach for predictive design in coating processes.

Purpose of the Study:

• To evaluate the predictive accuracy of Discrete Element Method (DEM) simulations for tablet coating processes.
• To assess the impact of experimental and simulation input uncertainties on DEM model predictions.
• To validate DEM predictions against a comprehensive set of experimental data across different scales and conditions.

Main Methods:

• Conducted extensive tablet coating experiments across various scales, conditions, and tablet shapes.
• Developed a water-soluble formulation for rapid UV/VIS spectroscopic analysis of coating amounts.
• Utilized Discrete Element Method (DEM) simulations to model and predict coating distribution and variability.

Main Results:

• DEM predictions consistently fell within experimentally determined confidence intervals.
• Achieved a mean absolute comparison error of 0.54% between DEM predictions and experimental estimates of coating variability.
• Identified spray area size parameterization as the primary source of simulation error, though smaller than experimental uncertainties at larger scales.

Conclusions:

• Discrete Element Method (DEM) simulations demonstrate high predictivity for tablet coating processes.
• DEM is a valuable tool for the model-predictive design of industrial coating operations.
• Understanding and minimizing simulation input uncertainties, particularly spray area, enhances model accuracy.