• Citer

A CFD-based approach to predict pea protein solubility improvement by high pressure homogenization in a complex valve geometry

par Nassouh, Soulaimane ;Pauline, FAUQUET;Abuabdou, Salahaldin ;Debaste, Frédéric
Référence 2026 International Food Operations and Processing Simulation Workshop (FOODOPS)
Publication Non publié, 2026-06-15

Communication à un colloque

• Accès en ligne
• Détails
• Contenu
• Statistiques

Résumé :

Plant-based proteins are increasingly considered as sustainable alternatives to animal-based proteins, but they often exhibit inferior functional properties, limiting their use in food formulations. Among these properties, protein solubility is regarded as one of the most critical, as it influences the industrial use of a product and most other functional qualities, including emulsification, gelation, and foaming. High-pressure homogenization (HPH) is often cited as an effective processing method to improve the solubility of plant protein suspensions by reducing particle size, unfolding protein structures, and exposing hydrophilic residues. Despite its growing industrial adoption, the mechanistic understanding and predictive modelling of solubility enhancement through HPH remain limited, particularly regarding the local hydrodynamic conditions within actual valve geometry used in most homogenizers. In this work, a laboratory-scale HPH valve geometry is investigated by a Computational Fluid Dynamics (CFD) approach. The CFD model resolves the local flow field within the valve, capturing key stresses, including turbulent viscosity (μ_t) and turbulent dissipation rate (ε), that are hypothesized to drive protein structural modifications. While the mathematical model successfully reproduced the overall experimental solubility enhancement measured, it also revealed that applying a strictly linear 1rst order source term overestimates the disruptive impact of mild, pre-gap wall shear. This finding highlights the fact that protein aggregates likely possess a critical mechanical yield stress that must be exceeded prior to fragmentation, leading to protein solubility increase. Ultimately, this work lays a robust computational foundation that, through the future integration of dissipation thresholds, can serve as a powerful predictive tool for HPH geometries.