Autumn 2017 Seminars and Abstracts Autumn 2017 Seminars and Abstracts

Title: A stable and dual consistent finite difference method

Date: Monday 2nd October, 2pm, (SCI 0.31)
Speaker: Dr Sofia Eriksson (TU Darmstadt)
Abstract: We study the numerical solutions of time-dependent systems of partial differential equations, focusing on the implementation of boundary conditions. The numerical method considered is a finite difference scheme constructed by high order summation by parts operators, combined with a boundary procedure using penalties (SBP-SAT). Recently it was shown that SBP-SAT finite difference methods can yield super-convergent functional output if the boundary conditions are imposed such that the discretisation is dual consistent. We generalise these results so that they include a broader range of boundary conditions and penalty parameters. The results are also generalised to hold for narrow-stencil second derivative operators. The derivations are supported by numerical experiments.

Title: Chemical front propagation in periodic flows: the role of large deviations

Date: Monday 9th October, 2pm, (SCI 0.31)
Speaker: Dr Alexandra Tzella (University of Birmingham)
Abstract: We discuss the propagation of chemical fronts arising in Fisher-Kolmogorov-Petrovskii-Piskunov (FKPP) type models in the presence of a steady cellular flow. In the long-time limit, a pulsating front is established. Its speed, on which we focus, can be obtained by solving an eigenvalue problem closely related to large-deviation theory. We employ asymptotic methods to solve this eigenvalue problem in the limit of small molecular diffusivity (large Peclet number, Pe) and arbitrary reaction rate (arbitrary Damkohler number, Da). We identify three regimes corresponding to the distinguished limits Da = O(1/Pe), Da = O(1/logPe) and Da = O(Pe) and, in each regime, obtain the front speed in terms of a different non-trivial function of the relevant combination of Pe and Da, determined by solving a (Pe-independent) one-dimensional problem. Our results are contrasted against front speed values obtained from the so-called G equation: a level-set approximation that is commonly used when Da = O(Pe).

(Joint work with J. Vanneste, Edinburgh).

Title: Acoustic-gravity waves, theory & applications

Date: Monday 16th October, 2pm (SCI 0.31)
Speaker: Dr Usama Kadri (Cardiff University)
Abstract: Acoustic–gravity waves (AGWs) are compression-type waves generated as a response to a sudden change in the water pressure, e.g. due to nonlinear interaction of surface waves, submarine earthquakes, landslides, falling meteorites and objects impacting the sea surface. AGWs can travel at near the speed of sound in water (ca. 1500 m/s), but can also penetrate through the sea-floor surface amplifying their speed, which turns them into excellent precursors. “Acoustic–gravity waves” is an emerging field that is rapidly gaining popularity among the scientific community, as it finds broad utility in physical oceanography, marine biology, geophysics, water engineering, and quantum analogues. This talk is an overview on AGWs, with emphasis on the propagation under elastic ice sheets.

Title: Modulation of multiphase wavetrains and nonlinear reductions

Date: Monday 23rd October, 2pm, (SCI 0.31)
Speaker: Dr Daniel Ratliff (Loughborough University)
Abstract: The modulation of single phase wavetrains (and more recently their generalisation, relative equilibria), instigated by Whitham, is a field that has been developed over the last 50 years. The procedure generates a set of dispersionless nonlinear PDEs that govern the local wavenumber and frequency of the wave. When these degenerate, it has been shown that dispersion emerges at such points leading to equations such as the Korteweg-de Vries (KdV) equation. Remarkably, such reductions possess coefficients that may be related to the conservation laws of the original system which can be calculated in advance. This property is known as ‘universal form’.

This talk concerns itself with taking these ideas and applying them to solutions that have more than one phase. The questions are now this – which nonlinear PDEs arise? Do these nonlinear reductions still emerge with universal form? It will be shown that yes, these properties generalise quite nicely to the multiphase problem and recover many of the same equations derived from single phased solutions (like the KdV). Unsurprisingly, the increase in the number of system parameters allows one to derive further nonlinear PDEs (and even some new ones).

The talk concludes (hopefully, time permitting) by discussing two applications of the theory. The first is a stratified shallow water system and the second is a set of coupled Nonlinear Schrodinger equations (which model ocean wave envelopes, Bose-Einstein condensates and electromagnetic waves), showing the possible reductions and how the conditions for each equation can be met.

This work is in collaboration with Tom Bridges (University of Surrey).

Title: How to make a splash: from high speed droplet impact to a novel methodology for calculating water catch on aircraft surfaces

Date: Monday 6th November, 2pm, (SCI 0.31)
Speaker: Dr Radu Cimpeanu (University of Oxford)
Abstract: A new methodology for the calculation of water collection efficiency on aircraft surfaces is discussed. The approach incorporates the detailed fluid dynamical processes often ignored in this setting, such as the drop interaction with the surrounding air flow, drop deformation, rupture and coalescence, as well as the motion of the ejected microdrops in the computational domain. Direct numerical simulations using the volume-of-fluid technique are performed using modelling assumptions which enable us to take advantage of the disparity of lengthscales in the system. Comparisons are performed in the pre-impact regime with available experimental data, while the early stages of the impact are validated using the analytical framework provided by Wagner theory, context in which recent developments are also presented. We then focus on quantifying useful information on the liquid movement on longer timescales. The analysis shows a high degree of flexibility and can be used efficiently when considering changes in geometry (aircraft design), flow conditions and cloud characteristics. The interaction with our industrial partners will also be a point of focus, in particular in the context of developing a framework that incorporates the above analysis in an industrial work pipeline with no additional computational cost, thus making direct use of several hundreds of thousands of hours of CPU time on local supercomputing facilities. The methodology is finally applied to representative test geometries in collaboration with our partners.

Title: Multilayer networks: a new framework for complex systems

Date: Monday 20th November, 2pm, (SCI 0.31)
Speaker: Dr Ginestra Bianconi (Queen Mary University London)
Abstract: Multilayer networks describe interacting complex systems formed by different interacting networks. Multilayer networks are ubiquitous and include social networks, financial markets, multimodal transportation systems, infrastructures, the brain and the cell. Multilayer networks cannot be reduced to a large single network. In this talk I will present first recent results showing how we can extract from multilayer networks more relevant information than from its single layer taken in isolation. Secondly I will provide evidence that dynamical processes on multilayer networks can display very novel properties that reflect the rich interplay between structure and multiplexity.

Title: When is the effective wavefield a useful tool to predict wave attenuation over long distances?

Date: Monday 27th November, 2pm, (SCI 0.31)
Speaker: Dr Luke Bennetts (University of Adelaide)
Abstract: Anderson localisation theory tells us that waves attenuate through disordered media, and this talk is motivated by seeking efficient ways to calculate the attenuation rate as a function of the incident wave properties (frequency) and properties of the given medium, including the statistical properties of the disorder. Effective media theory is an appealing way to approach the problem, as it provides analytical insight, circumventing the need to repeatedly compute individual wave fields for different realisations of the disorder, as well providing the opportunity for elegant mathematical analysis. I will discuss the usefulness of effective media theory in the setting of two canonical, linear 1D problems, along the way outlining some associated effective media methods. The findings are in some respects entirely intuitive, but in other respects surprising.

Title: Feedback control of thin film flows

Date: Monday 4th December, 2pm, (SCI 0.31)
Speaker: Dr Susana Gomes (Imperial College London)
Abstract: The flow of a thin film down an inclined plane is unstable when the Reynolds number is larger than a critical value depending on the slope angle. These flows are important for many industrial applications, including coating and heat transfer. While some applications benefit from a flat film, in many cases one wishes to explore the flow’s instabilities and drive the system towards a non-uniform state.

In this talk, I will present a control methodology based on same-fluid blowing and suction at the wall. Given a desired interface shape, we apply controls which are proportional to the deviation between the current state of the system and the chosen solution. We study the effect of these controls on three partial differential equations which model the interfaces of thin film flows in different limits: two long-wave models (the Benney equation and a first-order weighted residuals model) and in the weakly nonlinear regime (the Kuramoto-Sivashinsky (KS) equation). We show that for the KS equation we can use a finite number of point-actuated controls based on observations of the full interface to stabilise both the flat solution and chosen nontrivial solutions, and investigate the robustness of the designed controls to uncertain observations and/or parameter values. Furthermore, we show robustness of the controls between the more general models and to limited observations.

Title: Mathematical models for the shape of the Eiffel tower: historical perspective and new results

Date: Wednesday 6th December, 3pm, (Queens 2.21)
Speaker: Prof Patrick Weidman (University of Colorado, Boulder)
Abstract: Equations modeling the shape of the Eiffel Tower are investigated. One model, based on equilibrium of moments, gives the wrong tower curvature. A second model, based on constancy of vertical axial stress, does provide a fair approximation to the tower's skyline profile of twenty-nine contiguous panels. However, neither model can be traced back to Eiffel's writings. Reported here is a new model embodying Eiffel's concern for wind loads on the tower, as documented in his communication to the French Civil Engineering Society on March 30, 1885. The result is a nonlinear, integro-differential equation which may be solved to yield an exponential profile. An analysis of actual panel coordinates reveals a profile closely approximated by two piecewise continuous exponentials with different growth rates. This is explained by specific safety factors for wind loading that Eiffel & Company incorporated in the design and construction of the free-standing tower.

Title: Drowning by numbers: mathematical design & implications of Wetropolis' flood demonstrator

Date: Monday 11th December, 2pm, (SCI 0.31)
Speaker: Prof Onno Bokhove (University of Leeds)
Abstract: The Wetropolis Flood Demonstrator will be introduced and analysed. Wetropolis commenced as outreach model for the public to let them experience rainfall events causing river flooding. Wetropolis is a table-top model with a conceptual river, flood plain, city, porous moor representing the upper catchment and groundwater flow, and an upland reservoir. Key is the rainfall, in terms of rain amount per Wetopolis day (a day is WD=10s), it rains either 1s, 2s, 4s or an extreme 9s in a WD, and rainfall location, either in the moor, in both reservoir and moor, in the reservoir or not. These 4x4=16 rain amount times rain location combinations are visually drawn daily (so every 10s) from two skewed Galton boards (as two steel balls fall down), with the most extreme rainfall event: 90% rainfall in both moor and reservoir with a probability of 7/256~3%, causing floods in the city every 5 minutes on average, by mathematical design.

Although it started as outreach tool, Wetropolis has also triggered the thinking about flood mitigation amongst flood practitioners. Wetropolis was inspired by the extreme 2015 Boxing Day Aire River floods in and around Leeds. Straightforward and more advanced analysis of extreme floods peaks of several rivers in Yorkshire shows that it is useful to introduce the excess flood volume, the volume of water above a certain river level threshold that caused the flooding. Given these excess volumes, I will show which flood mitigation measures are expected to be useful.

Flood storage via controlled and enhanced flooding of certain sections of flood plains and reservoir storage seems hitherto the only mechanisms to create the required volumes or "space for water". Natural flood management solutions, whilst seemingly appealing, generally contribute (far too) little to flood mitigation, as simple estimates can illustrate. To wit, the Boxing Day floods in Leeds, taking a flooding threshold of 3.9m at the Armley river level gauge, lea to a flood excess volume/lake of 2.1k by 2.1km and 2m depth. When one is able to partition this lake in parts on upstream flood plains, then the flood damage can possibly be minimised or prevented.

For further details about the seminars, or to join our mailing list, please contact Anna Kalogirou. For details of previous talks, please use the menu links on the left.