Online batch recipe adjustments for product quality control using empirical models: application to a nylon-6,6...

Area of Science:

• Chemical Engineering
• Process Control
• Materials Science

Background:

• Process understanding is crucial for consistent product quality in batch manufacturing.
• Batch profile data contains valuable information about process dynamics and disturbances.
• Traditional methods may not fully capture the impact of disturbances on product quality.

Purpose of the Study:

• To apply batch profile characterization tools for enhanced process understanding.
• To identify primary disturbances affecting a nylon-6,6 process and their impact on product quality.
• To develop and evaluate an online recipe adjustment strategy for quality control.

Main Methods:

• Systematic study of historical profile data from a fixed-recipe nylon-6,6 process.
• Utilizing batch profile characterization to uncover disturbance signatures.
• Developing an online prediction model for final product quality.
• Implementing a simple online recipe adjustment strategy based on predicted quality deviations.

Main Results:

• Batch profile characterization successfully identified key disturbances impacting the process.
• Accurate online predictions of final product quality were achieved before batch completion.
• The proposed online recipe adjustments significantly reduced final product quality variation.
• Effectiveness of empirical prediction models from recipe-based data was discussed.

Conclusions:

• Batch profile characterization is an effective tool for enhancing process understanding and identifying disturbances.
• Online quality prediction and recipe adjustment can improve batch process control and product consistency.
• Further investigation into empirical prediction models for batch processes is warranted.