MR LIAM BOSTON | The University of Aberdeen

MR LIAM BOSTON

MEng, AMIChemE

Research PG

About

Email Address

l.boston.19@abdn.ac.uk

Office Address

Room 201, Fraser Noble Building

School of Engineering

King's College

Aberdeen

AB24 3UE

School/Department

Combined Studies

Research

Research Overview

Computational fluid dynamics; lattice Boltzmann method; optimisation; mixing; anaerobic digestion

Research Areas

Engineering

View Research Area

Research Specialisms

Bioengineering
Chemical Engineering
Computational Physics
Fluid Mechanics
Mathematical Modelling

Our research specialisms are based on the Higher Education Classification of Subjects (HECoS) which is HESA open data, published under the Creative Commons Attribution 4.0 International licence.

Current Research

Project Title: In Silico Optimisation of Viscous Mixing of Shear-Thinning Liquids

Project Summary: Mixing condition optimisation is vital within anaerobic digesters to ensure a suitable balance between the competing objectives of minimising power consumption and maximising mixing quality. Existing relevant optimisations tend to focus on methane production at low solids concentrations (≈ 1 %), using traditional computational fluid dynamics solvers such as the finite volume and element methods, and take a one-variable-at-a-time approach to helical ribbon impeller design. The main aim of this project is to employ a systematic design of experiments optimisation framework with an efficient Lattice Boltzmann Method hydrodynamic solver to allow efficient multi-objective optimisation of mixing systems, with a target application of high-solids anaerobic digester design for sustainable chemical production. The solver and framework are to be tested and validated to prove their applicability.

The optimisation framework consists of four stages: 1. Data screening using a definitive screening design to determine the most significant effects of input parameters on output responses. 2. Using an efficient in-house LBM solver to generate simulation data for training polynomial regression equations à la response surface methodology via a forward stepwise regression approach utilising the corrected Akaike's Information Criterion. 3. Validating regression equations against additional simulations. 4. Optimising a composite desirability function using the reduced gradient algorithm of Minitab® statistical software using the regression equations as surrogate models.

This framework is used to solve the multi-objective optimisation problem of minimising the impeller power number, percentage of dead volumes, and maximum shear rate within a mixing tank stirred by a helical ribbon impeller containing a non-Newtonian, shear-thinning, power-law liquid, for a range of flow behaviour indices, by altering impeller and tank geometry. Surrogate models are found to replicate simulation results with a deviation of 3 % to 18 %. This framework is also tested against an empirical correlation-based method, wherein the framework demonstrates greater flexibility. The power number predicted by the correlation method deviates by ≤ 24 % compared to simulations, while the surrogate model deviates by ≤ 8 %.

Teaching