Scale-independent solid fraction prediction in dry granulation process using a gray-box model integrating machine...

Area of Science:

• Pharmaceutical Engineering
• Materials Science
• Chemical Engineering

Background:

• Solid fraction prediction is vital for controlling product quality in roller compaction.
• The Johanson model, a first-principles approach, requires unmeasurable preconsolidation properties for roller compaction.
• Existing methods lack the ability to predict solid fraction accurately under varying roller compaction conditions.

Purpose of the Study:

• To develop a novel gray-box (hybrid) model for predicting solid fraction after roller compaction.
• To overcome the limitations of the Johanson model by incorporating measurable parameters.
• To provide a robust and scalable method for solid fraction prediction across different formulations and equipment.

Main Methods:

• Developed a statistical model to predict a novel preconsolidation parameter using material and process data.
• Integrated the statistical model with Johanson's first-principles model to create a gray-box model.
• Validated the model's performance across various roll speeds and conditions, including high throughput.

Main Results:

• The statistical model successfully predicted the preconsolidation parameter without requiring roller compaction experiments.
• The gray-box model accurately predicted solid fraction, even under conditions with powder velocity gradients.
• The model demonstrated robustness and applicability across different scales and formulations.

Conclusions:

• The proposed gray-box model offers a significant advancement in predicting solid fraction during roller compaction.
• This approach enhances product quality control by enabling accurate prediction without complex experimental setups.
• The model's scalability and adaptability make it a valuable tool for pharmaceutical and materials processing.