Micro & Nano Flows for Engineering
The micro & nano flows group is a research partnership between the Universities of Warwick and Edinburgh, and Daresbury Laboratory. We investigate gas and liquid flows at the micro and nano scale (where conventional analysis and classical fluid dynamics cannot be applied) using a range of simulation techniques: molecular dynamics, extended hydrodynamics, stochastic modelling, and hybrid multiscaling. Our aim is to predict and understand these flows by developing methods that combine modelling accuracy with computational efficiency.
Targeted applications all depend on the behaviour of interfaces that divide phases, and include: radical cancer treatments that exploit nano-bubble cavitation; the cooling of high-power electronics through evaporative nano-menisci; nanowire membranes for separating oil and water, e.g. for oil spills; and smart nano-structured surfaces for drag reduction and anti-fouling, with applications to low-emissions aerospace, automotive and marine transport.
EPSRC Programme Grant in Nano-Engineered Flow Technologies
Our work is supported by a number of funding sources (see below), including a 5-year EPSRC Programme Grant (2016-2020). This Programme aims to underpin future UK innovation in nano-structured and smart interfaces by delivering a simulation-for-design capability for nano-engineered flow technologies, as well as a better scientific understanding of the critical interfacial fluid dynamics.
We will produce software that a) resolves interfaces down to the molecular scale, and b) spans the scales relevant to the engineering application. As accurate molecular/particle methods are computationally unfeasible at engineering scales, and efficient but conventional fluids models do not capture the important molecular physics, this is a formidable multiscale problem in both time and space. The software we develop will have embedded intelligence that decides dynamically on the correct simulation tools needed at each interface location, for every phase combination, and matches these tools to appropriate computational platforms for maximum efficiency.
This work is strongly supported by nine external partners (see below).
- “Nano-Engineered Flow Technologies: Simulation for Design across Scale and Phase” EPSRC Programme Grant EP/N016602/1 01/16-12/20 (£3.4M)
- “The First Open-Source Software for Non-Continuum Flows in Engineering” EPSRC grants: EP/K038427/1 K038621/1 K038664/1 07/13-06/17 (£0.9M)
- “Multiscale Simulation of Interfacial Dynamics for Breakthrough Nano/Micro-Flow Engineering Applications” ARCHER Leadership Project 11/15-10/17 (£60k in supercomputer computational resource)
- “Skating on Thin Nanofilms: How Liquid Drops Impact Solids” Leverhulme Research Project Grant 08/16-08/19 (£146k funding a 3-year PDRA)
- Airbus Group Ltd
- Bell Labs
- European Space Agency
- Jaguar Land Rover
- Oxford Biomedical Engineering (BUBBL)
- TotalSim Ltd
- Waters Corporation
Latest news and blogs
Prof. Duncan Lockerby, University of Warwick
It's a pleasure to welcome three new starters to Warwick and the Micro Nano Flows team. Laura Cooper joins us as a post-doctoral researcher, and will be working on multi-phase flow in porous media with James Sprittles. Yixin Zhang and Jacqueline Misfud are in the first weeks of their PhDs. Yixin will be investigating nano film stability (with molecular dynamics) and Jacqui will be researching micro-bubble cavitation (with CFD) in partnership with Waters Limited. We wish all three the best of luck in their research!
Dr James Sprittles, University of Warwick
Dr Shiwani Singh has joined the Micro and Nano Flows team at Warwick, where she will be based in the Maths Institute. She will be looking into multiscale modelling of viscoelastic flows over the next couple of years.
Prof. David R Emerson, Daresbury Laboratory
Arnau Miro from the Universitat Politècnica de Catalunya won a HPC Europa-2 grant for a 13-week visit to the Daresbury group. Arnau will be working on advanced meshing and code coupling strategies and starts his visit mid-January.
The Mpemba Effect is a counter-intuitive phenomenon whereby, under certain circumstances, warm water can freeze faster than cold water. Although this phenomenon was known for a long time by many philosophers and scientist, including Aristotle, Rene' Descartes and Francis Bacon, it was brought to the attention of the scientific community by a Tanzanian schoolboy, Erasto Mpemba. In 1963, during a school project, he noticed that an ice-cream mixture, previously heated, froze more rapidly than one that was cold. He was so intrigued by this phenomenon that he continued experiments until, a few years later, he went on to work with a physics professor, Denis Osborne, and together they published a paper in the journal Physics Education in 1969.
A public competition announced by the Royal Society of Chemistry in 2012 to explain the Mpemba effect renewed the interest in this phenomenon.
The above video provides a very clear overview of the various explanations that have been put forward. These include evaporation (the evaporation rate of warm water is higher and, consequently, there is a smaller amount of water to cool down), dissolved gases (the warm water contains less dissolved gas, which, apparently, hinders the ability of water to conduct heat), supercooling (the warm water contains more nucleation sites which makes it unlikely the need to get temperatures less than zero degrees Celsius for freezing), convection (the convection currents in the warm water are stronger and increase heat dissipation).
None of these explanations is entirely convincing. A recent study has even concluded that the Mpemba effect is just an experimental artifact due to measurements' inaccuracies. On the other hand, another recent study has shown, by theory and simulation, that the Mpemba effect and its reverse (a cooler sample may heat faster than a hotter sample) may occur in uniform granular fluids composed of inelastic particles and an explaination can be given based on the kurtosis of particle velocity distribution function.
It's surprising that something as apparently simple as the freezing of water is by contrast still poorly understood.
Dr Rohit Pillai, Research Associate, University of Edinburgh
A recurrent theme on this blog has been posts describing talks presented at conferences. In keeping with that tradition, I’ll use this post to briefly introduce two talks from a recent-ish conference that I had the privilege of attending. Both talks relied on the use of what are known as slippery-liquid-infused-porous (SLIP) surfaces, or lubricant-impregnated surfaces. According to the folks who make them, SLIP surfaces 'combine a lubricated film on a porous solid material to create low-cost surfaces that exhibit ultra-liquid repellency, self-healing, optical transparency, pressure stability, and self-cleaning'. While this sounds impressive and worthy of further unpacking, I'll leave the discussion on the science behind such specialised surfaces for a future blogpost, and keep this one focused on the fluid dyamics instead, i.e. the drops.
Making a SLIP surface. Image credit: Wyss institute (Original here)
The first talk, by Zuzana Brabcova, discussed her recent results published in these two papers (one, two). Basically, there exist two popular techniques to electrically manipulate drops on surfaces. The first is electrowetting, where the ions in a conducting liquid droplet are transported by the electric field, resulting in an electrophoretic interfacial force. The other one is dielectrowetting, where the electric field acts on dipoles (in an effectively dielectric liquid droplet) at the solid-liquid interface, resulting in a dielectric interfacial force. One of the problems when using either mechanism is the observed contact-line hysteresis, which prevents a smooth transition from wetting to dewetting. This talk demonstrated that, by using a SLIP surface combined with custom spiral electrodes, the hysteresis for electrowetting and dielectrowetting could be completely removed. By modifying the applied signal, near-axisymmetric wetting was also realised, resulting in the formation of a circular thin film on demand.
The second talk, by Jian Guan, looked at translating drops on SLIP surfaces. To induce controlled drop motion on a SLIP surface, a patterned surface (before the lubricant layer is applied) was used, resulting in an uneven lubricant distribution on the surface. For a V-shaped channel, droplets moved towards the regions of higher lubricant deposition to minimise surface energy; these regions were invariably near the edges of the channel. The size of the droplet determined its net motion. For small droplets, either side of the channel was the energetically-preferred destination, while for larger droplets which straddled both ends of the channel, motion into along the channel was observed. Thus the motion and final equilibrium position of the droplet could be determined in advance using patterned SLIP surfaces. These results are discussed in further detail in the author's recent paper on the topic (link).
Dr Anirudh Rana, Research Fellow, University of Warwick
Our recent research article has recently been accepted in Journal of Fluid Mechanics. The general interest of this study was to explore boundary value problems for moderately rarefied gas flows, with an emphasis on evaporation from nanostructures. In this article we developed macroscopic models based on R26-moment equations to describe the transport process near the phase boundary between a liquid and its rarefied vapour due to evaporation and condensation. For evaporation from a spherical droplet, analytic solutions were obtained to the linearized equations from the Navier Stokes & Fourier (NSF), regularised 13-moment (R13) and regularised 26-moment (R26) frameworks. Results are shown to approach computational solutions to the Boltzmann equation as the number of moments are increased, with good agreement for Knudsen number smaller than 1.
Knudsen layer functions in temperature-driven flow (far field temperature is twice of the temperature in liquid) are shown in Figure above. Curves of the normalised pressure (left) and the temperature defect (right) are plotted against scaled distance from the interface, for the NSF, R13 and R26 theories with complete evaporation and compared with the solutions of the linearized Boltzmann equation (denoted by symbols).
Our results indicate that the R26 equations with evaporation boundary conditions yield an excellent quantitative description in all cases, which are not matched by the NSF or R13 theories. The R26 system provides three exponential functions to form the Knudsen layer, thus capturing the more complex behaviour in Knudsen layer.