Processing of lamellar structured liquids

Cunningham, Grace Ellie ORCID: 0000-0002-6214-8508 (2025). Processing of lamellar structured liquids. University of Birmingham. Eng.D.

Cunningham2025EngD.pdf
Text - Accepted Version
Available under License All rights reserved.
Download (4MB)

Abstract

In today’s society where health and sustainability are two key consumer trends, global FMCG companies like Unilever are driven to innovate, adapt, and adopt responsible practices. Aligning with consumer expectations enhances Unilever's market position and supports long-term business sustainability. Therefore, this thesis aims to improve insights into the hair conditioner manufacturing process to reduce energy and resource consumption while ensuring the desired product microstructure (a lamellar gel network) can be manufactured efficiently every time.
Hair conditioner improves manageability and enhances the appearance and shine of hair - a key indicator of hair health. Consumers expect a thick, opaque cream product that spreads evenly over hair and provides detangling and conditioning through deposition of beneficial ingredients. These physical and rheological properties are achieved through a lamellar gel network (LGN) microstructure - a product of the ingredients used and the manufacturing process. Insights into the influence of processing conditions on the LGN microstructure are limited due to the complexity of the product limiting the availability of process monitoring and characterisation techniques.
In this thesis, rheological mapping using a rheometer and 3D-printed scaled-down geometries to imitate the batch manufacturing vessel, was applied to generate viscosity-time profiles for LGNs at various speeds, times, and temperatures. Samples were characterised by their rheological properties and power requirements, comparing yield stress to qualitatively assess process optimisation strategies. Shorter mixing times, ending after the peak viscosity was reached, produced higher yield stress products, presenting energy-saving opportunities. The importance of monitoring viscosity during LGN mixing was highlighted, informing further studies on the application of mixer-viscometer approaches to
partially-filled batch vessels (25% to 100% liquid height to impeller length). The torque curve method and Couette analogy were applied to torque-speed data for various fluids, including LGNs. However, changes in fluid contact with the impeller as a function of speed, geometry, and fluid rheology showed no measurable relationships, limiting the application of mixer-viscometer techniques due to the consequent impact on torque measurement. Finally, data-driven models were investigated to predict online viscosity from torque for mixing systems at three scales (0.045 L, 2 L. 50 L). A random forest regression model used mixer diameter, speed, fill level and torque as inputs, and apparent viscosity as the output. The model indicated signs of over-fitting, likely due to an uneven dataset favouring lower viscosity values. Despite this, the work provides an initial contribution towards applying soft sensors to predict online viscosity during manufacture of formulated products.
This research provides Unilever with valuable techniques for monitoring microstructure formation using rheological approaches to enable process and product optimisation, aiding in achieving their net-zero targets and enhancing product superiority.

Type of Work:

Thesis (Doctorates > Eng.D.)

Award Type:

Doctorates > Eng.D.

Supervisor(s):