Characterization of Simultaneous Evolution of Size and Composition Distributions Using Generalized Aggregation...

Area of Science:

• Chemical Engineering
• Pharmaceutical Sciences
• Computational Modeling

Background:

• Multi-dimensional population balance equations (PBEs) are crucial for simulating complex granulation processes.
• Accurate prediction of particle compositional distribution requires precise modeling of component mixing.
• Existing numerical schemes often prioritize size number density over mixing accuracy.

Purpose of the Study:

• To compare the accuracy and applicability of the cell average technique (CAT) and the finite volume scheme (FVS) in predicting the mixing state of granulation.
• To evaluate the performance of these schemes in predicting the granule compositional distribution.
• To assess the prediction of the sum-square χ2 parameter for quantifying mixing.

Main Methods:

• Simulation of multi-component granulation using multi-dimensional PBEs.
• Implementation and comparison of the cell average technique (CAT) and finite volume scheme (FVS).
• Quantification of mixing using the sum-square χ2 parameter and analysis of integral moments.

Main Results:

• The finite volume scheme (FVS) accurately predicts integral moments but inaccurately predicts the mixing (χ2 parameter) in bicomponent systems.
• The cell average technique (CAT) offers moderate accuracy for moments but higher precision for the χ2 mixing parameter.
• Both CAT and FVS accurately predict average particle size, with FVS being computationally faster.

Conclusions:

• For pharmaceutical granulation simulations where compositional distribution is critical, the cell average technique (CAT) is preferred over the finite volume scheme (FVS) for its superior mixing prediction.
• While FVS is faster, its inaccuracy in predicting the mixing state can be a significant drawback.
• Accurate prediction of mixing is essential for understanding and controlling granule properties in pharmaceutical manufacturing.