Programme
Programme
Ten lectures across two days, paired with hands-on Cirrus sessions.
Day 1 builds from NEMD fundamentals to solid-liquid interfaces.
Day 2 moves to nanoscale hydrodynamics, liquid-vapour interfaces, and machine-learning potentials.
Solid-Liquid Interfaces
Fundamentals of NEMD for Fluids and Interfaces
This talk will introduce the main theoretical ideas underlying the design and interpretation of nonequilibrium molecular dynamics simulations of fluids subjected to boundary driven shear. It will consider why and when we should choose this technique, how we generate and control the flow, how to compute the properties of interest and what results we should expect. Both wide channel and narrow channel simulations will be considered. The physical modelling of effects that become strongly apparent at the nanoscale, such as slip, interfacial thermal resistance, spin coupling and density coupling will be briefly introduced in anticipation of more detailed treatments that will be given by the other speakers.
His research addresses the statistical mechanics and computer simulation of complex fluids, spanning equilibrium and non-equilibrium molecular dynamics, polymer and colloid physics, and molecular rheology.
Interfacial Properties of Solid-Liquid Systems
This lecture will introduce the structural and thermodynamic properties of solid-liquid interfaces, with a particular focus on interfacial free energy. It will build physical intuition for how a liquid is organised against a solid surface, from fluid layering and substrate-induced ordering through to the thermodynamic cost of forming an interface. It will then set out a working framework for relating microscopic interfacial structure to interfacial free energy and surface stress, highlighting the distinction between these quantities through the Shuttleworth relation. Finally, an overview of computational methods for determining solid-liquid interfacial free energies and stresses will be presented, together with common challenges and pitfalls in their calculation and interpretation.
Expert in multiscale modelling and the coupling of continuum and discrete methods, with particular emphasis on molecular simulations of solid–liquid interfaces. Specialized in surface interactions, thermodynamics, and adsorption phenomena.
Slip Lengths & Flows Past Surfaces
The purpose of this lecture is to illustrate how molecular dynamics simulations help characterize nanoscale hydrodynamic flows near solid-liquid interfaces, particularly slip length and osmotic coefficients. First, I will introduce the phenomenology of slip length and osmotic response. I will then discuss the various methods employed to extract interface transport properties from molecular dynamics simulations, including non-equilibrium and Green-Kubo simulations. I will examine the impact of several factors such as work of adhesion, solid roughness, surface charge and liquid temperature, covering also supercooled liquid dynamics. Subsequently, osmotic phenomena will be tackled with an emphasis on thermoosmosis. Finally, I will briefly extend these considerations to the case of thermal slip at solid-liquid interfaces. This tutorial will draw in part on research results obtained in collaboration with L. Joly, Y. Yamaguchi and T. Omori.
He uses molecular dynamics to study nanoscale transport, including thermal transport, nanofluidics, hydrodynamic slip, and interfacial thermal conductance at solid-liquid interfaces.
Heat Flux and Interfacial Thermal Transport Across the Wall
The lecture will introduce molecular dynamics simulations of heat transport across solid–liquid interfaces. A brief theoretical background will be provided to give an understanding on the physical interpretation of thermal transport inside both bulk material and over interfaces. A focus will be given to how heat flux propagates under non-equilibrium condition, resulting in temperature discontinuities that emerge at solid-liquid interfaces, and how this is connected to interfacial thermal resistance. The lecture will further present commonly used analysis methodologies for steady-state simulations, including the construction and interpretation of temperature profiles, evaluation of heat flux and obtaining interfacial thermal resistance values. Know-how will be provided on selecting appropriate simulation parameters, validating the results, and how to avoid common pitfalls. Finally, some more advanced analysis techniques will also be briefly introduced.
He studies fluid interfaces using molecular dynamics, focusing on heat flux, interfacial thermal transport, and wetting at solid-liquid boundaries.
Hands-on session: Solid-Liquid Interfaces
In the afternoon, participants put the morning’s theory into practice on Cirrus, a UK Tier-2 national HPC service, with support from EPCC. From a single nonequilibrium simulation of a fluid confined in a nanoscale channel under shear and a temperature gradient, participants extract four quantities, each tied to a morning lecture: the velocity profile and driving method (Daivis), the interfacial density profile (Di Pasquale), the slip length at the wall (Merabia), and the interfacial thermal resistance across the wall (Surblys).
Led by Edward Smith, Saikat Datta and Rohit Pillai · on Cirrus, supported by EPCC
Nanoscale Hydrodynamics, Liquid-Vapour Interfaces & Machine Learning Potentials
Nanoscale Hydrodynamics
Opening the second day, this lecture will examine fluid flow at the nanoscale, continuing from Day 1’s confined solid-liquid channel. To build a physical intuition for when and how classical hydrodynamics is expected to hold and break down, we first explore thermal relaxations in equilibrium comparing predictions from classical hydrodynamics to simulation data. After this we investigate fluid flows where the system characteristic length scale approaches just a few nanometers. We will highlight what happens near interfaces and under strong spatial gradients. Next, we introduce the non-local (or generalised) viscosity coefficient, where the mechanical stress at a point depends on the entire system strain-rate distribution; we will discuss how the generalised approach can be applied to nanoscale flows and what research challenges remain. Finally, if time permits, guidance will be given on extracting hydrodynamic quantities from molecular dynamics and on the pitfalls of doing so.
His work combines molecular dynamics with continuum and nanoscale hydrodynamics to study non-equilibrium processes, viscoelastic fluids, and the breakdown of classical hydrodynamics at the nanoscale.
The Liquid-Vapour Interface: Surface Tension and Structure
This lecture will introduce the liquid-vapour interface as a thermodynamic and structural object. It will build physical intuition for surface tension as the free-energy cost of keeping the two phases apart, the microscopic origin of the pressure-tensor anisotropy at the interface, and the diffuse density transition between liquid and vapour. It will then present commonly used methods for measuring surface tension and the density structure across the interface from a molecular dynamics trajectory, together with practical advice on setting these measurements up and reading them correctly. These foundations are built on by the following lecture, which treats the finer interfacial structure of the liquid-vapour interface.
His research applies molecular dynamics to nanoscale multiphase flows, including cavitation, surface nanobubbles, and their role in water treatment.
Microscopic Structure of Liquid Interfaces: Capillary Waves and Intrinsic Profiles
This lecture will introduce the microscopic structure of liquid interfaces. It will build physical intuition for what the apparent density profile of a liquid-vapour interface actually shows once capillary-wave broadening is taken into account, separating the thermal roughening of the interface from its underlying molecular structure. It will then turn to the intrinsic profile and how it is reconstructed from a simulation trajectory, and show how the picture extends from a simple model fluid to water. Finally, it will look at how to resolve interfacial structure reliably, and the analysis choices that most often mislead.
He studies the microscopic structure and dynamics of liquids and their interfaces, and develops computational-geometry methods that identify intrinsic interfaces and separate capillary-wave broadening from the underlying molecular structure.
Machine Learning Potentials: Fundamentals
This lecture will introduce the fundamentals of machine-learned interatomic potentials. It will explain how these potentials learn the relationship between atomic configurations and energies from quantum-mechanical reference data, offering close to first-principles accuracy at a fraction of the computational cost. It will describe the main ingredients shared by modern approaches, from the choice of training data and the representation of local atomic environments to the fitting and validation of the model. Finally, it will consider their use in molecular dynamics and the practical questions of accuracy, transferability, and behaviour beyond the range of the training set.
He develops machine-learned interatomic potentials trained on quantum-mechanical data and studies their behaviour in molecular dynamics, with applications to liquid electrolytes and molecular materials.
Machine Learning Potentials for Fluids & Interfaces
Building on the fundamentals, this lecture will show the use of machine-learned potentials for NEMD systems with fluids and interfaces. The advantage over classical force fields will be discussed, focused on modelling solid-liquid interfaces, where capturing the relevant interactions accurately is often difficult. A few practical demonstrations will be given of running MD codes to show the application of these potentials in NEMD systems. An overview of the local forms of stress and heat flux will be derived, along with the extension to machine learning potentials.
His work develops multiscale methods that couple particle and continuum descriptions, including atomistic-continuum coupling, and applies machine learning to fluid dynamics and multiscale modelling.
The Landscape of MD Applications: Nanobubbles for Energy and Fuels
This lecture will use nanobubbles as a worked example of how molecular dynamics can take an atomistic picture of a fluid through to real-world engineering applications. It will show how these simulations capture the formation and stability of nanobubbles in both polar and nonpolar liquids, and how their presence alters the thermophysical and thermochemical behaviour of the host fluid. It will then connect these effects to applications of current interest in energy and decarbonisation, from the thermal performance of engineered working fluids to carbon dioxide hydrate formation and the combustion of fuels. The growing use of data-driven methods to analyse such simulations will also be touched on.
A UKRI Future Leaders Fellow, he works on low-carbon fuels, combustion engines, and powertrain technologies, combining experiments, engine simulation, and molecular dynamics to link fundamental research with the industrial decarbonisation of transport.
Hands-on session: Nanoscale Hydrodynamics & Liquid-Vapour Interfaces
The Day 2 practical again runs on Cirrus with support from EPCC, across two systems. Participants first study a liquid confined in a nanoscale slit pore under flow, examining where a local continuum (Newtonian) description of the viscosity begins to break down (Hansen). They then turn to a liquid slab bounded by two free surfaces and, from a single trajectory, measure the surface tension and the apparent density profile across the liquid-vapour interface (Dockar) and reconstruct the underlying intrinsic profile by separating capillary-wave broadening from the molecular structure (Sega).
Led by Edward Smith, Saikat Datta and Rohit Pillai · on Cirrus, supported by EPCC