Multi-scale Exploration of MultiPhase Physics In FlowS (MEMPHIS)
This £5M EPSRC-funded project (www.memphis-multiphase.org), led by Omar Matar,is a collaboration between Imperial, Birmingham, Nottingham and UCL to create the next generation modelling tools for complex multiphase flows.
These flows are central to micro-fluidics, virtually every processing and manufacturing technology, oil-and-gas and nuclear applications, and biomedical applications such as lithotripsy and laser-surgery cavitation.
The ability to predict the behaviour of multiphase flows reliably will address a major challenge of tremendous economic, scientific, and societal benefit to the UK. The programme will achieve this goal by developing a single modelling framework that establishes, for the first time, a transparent linkage between input (models and/or data) and prediction; this will allow systematic error-source identification, and, therefore, directed, optimal, model-driven experimentation, to maximise prediction accuracy.
The framework will also feature optimal selection of massively-parallelisable numerical methods, capable of running efficiently on 105-106 core supercomputers, optimally-adaptive, three-dimensional resolution, and the most sophisticated multi-scale physical models. This framework will minimise the current reliance on correlations and empiricism resulting in a paradigm-shift in multiphase flow research worldwide.
- Dr Lyes Kahouadji (numerical simulation of multiphase flows using front-tracking and domain decomposition methods)
- Dr Zhizhao Che (ensemble methods, data assimilation methods, and experiments on droplet impact on falling films)
- Dr Zhihua Xie (Control volume finite-element multiphase flow simulations on unstructured meshes)
- Dr Ivan Zadrazil (experiments on stratified-stratifying flow, vertical upwards and downwards annular flows)
- Dr James Percival (numerical simulations of two-phase flow past a cylinder)
- Mr Idris Adebayo (experiments on droplet impact on flowing films)
- Mr Habib Abubakar (stability of bubbles rising in vertical tubes)
- Ms Renad Ismail (experiments on droplet impact on flowing films)
- Ms Yan Wang (numerical simulations of three-phase slug flows)
- Mr Thomas Ewers (analysis and simulation of downwards annular flows)
- Dr Huma Lateef
TMF: Transient Multiphase Flows
Since 1996 the Programme has undertaken research aimed at improving commercial computer programs' performance and supplying validation data.
Following on from the success of the original work in 1999 the sponsors requested several successive stages to the research programme, which continues to the present day (http://www.tmf-consortium.org).
The TMF Co-ordinated Projects have brought together leading researchers from 4 universities: Bristol University, Cranfield University, Imperial College London and Nottingham University.
This phase of co-ordinated research topics continues the very successful investigations to improve industry's ability to understand Transient Multiphase Flows. It is jointly funded by Industry and the UK's Engineering and Physical Sciences Research Council (EPSRC) and the UK's Department of Trade and Industry.
The industrial companies involved include many of those active in the exploration and development of oil and gas reserves and organisations developing and marketing computer programs for transient multiphase flow analysis.
Manufacturing with light: surface wrinkling and frontal photopolymeriation
This project focuses on developing fast, inexpensive and scalable approaches for surface patterning and microfabrication. Specifically, we study the mechanism of wrinkling as a way to impress regular patterns on polymers.
Wrinkled surfaces can be applied for a wide range of applications including drag reduction, increased surface hydrophobicity, and photonics. We have successfully managed to obtain sinusoidal wrinkles with wavelength of 100 nm on plasma oxidised polydimethylsiloxane. Regarding microfabrication, we investigate the manifacturing of microfluidic devices using photolithographic techniques.
Moreover, we also working on understanding how photopolymerisation can be coupled with wrinkling in order to rapidly manufacture three-dimensional structures using light. Photopolymerisation is a versatile solidification process that occurs when a photosensitive monomer-rich bath is exposed to light. Wrinkling is a mechanical instability that can occur when a heterogeneous elastic body is subject to lateral compression. During photopolymerisation, large compressive stresses can be generated in a thin gel layer that forms between the solid and liquid phases. We believe these stresses are a result of monomer diffusing into the gel layer from the liquid, thereby causing the gel to swell. The aspect of this project involves developing and simulating a mathematical model that captures the growth, composition, and mechanics of the solid/gel system.
- Dr Matt Hennessy
- Ms Manuela Nania
Funded by EPSRC (MH) and Departmental Scholarship (MN)
Numerical simulation of crude oil fouling in heat exchanger units
The general subject of this project is computation of fluid flow and heat- and mass- transfer using finite volume computations. The interests cover a wide variety of topics including single and multi-phase flows, compressible liquid, rheology/fluid mechanics of non- Newtonian fluids, turbulence modelling using LES and RANS methods, turbulence/chemistry interaction, surfactant driven flow, heat transfer in wavy annular flow, instability of Taylor bubble, laser radiation and multi-scale modelling. This project aim is to contribute to the in-depth understanding of fundamentals of multiphase flows and the development of concepts that tackle industrial problems, in particularly crude oil fouling and air-water-oil separation in oil and gas industry, complex fluid motion in rotating machinery and reciprocating engines, thermosyphon heating in refrigerant system and surfactant driven jets in microfluidic device.
“Up-scaling Algorithms for Crude-Oil Fouling in Heat Exchanger Networks” involves the validation of fouling experiments performed on the High Pressure Oil Rig (HIPOR) through 2D and 3D CFD simulations and aims to develop algorithms that up-scale the results for the whole heat exchanger.
- Dr Junfeng Yang
- Ms Lydia Laghaditi
- Mr Misha Crastes
Funded by the Skolkovo Foundation
- Prof. Sandro Macchietto(IC, Chem. Eng. Dept.)
- Prof. Erich Muller (IC, Chem. Eng. Dept.)
The effect of surfactants on flooding phenomena in vertical counter-current gas-liquid flows
Flooding in counter-current gas-liquid systems is characterised by an onset of flow reversal of at least a part of the liquid phase from a counter-current to a co-current state adverse to gravity in annular flow regime. Although flooding has been investigated extensively both phenomenologically and experimentally, the criteria for the onset of flooding and related interfacial instabilities remain unclear. This phenomenon is of high industrial importance; it can be found in the oil-and-gas industry (e.g. raisers in gas-liquid oil wells) or during emergency events in nuclear reactor core.
In this project we focus on the investigation of the dynamics of the flooding phenomena in the presence of surface active chemicals, i.e. surfactants. The phenomena is studied on an in-house build flow facility (see Figure 1) consisting of 4 m long, 32,4 mm nominal bore poly(methyl methacrylate) test section. The experimental techniques utilised include high-speed shadowgraphy, axial imaging as well as advanced non-intrusive techniques for the measurements of gas-liquid topology and velocity fields, namely Planar Laser Induced Fluorescence and Particle Image Velocimetry.
The effect of the addition of surface active agents on the structure of counter-current gas-liquid annular flow can be seen in Figure 2. The presence of surfactants leads to a significant increase in the level of the gas-liquid interfacial complexity, which arises from the associated reduction in interfacial tension (the so-called Marangoni interfacial stresses).
Figure 1: A schematic illustration of the flow facility used for the investigation of counter-current gas-liquid annular flow.
Figure 2: Side-view photographs of counter-current annular flow in (a) water and (b) in the presence of surface active agents.
- Dr Ivan Zadrazil
Funded by Shell (part of the Transient Multiphase Flows Consortium www.tmf-consortium.org)
Thin film flows over spinning discs with chemical reactions
The flow of thin liquid films subjected to centrifugal forces is accompanied by the formation of large-amplitude waves which gives rise to intense mixing on the surface of the disk and considerable increase in heat and mass transfer, commonly referred as “process intensification”.
Therefore, such flows have wide industrial applications, ranging from liquid atomisation to manufacturing of pharmaceuticals and fine chemicals. We develop a mathematical model studying the hydrodynamics, mass and heat transfer and chemical reaction with regards to a thin liquid film flowing over a spinning disk.
We apply the thin-layer approximation in conjunction with the Karman–Polhausen method to derive axisymmetric and non-axisymmetric evolution equations for the film thickness and the volumetric flow rates in both the radial and azimuthal directions satisfying boundary conditions and equation of state.
- Mr Kun Zhao
Multi-scale analysis of flows through nanochannels
Nanoporous materials have been the subject of extensive research in recent years. This includes carbon nanotubes (CNTs) or porous graphene sheets used for water desalination or gas separation. Finding more efficient and cheaper methods of separating solutes from a solution can make these separation mechanisms more economically viable and accessible which makes it an important area of research.
Typically, fluid flow is described by the solutions to the Navier-Stokes equations, such as Hagen-Poiseuille flow through cylindrical pipes. However, these first of all do not take interactions between the fluid and the pore wall into account. Secondly, the NS equations are generally solved for incompressible Newtonian fluids which follow a linear stress-strain relation.
We aim to find a description of fluid flow through nanochannels based on the continuum approach which nevertheless incorporates small-scale effects and can be generalised for different types of fluids. The following step is to look into solute rejection mechanisms, which factors contribute to them on a molecular level and finding a coherent description of such processes.
- Ms. Frederika Jaeger
Funded by EPSRC (CDT in Theory and Simulation of Materials)
Stability of rising bubbles in vertical pipes
Different distribution of phases, termed flow regimes, are encountered in gas-liquid flows through vertical pipes. One of these regimes is the ‘slug flow’. It occurs over a wide range of flow conditions and is characterised by the presence of bullet-shaped bubbles, known as Taylor bubbles, with diameter of about the same size as the pipe. For flows in sufficiently large diameter pipes, this flow regime is not observed. This project seeks to numerically simulate gas-liquid flows in vertical ‘large-diameter’ pipes. A particular objective would be the investigation of the stability of Taylor bubbles under the flow condition encountered in these pipes. From this, we hope to be able to offer qualitative and quantitative explanations as to the nonexistence of slug flow regime in gas-liquid flows in large diameter risers.
- Mr Habib Abubakar
Funded by the Nigerian Government
Jet-mixing of liquid-liquid flows in horizontal pipes
This project aims researching the interaction of transverse jets with horizontal stratified pipeline flows. Monitoring the jet breakup and the evolution of the resulting dispersions through using PLIF, PIV, PTV laser based techniques. Experimental data is compared with CFD models to study the ability of the models to capture the important details in these complex flows.
- Mr Stuart wright
Modelling and simulation of downwards annular flows
Numerical models will be developed specifically for downwards-annular flows, which are flows that involve a film falling under the action of gravity on the inside of a tube, whilst simultaneously being forced by a turbulent, co-flowing gas phase. These models will be capable of simulating the complex interfacial flows associated with falling film reactors, and will account for the effects of turbulence in the co-flowing gas core, which will enhance the formation of nonlinear waves (already present naturally in falling films). The development of these models will then facilitate the inclusion of non-isothermal effects arising from the presence of chemical reactions, and the increase in the film viscosity associated with these reactions. Solution of the equations underlying these models will complement the front-tracking methods, and massive-parallelisation capabilities that will be developed as part of the MEMPHIS programme. The deliverables will include models that can delineate the transitions to the various flow regimes in downwards-annular flow, which have been observed experimentally that have a profound influence on the rates of heat and mass transfer. These deliverables will also include models that can trace the concentration (of the various species), thermal, and viscosity profiles in space and time.
- Mr Thomas Ewers (Funded by Departmental Scholarship)
- Dr Hua Hu (Procter & Gamble)
Gel deformation induced by ultrasound-driven bubble oscillations
Acoustically driven oscillating microbubbles are used in biomedical applications, for instance as contrast agent for ultrasound imaging, or to promote uptake of therapeutic drugs by cells. However, the control of the stresses generated by oscillating microbubbles on cells and tissues is still lacking. We use high-speed video microscopy to observe the dynamics of oscillating bubbles near tissue phantoms.
The mechanical properties of the phantoms can be tuned to mimic different biological tissues. Tissue phantoms with different geometries can be made in order to study the effect of confinement on the bubble dynamics. The deformation is measured by tracking the displacement of tracer particles embedded in the phantom. Accurate measurements of the bubble dynamics and of the deformation of the phantom will help develop models to predict the interaction between a bubble and a soft interface, and ultimately help develop protocols for ultrasonic drug delivery.
- Dr Marc Tinguely
Funded by the Swiss National Foundation (MT)
Scale-up of graphene production
Liquid-liquid flows in horizontal and near-horizontal pipes
Mixtures of oil and water are commonly encountered in oil transportation pipelines, especially for subsea developments. The extracted fluids from a well are processed in offshore platforms which have a limited processing capability. This limitation means that fluids are not fully separated and a residual water content is typically carried through export oil pipelines. For high oil flow rates, water is transported along the pipe as droplets in the oil phase. As the oil flow rate decreases, the water droplets coalesce to form slugs in the oil phase in which the water volume is unknown. This behaviour has not been widely studied for horizontal and slightly inclined pipes having small water fractions (i.e. <5%).
Experimental and theoretical investigations are being carried out to study the transition from stratified to non-stratified flow in oil-water mixtures at pipe inclinations between ±5°. The work is focused on the determination of the in-situ water fraction as a function of phase velocity, pipe inclination and inlet configurations. A number of measurement techniques are being used to characterise the flow, namely, Planar Laser Induced Fluorescence (PLIF), Particle Image Velocimetry (PIV) and pressure drop measurements. These techniques provide detailed information of the flow such as phase distribution, velocity vectors, velocity profiles, flow patterns, and turbulent measurements. This information will be used to develop a mechanistic model to predict the in-situ water fraction and the transition between stratified to non-stratified flows.
- Mr Roberto Ibarra-Hernandes
Funded by BP (part of the Transient Multiphase Flows Consortium www.tmf-consortium.org)
- Dr Christos Markides (IC, Chem. Eng. Dept.)
Surfactant-assisted superspreading of aqueous sessile drops on hydrophobic substrates: MD simulations and molecular design
The intriguing ability of certain surfactant molecules to drive the superspreading of liquids to complete wetting on hydrophobic substrates is central to numerous applications that range from coating flow technology to enhanced oil recovery. Despite significant experimental efforts, the precise mechanisms underlying superspreading had remained unknown. We use molecular dynamics simulations of coarse-grained models based on the SAFT force-field to simulate surfactant-laden droplets containing surfactants of different molecular architecture on substrate of different affinity. Our simulations have identified two key conditions for superspreading, which must be simultaneously satisfied: the adsorption of surfactants from the liquid–vapor surface onto the three-phase contact line augmented by local bilayer formation. Crucially, this must be coordinated with the rapid replenishment of liquid–vapor and solid–liquid interfaces with surfactants from the interior of the droplet. Furthermore, we explore differences between superspreading and conventional surfactants, paving the way for the design of molecular architectures tailored specifically for applications that rely on the control of wetting. This work is performed in collaboration with the Department of Mathematics (Prof. Richard Craster) and Loughborough University (Prof. Victor Starov and Dr. Nina Kovalchuk), where, also, pertinent experiments validating our simulation observations are carried out. The work is financed by an EPSRC grant (EP/J010502/1)
- Dr Panos Theodorakis
- Dr Ed Smith
- Dr Nina Kovalchuk
Funded by EPSRC
Development of wax-inhibitors
Wax deposition is a phenomenon that plagues crude oilfields with significant economic impact that causes financial losses through the cost of prevention and remediation, reduced or deferred production, pipeline replacements and/or abandonments and equipment failures. A number of wax control technologies are currently being applied in the oilfield which includes mechanical methods for wax removal and thermal management strategies which primarily focuses on remediation rather than prevention. Chemical injection technology is a preventive, cost-effective alternative to combat wax deposition. The wax control chemicals currently in the market however are found not to be effective in preventing wax crystallization especially for crudes with high WAT and pour point temperatures. Therefore, there is a need to develop fundamental understanding of the wax crystallization mechanism at the molecular level and to develop a good and effective wax control chemical to suppress wax crystallization to the lowest possible temperature.
The aim of this study is to develop a framework of approaching the wax control chemical development in a systematic way. This will be done through a three-pronged approach: computer-aided molecular design; chemical synthesis and testing; embedding the molecular-scale chemistry into a continuum-scale model for simulations at the macro-scale. Coarse-grained molecular dynamics (MD) and the computational fluid dynamics (CFD) continuum-level simulation will be used to guide the synthesis of a new chemical, which will be tested against chemical systems from the oilfields. The MD-synthesis-testing steps will be iterative, culminating in the development of an effective wax inhibitor. Information from the MD step will be passed to the continuum-level modeling step for the development of simulation tools of wax formation/inhibition in flow processes, which is my main focus area.
- Ms Sara Shahruddin
- Mr Zaid Zolkiffly
Funded by Petronas
Crack propagation in the drying of blood drops
This Ph.D. project will involve the development of accurate, reliable, and efficient models for the direct simulation of the pattern formation associated with the drying of blood droplets (see Figure 1) whose complex stain morphologies are influenced by the original blood composition. The models will be formulated from the three-dimensional (3D) equations of mass, momentum and energy conservation, complemented by a 3D advective-diffusion equations for the concentrations of all species present within the drop. To simulate naturally crack-formation during drop drying, we will use a first-principles approach and couple the 3D fluid mechanical, and particle transport problem, to a mean-field, 3D solid mechanics problem by taking the densely-packed particles to be deformable Hertzian spheres; the latter are compressed by capillary pressure gradients due to non-uniform evaporation. These problems will, in turn, be coupled to Darcy flow in the interstices of the porous solid. A steady diffusion equation will be solved for the quasi-static vapour concentration to furnish the evaporative flux, and the energy equation will also be solved in the solid wall underlying the drop.
A hierarchy of models will be generated reflecting the increasing level of complexity. The models will account for all the relevant physical processes: evaporation, capillarity, Marangoni flow, diffusion, solidification/crystallisation, adsorption and changes of substrate wettability, contact line motion, sol-gel transition, and crack-formation. Thus, we will develop a simulation capability for the deposition morphology as a function of composition changes of the drying blood drops. This research will have application in the development of devices for rapid medical diagnosis.
This work will be carried out in collaboration with the group of Professor Khellil Sefiane (University of Edinburgh) who has world-leading expertise in evaporating, particle-laden drops, and with the group of Professor Rhodri Williams (University of Swansea) who has expertise in haemorheology.
- Arandeep Uppal (Funded by EPSRC through the Centre for Doctoral Training in Fluid Dynamics across Scales)
- Prof. Richard Craster (IC, Chem. Eng. Dept.)
- Prof. Khellil Sefiane (University of Edinburgh)
- Prof. Rhodri Williams (University of Swansea)