Abstract
dc:description.abstractEffective cleaning is a critical industrial process aimed at minimising cross-contamination in the production of food, beverages, and pharmaceutical products by removing unwanted deposits or soils. However, accurately predicting cleaning performance remains challenging due to limited understanding of the fundamental interactions between fluids and soils that govern the cleaning process. In this study, three different flow configurations that simulate typical industrial flow characteristics were investigated: a radial flow cell, a slit flow cleaning cell and a backward-facing step. Despite their geometric simplicity, these configurations often exhibit complex flow features such as turbulence and recirculation zones, which have significant impacts on the cleanability of the processing equipment and the kinetics of soil removal. Instant coffee was identified as a suitable soluble model soil for the experiments, with thin, dry layers of soil prepared using softened water as the cleaning fluid. While these layers were generally flat in the centre, they formed raised edges at the periphery. Cleaning experiments were conducted for each configuration, and kinetic data were collected. A fundamental zeroth-order kinetic model for the soluble soil was developed, showing strong agreement with experimental results and accurately predicting soil thickness profiles. A computational modelling approach was developed that assumed the flow field and mass transport are decoupled. The soluble soil was modelled as a boundary condition using the species transport model to simulate the pseudo-steady-state behaviour of the transient process, rather than employing conventional multiphase flow simulations. Previous work using this approach, which relied on experimentally fitted model parameters, showed reasonable agreement in some cases. To improve accuracy, an enhanced model incorporating experimentally measured parameters was developed to evaluate whether it could more reliably predict cleaning performance. The findings indicate that while accurate predictions can be achieved using a cost-effective modelling approach that simplifies the geometry by utilising 2D rather than 3D simulations, this approach often requires high mesh resolution near the wall to accurately capture the cleaning dynamics.
Degree
thesis:*- Name dc:type.qualificationname
- Doctor of Philosophy (PhD)
- Level dc:type.qualificationlevel
- Doctoral
- Grantor dc:publisher.institution
- University of Cambridge
- Year dc:date.issued
- 2024
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Deshmukh, Karthikeya Prashant
- Advisor dc:contributor.advisor
-
- Wilson, David Ian
Subjects
dc:subject × 8Rights
dc:rightsIdentifiers
dc:identifier.*- DOI dc:identifier.doi
- https://doi.org/10.17863/CAM.116594
- OAI identifier oai:identifier
- oai:www.repository.cam.ac.uk:1810/381395