Multivariate modeling of protein ingredient properties for optimized extrusion processing

Overview

Protein ingredients used in extrusion processing exhibit highly variable technological properties depending on species, composition, and extraction method. These properties, such as water absorption capacity and protein solubility, directly affect critical extrusion parameters including moisture levels, screw speed, and energy requirements, ultimately shaping the quality of the final extrudate. This research proposes a multivariate approach to identify which technological properties most strongly predict product quality, enabling the design of tailored ingredient mixes and optimized processing conditions for targeted extrudate characteristics.

Technical specifications

Research approach:

• Design of experiments combined with multivariate statistical analysis to model relationships between ingredient properties and extrusion outcomes
• Evaluation of protein concentrates from dry-fractionation of pulse flour versus protein concentrates and isolates from wet extraction
• Measurement of key technological parameters including bulk density, water absorption capacity, water solubility index, and rheological behavior
• Assessment of product responses through microstructure analysis, textural profiling, and sensory evaluation
• Development of a predictive model to adjust extrusion conditions based on ingredient quality and desired extrudate characteristics

Key focus areas:

• Understanding how extraction technology influences protein functionality
• Correlating chemical composition with processing behavior and end-product quality
• Identifying critical control parameters for ingredient selection in extrusion applications

Technology readiness level

This research is at an early-to-mid stage of development. Prior literature and preliminary observations confirm that protein blend composition significantly affects extrusion conditions and product quality, with dry-fractionated concentrates showing lower water absorption and higher solubility compared to wet-extracted counterparts. However, the systematic multivariate investigation and predictive modeling proposed here represent novel work that has not yet been completed. Validation is planned through structured experimental design linking ingredient properties to extrusion performance and product quality outcomes.