Optimizing model parameters for accurate prediction of ribbon density in roller compaction

Abstract

The accurate prediction and adjustment of ribbon density is crucial in roller compaction; a continuous dry granulation process widely used in pharmaceutical manufacturing. This study presents an optimization of model parameters for the prediction of relative ribbon density (RRD) using a grid search method.

Unlike previous approaches that assumed constant inlet pressure (P₀) and pre-consolidated density (γ₀), this work establishes their dependency on process parameters, including roll speed, roll gap, roll force, and screw speed. The optimized model parameters significantly improved the agreement between theoretical predictions and experimental results, achieving a root mean square error (RMSE) of 0.024.

Highlights

Grid search optimization was applied to predict ribbon density.
Inlet pressure and pre-consolidated density linked to process parameters.
Empirically calibrated pre-consolidated density enhanced ribbon density prediction.
Experimentally inaccessible variables are effectively fitted using the proposed approach.
Enhanced process control understanding in pharmaceutical roller compaction within the investigated setup.

Experimental studies were conducted on a roller compactor with varying process conditions to generate data for model calibration and validation. An optimized fitted compressibility constant (K) of 12.88 was determined, followed by the effective fitting of P₀ and γ₀ using grid search. Sensitivity analysis based on the formulation and the operating conditions from experiments, revealed that the fitted pre-consolidated density, influenced by all process parameters, has a strong model derived association within the calibrated model for RRD prediction, whereas roll force had a comparatively minor impact.

The findings underscore the importance of incorporating dynamic model parameters in predictive frameworks to enhance the accuracy of ribbon density estimations. This approach offers an empirical calibration basis for optimizing roller compaction processes in pharmaceutical applications.

Download the full article as PDF here Optimizing model parameters for accurate prediction of ribbon density in roller compaction

or continue reading here

Table 1. Details of materials used in this work.

Material	Product name	Supplier	Weight ratio / %
Mannitol	Pearlitol 100SD	ROQUETTE FRERES, Lestrem, France	79
Microcrystalline cellulose	Ceolus UF-711	Asahi Kasei Co., Tokyo, Japan	20
Magnesium Stearate	HyQual 5712	SpecGX LLC, Missouri, USA	1

Shubhangini Awasthi, Srikanth R. Gopireddy, Shuichi Tanabe, Susumu Hasegawa, Thomas Profitlich, Regina Scherließ, Nora A. Urbanetz, Optimizing model parameters for accurate prediction of ribbon density in roller compaction, Powder Technology, Volume 487, Part 1, 2027, 123376, ISSN 0032-5910, https://doi.org/10.1016/j.powtec.2026.123376.