Creating macroscale effective interfaces encapsulating microstructural physics

Lead Research Organisation: Imperial College London
Department Name: Dept of Mathematics

Abstract

This proposal seeks funding for a comprehensive three year research
program into
methodologies for modeling microscale interfacial phenomena on the macroscale
level. A fundamental question stretching across many disciplines is:
Given a microstructured interface
can it be replaced by an effective ``averaged'' boundary condition
entirely posed upon a macroscale. If so, can it accurately reproduce the physical
effects created by the microstructure? Can this effective boundary condition be
derived rigorously, rather than in some ad-hoc fashion, and what are
the limitations in so doing?

The proposal aims to answer these questions, with the goal of being
able to accurately and efficiently predict complex physical behaviour
in three apparently unconnected
fields: in wave propagation for surface Rayleigh-Bloch
waves and for the reflection of waves from designer structured
surfaces, in the statistical mechanics of phase transitions on
micropatterned surfaces, and in modeling diffusions through
structured domains. These fields all share a complex structured
interface and the generic overarching Mathematical approach we propose will lead to effective
boundary conditions
encapsulating the dominant microscale Physics; this will represent a
considerable advance in each of these areas.

The primary Mathematical approach will be based around
Homogenization theory utilizing the discrepancy in lengthscales to
create asymptotics from multiple scales analysis. Homogenization is conventionally
used when the bulk material has short-scale fluctuations and the
solution varies on a long-scale, its use for
interfaces is much less well explored.
Importantly we also aim to enhance the
range of validity of homogenization theory away from long-wave,
quasi-static, regimes to ones that can vary on the same scale as
the microstructure. This analytical work will be complemented by
detailed
numerical simulations that will act to verify the efficacy of the
developed interfacial models. The work will be undertaken by a team
from the Mathematics Department at Imperial College London
with complementary skills and strengths: Pavliotis (Homogenization
theory, stochastic processes), Parry (Statistical mechanics, phase
transitions) and Craster (Wave propagation, homogenization, fluid mechanics).

Planned Impact

This proposal is unique as it bridges between the research programs of three well-established Investigators with complementary expertise with the aim to tackle a fundamental
scientific issue regarding structured surfaces. The immediate impact will be academic and felt across a spectrum of disciplines, such as applied mathematics, applied physics, statistical physics, optics, fluid dynamics and mathematical biology. The applications are truly vast as structured surfaces are all around us. The results and advances in modelling obtained through this proposal will impact these applications, for instance, the potential of the effective bulk modelling of waves upon biomedical sensors and the accurate modelling of thin fluid films, spreading and wetting over structured surfaces. Wetting phenomena have an enormously wide variety of industrial applications ranging from the macroscopic scale (oil recovery, drainage of water, cooling of reactors, deposition of pesticides) to the microscopic (microfluidics, ink jet printing). Furthermore, the development of efficient numerical and analytical techniques for the study of the narrow escape problem will be useful for researchers in cellular biology. Indeed, the techniques that methodologies that we will develop will lead to efficient analytical and computational techniques for calculating quantities such as the time needed for a reactive particle released from a specific organelle to activate a given protein on the cell membrane.

The new modelling to be undertaken will lead to effective boundary conditions that truly incorporate the microstructure. On one level this will lead to more rapid and simple computations and on another will allow the rapid design of smart structures or designer surfaces for real applications. This is in addition to the academic impact upon each subject and topic.

Publications

10 25 50
 
Description Analysis of stochastic systems with a multiscale structure.

Understanding of noise induced transitions for multiscale stochastic systems.

Use of these techniques to study problems in nonequilibrium statistical mechanics.

Development of improved Markov Chain Monte Carlo techniques and application to problems in statistical mechanics.

Development of new homogenization techniques for the study of complex media. Study of diffusion problems in complex media and composite materials.

Data-driven approaches to the derivation of coarse-grained models.

Control of thin film flows, falling liquid films. Control of spatiotemporal chaos.

Rigorous and systematic study of wetting transitions.

Rigorous study of time dependent density functional theory.
Exploitation Route Noise induced transitions for stochastic dynamical systems with multiple scales.

Applications of homogenization theory/multiscale techiques to metamaterials.

Time dependent density functional density theory.
Sectors Aerospace, Defence and Marine,Other

 
Description The analytical and computational techniques that have been developed are being used in order to develop algorithms for the calculation of transport coefficients and in order to optimize the performance of statistical algorithms such as Markov Chain Monte Carlo. The research supported by this grant also led to advances in the study of metamaterials and also in the physics of wetting phenomena.
First Year Of Impact 2014
Sector Other
 
Description EPSRC Platform Grant
Amount £1,616,110 (GBP)
Funding ID EP/L020564/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Public
Country United Kingdom
Start 06/2014 
End 06/2019
 
Description EPSRC Programme Grant
Amount £2,551,402 (GBP)
Funding ID EP/L024926/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Public
Country United Kingdom
Start 07/2014 
End 07/2019
 
Description EPSRC responsive mode
Amount £379,072 (GBP)
Funding ID EP/L025159/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Public
Country United Kingdom
Start 09/2014 
End 08/2017
 
Description Control theory for multiscale stochastic systems 
Organisation Free University of Berlin
Department Institute of Mathematics
Country Germany 
Sector Academic/University 
PI Contribution I brought my expertise on multiscale analysis for diffusion processes such as homogenization theory, averaging, and also statistical inference for multiscale diffusions
Collaborator Contribution Stochastic Optimal Control
Impact Optimal control of multiscale systems using reduced-order models (with W. Zhang, J.C. Latorre and C. Hartmann). J. Comp. Dyn. 1(2), pp 279-308, (2014). Numerical Methods for Computing Effective Transport Properties of Flashing Brownian Motors (with J.C. Latorre and P.R. Kramer). J. Comp. Phys., 257 Part A, pp. 57-82, (2014). Corrections to Einstein's relation for Brownian motion in a tilted periodic potential (with J.C. Latorre and P.R. Kramer). J. Stat. Phys. 150(4), 776-803 (2013).
Start Year 2012
 
Description Optimizing the performance of Markov Chain Monte Carlo algorithms 
Organisation University of Edinburgh
Department School of Mathematics
Country United Kingdom 
Sector Academic/University 
PI Contribution This collaborative research project is concerned with the optimization of the performance of Markov Chain Monte Carlo algorithms that used for sampling from probability distributions in high dimensional spaces. My contribution to this project was the necessary mathematical techniques for studying this problem, in particular the development of techniques for studying spectral problems for hypoelliptic operators
Collaborator Contribution They brought their expertise on molecular dynamics and on spectral theory for differential operators. They also made contributions on the numerical analysis of algorithms used in Molecular dynamics simulations.
Impact Variance Reduction using Nonreversible Langevin Samplers (with A. Duncan and T. Lelievre). J. Stat. Phys. (2016) doi:10.1007/s10955-016-1491-2 Langevin dynamics with space-time periodic nonequilibrium forcing (with R. Joubaud and G. Stoltz). J. Stat. Phys., to appear J. Stat. Phys., 158(1) pp 1-36 (2015). Optimal nonreversible linear drift for the convergence to equilibrium of a diffusion (with T. Lelievre and F. Nier). J. Stat. Phys. 152(2) 237-274 (2013).
Start Year 2012
 
Description Optimizing the performance of Markov Chain Monte Carlo algorithms 
Organisation École des ponts ParisTech
Country France 
Sector Academic/University 
PI Contribution This collaborative research project is concerned with the optimization of the performance of Markov Chain Monte Carlo algorithms that used for sampling from probability distributions in high dimensional spaces. My contribution to this project was the necessary mathematical techniques for studying this problem, in particular the development of techniques for studying spectral problems for hypoelliptic operators
Collaborator Contribution They brought their expertise on molecular dynamics and on spectral theory for differential operators. They also made contributions on the numerical analysis of algorithms used in Molecular dynamics simulations.
Impact Variance Reduction using Nonreversible Langevin Samplers (with A. Duncan and T. Lelievre). J. Stat. Phys. (2016) doi:10.1007/s10955-016-1491-2 Langevin dynamics with space-time periodic nonequilibrium forcing (with R. Joubaud and G. Stoltz). J. Stat. Phys., to appear J. Stat. Phys., 158(1) pp 1-36 (2015). Optimal nonreversible linear drift for the convergence to equilibrium of a diffusion (with T. Lelievre and F. Nier). J. Stat. Phys. 152(2) 237-274 (2013).
Start Year 2012
 
Title IN SITU CONTROL OF FLUID MENISCI 
Description A system includes a non-vertical channel containing a fluid forming a fluid meniscus having a capillary length and a contact angle ?. The channel in cross-section has a perimeter length |S |and an area |O|. The cross-section of the non-vertical channel is selected so as to define a constant Lagrange multiplier ?, where ?= |S |cos ? / |O|. A functional F [G*] ? |G*| - cos ? |S* | + (1/a2)G* +? |O*| is minimised to define a minimum value F0 =MinF. At a critical transition where F=0, the fluid defines a smooth arc of length [G*] that divides the cross-section of the channel into two parts. |O*| is the cross-sectional area of the fluid, which has a curve of length |S* | in contact with the channel, and G* represents a vertical position of the centre of mass of the fluid multiplied by the cross-sectional area |O*|. How far the fluid meniscus extends along the channel is controlled by one or more parameters of the functional F [G*]. 
IP Reference WO2015121629 
Protection Patent granted
Year Protection Granted 2015
Licensed No
Impact ISIS Project 9544 involves a method for controlling fluid menisci in micro- and nano-fluidic devices from mesoscopic to a nanoscopic scale. The method can be used to control how far the fluid meniscus extends along the channel without emptying, and can be used to force the draining of much smaller channels than previously possible. This meniscus deformation has several advantages that could find applications in a number of science and technology sectors: Increasing the liquid-gas or fluid-fluid interfacial area for chemical reaction, analysis or catalysis Forcing the draining of liquids and gases out of the capillary Measuring any of the system control parameters with high accuracy; contact angles, capillary length and surface tensions.