GCEC seminar - external speaker

Group: Glasgow Computational Engineering Centre
Speaker: Dr Luca Heltai, University of Pisa
Date: 26 April, 2024
Time: 13:00 - 14:00
Location: James Watt Building: 334 Studio

Title: The Reduced Lagrange multipliers approach: dimensionality reduction and coupling across heterogeneous dimensions in multi-physics problems

Abstract:

Many physical problems involving heterogeneous spatial scales, such as the flow-through fractured porous media, the study of fiber-reinforced materials, or the modeling of blood circulation in living tissues -- just to mention a few examples -- can be described as coupled partial differential equations defined in domains of heterogeneous dimensions that are embedded into each other. The definition and the approximation of coupling operators that are suitable for such problems remain challenging, both theoretically and computationally.

In this talk I will present the combination of two mathematical techniques that can be used to construct natural dimensionality reduction techniquesœ and coupling operators across heterogeneous dimensions. In particular, I will show how non-matching and immersed finite element methods can be used to enable the mesoscale resolution of problems with embedded structures, and how they provide a natural way to perform geometric dimensionality reduction. I will discuss how to define systematically geometrical reduction operators that, combined with non-matching techniques, and how this allows the formulation of coupled multi-physics and multi-dimensional problems in a mathematically consistent way. I will analyze the model error deriving from the dimensionality reduction, and provide some relevant numerical examples.

GCEC seminar - Riaz Akhtar

Group: Glasgow Computational Engineering Centre
Speaker: Riaz Akhtar, University of Liverpool
Date: 23 February, 2024
Time: 11:00 - 12:00
Location: Rankine 629

Title: The Aorta: Structure and Rupture through the lens of micromechanics

Abstract: The aorta is the most important conduit for blood flow in the human body. The extracellular matrix (ECM) provides a structural framework for normal, healthy function of the aorta. This ECM framework is composed of a 3D arrangement of elastin, collagen, proteoglycans which are synthesised during fetal development. With ageing and with aortic disease such as aneurysm formation, protease activity and degradation of the ECM increases.

Amongst the various aortic diseases, Thoracic aortic aneurysms (TAAs) are a devastating condition where the aorta can dilate significantly beyond its normal diameter and rupture. Often these go undetected and misdiagnosis can be fatal. Sometimes the cause of TAAs is unknown (non-syndromic) but the most common cause of TAAs are genetic conditions such as Bicuspid Aortic Valve (BAV) disease. This talk will focus on micromechanical and biochemical differences in patients with BAV as compared to non-syndromic patients. Implications of current surgical treatment and guidelines will be discussed.

Bio: Riaz Akhtar is a Reader in Biomedical Engineering at the University of Liverpool. His research largely focusses on structure-property relationships in ageing and diseased soft tissues at the nano- and micron- level. His work spans a number of biomaterials and tissues including hydrogels, cardiovascular tissues, ocular tissues and skin. The main thrust of Riaz’s research is on aortic disease biomechanics and he co-established the LABB Group (www.labb-group.com) and is a member of the Liverpool Centre of Cardiovascular Sciences. Riaz has published over 70 peer reviewed papers and has received several honours and awards for his research including the Bob Bonser Prize for Aortic Surgery (2019) and held a Royal Academy of Engineering/Leverhulme Trust fellowship in 2017-18.

GCEC Seminar: Modelling and Simulation of Bacterial Colony Formation: Challenges in Health and Computations

Group: Glasgow Computational Engineering Centre
Speaker: Prof Paul Steinmann, University of Glasgow and FAU Erlangen-Nurenberg
Date: 05 December, 2023
Time: 10:00 - 11:00
Location: JAMES WATT STH: 603 CREATIVITY

Title: Modelling and Simulation of Bacterial Colony Formation: Challenges in Health and Computations

GCEC seminar: external

Group: Glasgow Computational Engineering Centre
Speaker: Evzen Korec, Imperial College London
Date: 29 November, 2023
Time: 15:00 - 16:00
Location: Rankine 629

Zoom link: https://uofglasgow.zoom.us/j/84366078649?pwd=ckxnRFl3TVJ5b2ZRTjJrOWFUYi9Ldz09&from=addon

Title: A phase-field-based chemo-mechanical model for corrosion-induced cracking in reinforced concrete

 Abstract: We propose a new model for the corrosion-induced cracking of reinforced concrete [1]. This complex chemo-mechanical phenomenon is of a great practical importance because it results in the spalling/delamination of concrete cover, and corrosion-induced damage is responsible for 70-90 % of prematurely deteriorated reinforced concrete structures. The state-of-the-art knowledge of the underlying chemo-mechanical processes has been incorporated in three interconnected sub-models – (i) a reactive transport model for: (i.A) the transport of aggressive corrosion-activating species (such as chlorides) to the steel surface and (i.B) the transport and precipitation of iron ions released from the steel surface in the concrete porosity, (ii) model for the corrosion-induced pressure resulting from the concurrent constrained accumulation of: (ii.A) the rust layer in the vicinity of corroding steel rebar and (ii.B) the rust in the concrete pore space, (iii) a phase-field fracture model calibrated to accurately describe the quasi-brittle fracture of concrete. In addition, a damage-dependent diffusivity tensor has been employed to capture the enhanced transport of chlorides to the steel surface and the local reduction of the rate of accumulation of rust due to the enhanced transport of iron ions through cracks away from the steel surface. The proposed model was numerically solved with the finite element method in COMSOL Multiphysics software. Several two-dimensional and three-dimensional numerical case studies were investigated including impressed current tests carried out experimentally in laboratory conditions. The comparison of experimental and numerical results revealed that the model can simultaneously capture the steel mass loss and the evolution of surface crack width for different concrete cover depths and applied corrosion currents. Also, for the samples with multiple rebars, the model accurately predicted the spalling or delamination patterns depending on the rebar spacing. There are two critical findings. Firstly, that fracture can proceed even with the partial saturation of the concrete pore space with rust, which can last for years in natural conditions. Secondly, we propose a solution to a more than 25-year-old problem of why accelerated impressed current tests lead to slower propagation of cracks (with respect to corrosion penetration) than in natural conditions and thus underestimate sustained corrosion-induced damage. In view of recent experimental findings, we suggest that the phenomenon can be attributed to the variability of rust composition and density, specifically to the variable ratio of the mass fractions of iron oxide and iron hydroxide-oxide, which is affected by the magnitude of the applied corrosion current density.

[1] Korec, E., Jirásek, M., Wong, H. S., & Martínez-Pañeda, E. (2023). A phase-field chemo-mechanical model for corrosion-induced cracking in reinforced concrete. Construction and Building Materials, 393, 131964.

 Bio: Born and raised in Prague, Czech Republic, I joined the Czech Technical University in Prague (CTU), where I studied successively for a first-class bachelor (2018) and master’s degree (2020) in structural and transportation engineering. There I was privileged to join its Department of Mathematics (2015- 2016) and later the Department of Mechanics (2016 - 2020). There, under the supervision of Prof Milan Jirásek, I was analysing and developing models for the buckling of elastic beams and formulating the model for torsion of elastic beams with thin-walled closed cross-section. During my master’s degree, I had a privilege to join Prof Garth N. Wells and his group at the Cambridge Center for Micromechanics at the University of Cambridge as an academic visitor (2020). While in his group, I was working on suitable methods (especially Nitsche methods) for enforcing contact boundary conditions in large scale (specifically exascale) multi-physical problems. This research was conducted for ASiMoV Project, one of the largest high-performance computing focused research projects in the UK in recent times, worth $14.7M. Currently, I am a PhD student in the Mechanics and Infrastructure Materials Lab and President’s PhD Scholar at Imperial College London. My research is focused on the chemo-mechanical modelling of corrosion-induced fracture in reinforced concrete and is supervised by Prof Emilio Martínez-Pañeda (University of Oxford), Prof Milan Jirásek (Czech Technical University in Prague) and Prof Hong S. Wong (Imperial College London).

GCEC Seminar

Group: Glasgow Computational Engineering Centre
Speaker: Lukasz Kaczmarczyk, University of Glasgow
Date: 17 November, 2023
Time: 15:00 - 16:00
Location: Rankine 629

Title: An arbitrary order contact formulation using Lagrange Multipliers in Rivart-Thomas space

GCEC seminar

Group: Glasgow Computational Engineering Centre
Speaker: Dr. Sang-ri Yi, UC Berkeley and NHERI SimCenter
Date: 14 July, 2023
Time: 15:00 - 16:30
Location: Rankine 629

GCEC seminar

Group: Glasgow Computational Engineering Centre
Speaker: Prof Junho Song, Seoul National University, South Korea
Date: 14 July, 2023
Time: 15:00 - 16:30
Location: Rankine 629

GCEC Seminar: Hybrid Rayleigh waves along with nonlocal nonlinear metasurface

Group: Glasgow Computational Engineering Centre
Speaker: Santanu Manna, Indian Institute of Technology Indore
Date: 24 May, 2023
Time: 15:00 - 16:00
Location: Rankine Building, Room 630

Abstract: 

Seismic metasurface shows promising applications for the development of earthquake-resistant structures and improved seismic imaging and sensing capabilities. This work presents a new design for a metasurface consisting of an array of 2-DOF nonlinear spring-mass systems attached via a linear elastic spring to a nonlocal elastic substrate. Our research has integrated the concepts of nonlinearity in a double spring-mass system with the nonlocal elasticity effects in the substrate to investigate the dispersive properties of hybrid Rayleigh waves. This compact construction has resulted in providing multi-frequency attenuation via multi-frequency bandgaps. The presence of nonlinearity in the multi-resonators can be argued to be responsible for the existence of these multi-frequency bandgaps. Numerous plots are plotted in MATLAB to demonstrate the effects of nonlocal elasticity and the velocity ratio of the host substrate, nonlinearities, and relative amplitude inputs of the spring-mass system on the spectral bandgaps.

Speaker Bio:

Dr. Santanu Manna is an Assistant Professor at the Department of Mathematics, Indian Institute of Technology Indore. Prior to that, he spent around one year as Newton International Fellow by the Royal Society at the School of Computing & Applied Mathematics, Keele University, United Kingdom. He was a Post-Doctoral Research Scientist at Indian Institute of Science Education and Research Kolkata during 2015-2016. Dr. Manna received the Ph.D. degree in Applied Mathematics from Indian Institute of Technology (ISM) Dhanbad in 2015.

His research interests include Applied PDEs: Elastic wave propagation aspects, Geomechanics, Plate theory, & earthquake statistics. He has published more than 35 refereed journal papers of international repute (SCI journals). 

GCEC Seminar: The case for understanding soft-tissue mechanics to enable new wearable and implantable technologies

Group: Glasgow Computational Engineering Centre
Speaker: Michael Crichton, Heriot-Watt University
Date: 12 May, 2023
Time: 11:00 - 12:00
Location: James Watt South Building, Room 603

Title: The case for understanding soft-tissue mechanics to enable new wearable and implantable technologies

Abstract: In the past decade, skin and other soft tissues have received substantial research focus due in part to the surge in popularity of wearable and implantable technologies. Researchers developing sensor technologies often use the term ‘wearable’ for the purpose of indicating ‘pain free’, ‘non-invasive’, ‘personalised medicine’ and ‘AI-driven diagnostics’, but the number of devices on the market that can do substantially more than measure heart rate and movement is limited. Part of the reason for this is that the complexity of the soft tissue interfaces of tissues are often overlooked. In this presentation we will provide an overview of how an understanding of biological tissue mechanics is allowing us to develop new sensing modalities for health conditions including wound healing, the birthing process, incontinence and traumatic injury.

Bio: Michael Crichton is an Associate Professor in Biomedical Engineering at Heriot-Watt University where his research focusses on the interfaces between biology and engineered technologies. He is interested in how the mechanical interactions of our bodies can tell us about an individual’s health and their longer term progression through health states. Michael’s research takes an experimental  multi-disciplinary approach to the development of novel biomedical devices centred on clinical needs. He is passionate about working to understand how to derive impact for patients from technology advances, and is involved in a range of initiatives to help progress medical technologies. Michael has a PhD in microneedle vaccine delivery from the University of Queensland, and an undergraduate engineering degree from Glasgow University. He also has spent time in industry in a startup medical device company, and holds licenced patents in this area.

GCEC seminar

Group: Glasgow Computational Engineering Centre
Speaker: Dr Benjamin Cameron, University of Southampton
Date: 05 May, 2023
Time: 15:00 - 17:00
Location: https://uofglasgow.zoom.us/j/84206291369?pwd=VUxRWkpOcUczQjFlaHdPYTNrZlF0QT09

Title: “Partial differential equations to obtain full-field stress from measured kinematic fields”

Speaker: Dr Benjamin Cameron, University of Southampton

ABSTRACT:

I’ll introduce new theoretical/computational method to measure material properties in places where they are really difficult to measure: e.g. in a shock wave, in a shear band, or at a crack tip. These are all of critical importance in many engineering problems such as car crashes, manufacturing processes, or geotechnical slope stability – but it remains an ongoing technical challenge. I introduce a new experimentally driven approach to quantifying material behaviour using high spatial resolution datasets such as DIC/DVC. I’ll focus on developing novel physics-based partial differential equations that enable one to predict material properties, based on an observed strain-field [1]. This will allow measurement and quantification of material behaviour even in challenging cases such as high strain rate deformation, plastic flow instabilities, or heterogeneous materials. 
[1]  B.C. Cameron, C.C. Tasan, Full-field stress computation from measured deformation fields: a hyperbolic formulation, J. Mech. Phys. Solids. 147 (2021). https://doi.org/10.1016/j.jmps.2020.104186 

BIOGRAPHY:

Dr Benjamin Cameron is a lecturer in the department of mechanical engineering at the university of Southampton. He obtained his doctorate in 2021 from Massachusetts institute of technology. He is focused on the deep integration of computation and experiment in order to design the next generation of materials from the microstructure up. He works on new theoretical models of plastic deformation and statistical techniques to interpret complex microstructures.

GCEC Seminar: Development of a finite element code for multiscale modelling of large plastic deformations in polycrystalline materials at elevated temperatures

Group: Glasgow Computational Engineering Centre
Speaker: Benjamin Alheit, University of Cape Town
Date: 16 December, 2022
Time: 15:00 - 16:00
Location: Zoom meeting

Title: Development of a finite element code for multiscale modelling of large plastic deformations in polycrystalline materials at elevated temperatures

Abstract: Metal alloys commonly contain non-metallic elements, which, during thermomechanical processing, precipitate out of solution to form intermetallic particles (IMPs) embedded in a polycrystalline matrix. It has been observed, at least anecdotally, that IMPs endow the bulk material with mechanical properties in certain industrial applications. For example, in the forming of beverage can body, IMPs are thought to improve galling resistance. Additionally, it has been hypothesized that IMPs affect texture evolution, which, in turn, affects macroscale anisotropic yield behaviour. Clearly, it would be useful to have a computational tool to investigate the validity of the afore mentioned hypotheses and anecdotal observations. The development of such a tool using the finite element library deal.ii is the topic of this presentation.

The outline of the presentation is as follows:

1. Some high-level aspects of the code design;
2. Illustration of some of the code features:
• Finite strain rate-dependent polycrystalline plasticity;
• Finite strain rate-dependent single crystal plasticity;
• Periodic boundary conditions;
• Homogenization operations.
3. Preliminary studies on representative volume elements (RVEs) containing IMPs and face centred cubic (FCC) crystals;
4. Preliminary studies of plane strain compression tests;
5. Implementation of an arbitrary Lagrangian Eulerian (ALE) formulation for large plastic deformations – work in progress.

Bio: Benjamin Alheit recently completed his PhD at the University of Cape Town and currently holds a Postdoctoral Fellowship to conduct research within the Centre for Research in Computational and Applied Mechanics (CERECAM) and the Centre for Materials Engineering (CME). His research interests are in the broad domain of computational solid mechanics, with a particular interest in scientific software development, material characterisation, and multiscale modelling. His PhD work concerned multiscale modelling of biological structures and has been published in the Journal of the Mechanics and Physics of Solids, Acta Biomaterialia, and the Journal of the Mechanical Behaviour of Biological Materials.

Zoom link: https://uofglasgow.zoom.us/j/81996798950?pwd=ZUFIV3pKZVczakxhTW9wS05jOGI0Zz09

GCEC Seminar: Robust and efficient approximation of the compressible Euler and Navier-Stokes equations

Group: Glasgow Computational Engineering Centre
Speaker: Matthias Maier, Texas A&M University
Date: 30 September, 2022
Time: 15:00 - 16:00
Location: Online Zoom Webinar

  Title: Robust and efficient approximation of the compressible Euler and Navier-Stokes equations
 
  Abstract: Structure preserving numerical methods provide theoretical guarantees of
  reliability for situations where ad-hoc stabilization techniques can fail.
  In this talk we present fully discrete approximation techniques for the
  compressible Euler and Navier-Stokes equations that is second-order
  accurate in time and space and guaranteed to be invariant domain
  preserving. This means the method maintains important physical invariants
  and is guaranteed to be stable without the use of ad-hoc tuning parameters.

  We discuss the underlying algebraic discretization technique based on
  collocation and convex limiting, and briefly comment on a high-performance
  implementation utilizing SIMD (single instruction multiple data)
  vectorization and OpenMPI parallelization. We conclude with a short
  overview of concrete applications to related hyperbolic equations such as
  the shallow water equations and the Euler-Poisson system.

Zoom link: https://uofglasgow.zoom.us/j/87524131080?pwd=czMxaENwOFc0aTh0c1dQRHFqOW8wZz09

GCEC seminar by Dr Stefanos Aldo Papanicolopulos

Group: Glasgow Computational Engineering Centre
Speaker: Stefanos Aldo Papanicolopulos, University of Edinburgh
Date: 23 September, 2022
Time: 15:30 - 17:00
Location: Rankine Building, Room 816

GCEC seminar

Group: Glasgow Computational Engineering Centre
Speaker: Dr Vladislav Yastrebov, MINES ParisTech, PSL University (France)
Date: 20 September, 2022
Time: 15:30 - 17:00
Location: Wolfson Building, Seminar Room 1 (Yudowitz)

GCEC seminar by Dr Stefanos Aldo Papanicolopulos

Group: Glasgow Computational Engineering Centre
Speaker: Stefanos Aldo Papanicolopulos, University of Edinburgh
Date: 26 August, 2022
Time: 15:30 - 17:00
Location: Rankine Building, Room 816

GCEC Seminar: Statistical finite element method for physics-informed digital twin of infrastructures

Group: Glasgow Computational Engineering Centre
Speaker: Eky Febrianto, University of Glasgow
Date: 20 May, 2022
Time: 15:00 - 16:00
Location: University of Glasgow

Title: Statistical finite element method for physics-informed digital twin of infrastructures

Abstract: The monitoring of infrastructure assets using sensor networks is becoming increasingly prevalent. A digital twin in the form of a finite element (FE) model can help make sense of the copious amount of collected sensor data. This talk will introduce a statistical finite element method (statFEM) that provides the means of synthesising measurement data and FE models. In statFEM, the data and FE model are random variables with uncertainties associated with measurement error, random inputs, and modelling assumptions. A physics-informed prior density distribution of the system response, e.g., strain or displacement, is given by a conventional stochastic forward problem. The posterior density of the system response is obtained through the Bayes rule from the postulated prior density and a data-dependent likelihood function. This talk will also present the application of statFEM to predict the structural response of an instrumented steel railway bridge and nonlinear continuous welded rail system.

Biosketch: Eky is a lecturer in computational mechanics and a member of the Glasgow Computational Engineering Centre (GCEC) at the University of Glasgow. He obtained his PhD in computational mechanics from the University of Cambridge. He did his postdoctoral research on a strategic theme of data-centric engineering at The Alan Turing Institute and was a visiting research fellow at the University of Cambridge. His research combines aspects of advanced computational methods, geometry, and statistics for developing physics-informed digital twins of engineering structures across scales.

Location: James Watt South Building, Room 603, Creaitivity Lab

GCEC Seminar

Group: Glasgow Computational Engineering Centre
Speaker: David Kamensky, University of California, San Diego
Date: 31 March, 2022
Time: 16:00 - 17:00
Location: Online Zoom Webinar

GCEC Seminar: Of Microbes, Mechanics and Materials

Group: Glasgow Computational Engineering Centre
Speaker: Anupam Sengupta, University of Luxembourg
Date: 09 June, 2021
Time: 15:30 - 16:30
Location: Online

Title: Of Microbes, Mechanics and Materials

Abstract: Understanding how microbes interface, exchange and communicate with their local surroundings is central to the grand quest for a theory of microbial ecology. From simple to complex fluids, from compliant to rigid surfaces, microbes inhabit plethora of micro-environments spanning vastly different structures, dynamics, and internal energies. Currently we lack a biophysical framework that could explain, generalize, and crucially, predict the if-s, the how-s, and the why-s of the microbe-environment interactions. Research in my lab aims to fill this gap by interfacing soft matter physics and fluid mechanics with microbiology and genetic engineering. In this talk I will discuss that microbes – across individual, species and community scales – are inherently coupled to their micro-environments, and that their behavioural and physiological traits emerge as a consequence of active biophysical feedbacks between the material, information and energy transport processes. Using vignettes from our recent experiments in model gut and aquatic microbial systems, I will demonstrate how microbes and their micro-environments crosstalk via biomechanical coupling, leading to emergence of traits that ultimately translate into ecological and eco-physiological functions. I will discuss the generality of our results across microbial worlds, specifically touching upon the role and ramification of fluctuations in microbial environments. I will conclude by discussing why our efforts to unpack the microbe-mechanics-materials nexus are central to deciphering microbial fitness, succession, and selection, not least for their emerging prospects in medical diagnostics, biotechnology, and bioremediation during current climatic trends.

Bio: Anupam Sengupta is an ATTRACT Fellow and tenure track Professor of Biological Physics at the University of Luxembourg. Anupam directs the Physics of Living Matter Group, a cross-disciplinary team of scientists working on emergent functionalities in biological systems. Research in the Sengupta Lab spans soft and living matter physics, microfluidics, optics and microbiology. Anupam holds a B.S. and M.S degree (Dual Degree) in Mechanical Engineering from IIT Bombay, India, and a Ph.D. in Soft Matter Physics for his thesis on Liquid Crystal Microfluidics, carried out at the Max Planck Institute for Dynamics and Self Organization, Göttingen, Germany (2013). As a postdoc at MIT (Cambridge, USA) and ETH Zurich (Switzerland), Anupam switched fields to work on the physical ecology of aquatic microbes, and the mechanics of microbial growth and adaptation. Anupam is a recipient of the Human Frontiers Cross-Disciplinary Fellowship (2014-2017) by the International Human Frontier Science Program Organization, and was selected as one of the six “promising young scientists for future” by Nature during the 65th Lindau Meeting for Nobel Laureates in Lindau, Germany. Anupam is the Director for Undergraduate Physics Studies at the University of Luxembourg.

Zoom link: https://uofglasgow.zoom.us/j/97254708828?pwd=bk9KR094WkZEQWZJT09xdHFSak0yUT09

GCEC Seminar: Multiscale modeling of lung biomechanics

Group: Glasgow Computational Engineering Centre
Speaker: Prof Daniel Hurtado, Pontificia Universidad Católica de Chile
Date: 12 May, 2021
Time: 15:30 - 16:30
Location: Online Zoom Webinar

Title: Multiscale modeling of lung biomechanics

Abstract: Covid-19 pneumonia has quickly become a leading cause of death worldwide, boosting the interest of the scientific computing community in creating accurate models of the respiratory system for in silico experimentation and medical discovery. In this talk, I will present our current efforts towards creating a multiscale framework to achieve whole-lung predictive simulations. Drawing concepts from finite-deformation homogenization theory, I will introduce a microstructural model for the poroelastic behavior of the lung tissue. I will further discuss our validation efforts and comment on the predictiveness of the tissue model. Finally, I will present how the proposed micromechanical model can be integrated into whole-lung simulations of healthy and diseased conditions and discuss future directions in lung modeling.

Short Bio: Daniel Hurtado is an associate professor with the School of Engineering and the Institute for Biological and Medical Engineering at Pontificia Universidad Catolica de Chile. He leads the Computational Medicine Group, an interdisciplinary team that focuses on creating physiology-based digital replicas of the human lungs, with applications in the study of mechanical ventilation therapies and early diagnosis of pulmonary diseases. Prof. Hurtado received his M.S. and Ph.D. degrees from the California Institute of Technology as a Fulbright fellow. He is an elected member of the World Council of Biomechanics since 2018.

Zoom link: https://uofglasgow.zoom.us/j/92841584390?pwd=aHZFYUtBS2kyU0duK3ZmanVEeXl4Zz09

GCEC Seminar: Adaptive tensor methods for scientific computing

Group: Glasgow Computational Engineering Centre
Speaker: Damiano Lombardi, Inria Paris
Date: 30 April, 2021
Time: 15:00 - 16:00
Location: Online Zoom Webinar

Title: Adaptive tensor methods for scientific computing

Abstract: High-dimensional problems arise in many areas of science and engineering. The so called "curse of dimensionality" makes the solution approximation a challenging task, preventing us from using standard discretisations. Several classes of methods have been proposed (and are currently investigated) in the literature in order to approximate the solution of high-dimensional problems: tensor methods are one of them. They consist in using in a systematic way the principle of separation of variables in order to provide an approximation of the solution. In the methods which are proposed in the literature, the "size" of the tensor (called rank) is often fixed a priori, which could lead to severe limitations on realistic problems. The main goal of the work is to propose adaptive tensor methods: the memory used and the tensor format (to some extent) are computed along with the solution approximation in order to fulfil a prescribed accuracy and optimise the computational resources.

In this talk, we are going to discuss first a possible discretisation of a 3d-3d Vlasov-Poisson system. The results obtained are encouraging and provide a motivation for two methodological investigations: the first one consists in defining an adaptive piece-wise tensor decomposition; the second one is an investigation of a multi-linear system solver which is particularly well suited for systems of parametric Partial Differential Equations. Some theoretical results and some numerical experiments are shown.

Brief bio: Damiano Lombardi is a researcher in scientific computing at Inria Paris since 2013, working in the COMMEDIA team (mathematical modeling of the cardiovascular system).

Zoom link: https://uofglasgow.zoom.us/j/96209524406?pwd=Snd3V0VFKy9aR3BzWm5rM0Z5bERVQT09

From discrete particles to continuum fields

Group: Glasgow Computational Engineering Centre
Speaker: Thomas Weinhart, University of Twente
Date: 04 March, 2021
Time: 13:00 - 14:00
Location: https://uofglasgow.zoom.us/j/96893295506?pwd=ZHptZ2JFdjIwUGFVc2ZSSWJaVVM5QT09

Micro–macro transition methods are used to calibrate and validate continuum models from discrete data, obtained from either experiments or simulations. Such methods generate continuum fields such as density, momentum, stress, etc, from discrete data, i.e. the particles' positions, velocity, and forces. Performing this micro–macro transition step is especially challenging for heterogeneous and dynamic situations.

Here, we present a mapping technique, called coarse-graining, to perform this transition. This method has several advantages: (1) By construction, the obtained macroscopic fields are consistent with the continuum equations of mass, momentum and energy balance. (2) boundary interactions are accounted for in a self-consistent way and thus allow for the construction of locally accurate stress fields, even near boundaries. (3) Partial stresses and drag forces can be determined for individual constituents, which is critical for developing mixture models, e.g. for segregation. (4) The method does not require ensemble-averaging and thus can be efficiently exploited to investigate static, steady and time-dependent flows. 

The method is valid for any discrete data, e.g. particle simulations, molecular dynamics, experimental data, etc. However, for the purpose of illustration we consider data generated from discrete particle simulations. We show how to practically use coarse-graining for both steady and unsteady flows using our open-source tool MercuryCG, which is available as part of the discrete particle solver MercuryDPM (http://www.mercurydpm.org/).

Phase-field modeling of microstructure formation in phase-change materials for memory applications

Group: Glasgow Computational Engineering Centre
Speaker: Professor Mathis Plapp, École Polytechnique
Date: 12 February, 2021
Time: 15:30 - 16:30
Location: https://uofglasgow.zoom.us/j/99466185021

Phase change materials (PCM) can exist in two different states, namely, as amorphous and crystalline solids, which can have markedly different physical properties. This is the basis for the usage of PCM for data storage applications. The ternary alloy of germanium, antimony, and tellurium (GST) has been intensely studied because at the stoichiometric composition Ge2Sb2Te5, the amorphous phase can rapidly crystallize without a change in composition. This makes it a material of choice for random access memory devices (PCRAM). The amorphous phase spontaneously crystallizes at a relatively low temperature, 160 °C. For numerous applications, this temperature has to be increased, for example by enriching the stoichiometric GST with germanium. However, then the crystallization is accompanied by a composition change: upon crystallization, the additional Ge segregates into the amorphous phase. This can lead to the nucleation and growth of a new crystalline phase which is rich in Ge. For the development of new materials, it is important to better understand this complex microstructure formation process.

A grand-canonical multi-phase-field model for the crystallization of GST has been developed, which takes into account three phases: the amorphous state, stoichiometric GST, and a Ge-rich crystalline phase. Within a pseudo-binary approximation of the phase diagram, the phase fields are coupled to a single concentration field; the grain structure of the polycrystalline material is described using orientation fields. The kinetic parameters of the model are determined from data on stoichiometric GST available in the literature. By coupling this model with an electrothermal solver that calculates the time-dependent temperature field on the device scale, the microstructure and composition distribution observed in real memory cells after device operations are qualitatively reproduced.

After studies of physics in Freiburg (Germany) and Paris (France) and a postdoc in Boston (USA), Professor Plapp joined the Condensed Matter Physics Laboratory of Ecole Polytechnique and Centre National de la Recherche Scientifique (CNRS) in 1999. Since then, his research is centered on pattern formation in materials, mainly during solidification. He has contributed to the development of the phase-field method, which has become a standard method for the modeling of microstructure evolution in materials. He has co-authored 80 publications in peer-reviewed journals and given more than 100 invited lectures and seminars. Since 2015, he is the director of the Condensed Matter Physics Laboratory.

GCEC Seminar: High-Fidelity Simulation of Brittle Fracture Problems

Group: Glasgow Computational Engineering Centre
Speaker: Prof Adrian Lew, Stanford University
Date: 09 February, 2021
Time: 16:00 - 17:00
Location: Online Zoom Webinar

Deformable particle modelling

Group: Glasgow Computational Engineering Centre
Speaker: Saeid Nezamabadi, Université de Montpellier
Date: 27 January, 2021
Time: 16:00 - 17:00
Location: https://uofglasgow.zoom.us/j/95020619485?pwd=TXZmY24wRmQ1eVIyZmRVTnFOV0F4dz09

Soft granular materials such as many pharmaceutical and food products can deform elastically or plastically without rupture under low confining pressure. It allows them to achieve high packing fractions above the jamming state of hard granular materials. In the context of soft granular materials, a combination of grain rearrangements and grain shape change hence controls their rheological properties. We proposed three numerical approaches to study the rheology of these materials. Two methods are based on coupling the continuum approaches (Material Point Method and Finite Element Method), accounting for a broad class of mechanical behaviors (elastic, plastic…) of individual grains, with the Contact Dynamics method to deal with contact interactions between grains [1-4]. Another approach is the Bonded Particle Model, which consists in modeling each grain as an aggregate composed of hard primary particles with cohesive interactions such that a grain can deform as a result of the relative motions of the particles while staying together as a solid grain [2]. In this work, we investigate the compaction of assemblies of elastic and plastic grains using the above numerical approaches. The effect of several parameters (friction, compressibility, plasticity…) of the particles on the properties (stress, packing fraction, connectivity…) of the soft granular systems are investigated beyond the jamming state. Some of these numerical results are compared with experiments on soft disks subjected to compressive loading and analyzed by means of the Digital Image Correlation (DIC) technique. The quantitative agreement between these approaches makes it possible to validate both the experimental method for the study of granular media with deformable particles, and the numerical approaches proposed.

References.
[1] S. Nezamabadi, F. Radjai, J. Averseng, J.-Y. Delenne. Journal of the Mechanics and Physics of Solids, 83: 72-87, 2015.
[2] S. Nezamabadi, T.-H. Nguyen, J.-Y. Delenne, F. Radjai. Granular Matter, 19: 8, 2017.
[3] S. Nezamabadi, X. Frank, J.-Y. Delenne, J. Averseng, F. Radjai. Computer Physics Communications, 237: 17-25, 2019.
[4] T.-L. Vu, S. Nezamabadi, S. Mora. Soft Matter, 16: 679-687, 2020. 
[5] T.-L. Vu, J. Barés, S. Mora, S. Nezamabadi. Physical Review E, 99 :062903, 2019.

Bio-sketch: 

Saeid Nezamabadi is an Associate professor in the Laboratory of Mechanics and Civil Engineering (LMGC) at the University of Montpellier, France, with research activities on powders and granular materials, and focussing on the link between the rheology of granular media and single particle properties with the aid of numerical simulations by combined continuum and discrete approaches. Application areas of interest are specially compressive and shear behaviours of soft particle materials. For details of research works, please see his personal page (http://www.lmgc.univ-montp2.fr/perso/saeid-nezamabadi/) and research team in LMGC (http://www.lmgc.univ-montp2.fr/perso/pmmd/).

GCEC Seminar: Micromorphic Tissue Mechanics Accounting For Non-Affine Myocardial Deformation Characteristics

Group: Glasgow Computational Engineering Centre
Speaker: Dr Sebastian Skatulla, Department of Civil Engineering, University of Cape Town
Date: 12 November, 2020
Time: 13:00 - 14:00
Location: Zoom

Zoom link: https://uofglasgow.zoom.us/j/93968561981?pwd=aHliYjd4bHNhN2pib1pLU1VNT24yZz09

Title: Micromorphic Tissue Mechanics Accounting For Non-Affine Myocardial Deformation Characteristics 

Abstract: Cardiovascular diseases are among the most common causes of death in the world. Computational modelling in combination with medical imaging techniques, mechanical tissue testing, as well as cell and molecular biological analysis has the potential to help better understanding the underlying physiological mechanisms of heart failure and guide decision making in finding patient-specific treatment options in the future.

Computational models, however, need to be realistic enough to accurately describe the highly heterogeneous and non-uniform myocardial material composition [1], its anisotropic mechanical properties, the electro-mechanical interaction during muscle contraction and other biological effects, such as residual stresses and remodelling processes.

In this contribution we want to focus on the passive response of the myocardium which is very compliant exhibiting large strains, in particular, while the heart is contracting and twisting to eject oxygenated blood into the circulatory system during the systolic phase of the heart cycle. In the past it has been discovered that the initially crimped and coiled collagen fibres straighten during passive filling [2] and that cardiac myocytes exhibit a certain degree of motion flexibility within the constraining cytoskeleton [3, 4]. In contrast to classical models of phenomenological nature, this work proposes a micromorphic continuum-based formulation [5] which features extra degrees of freedom and corresponding strain and stress measures. The approach can therefore account for the hierarchical fibrous characteristics of the myocardium which are associated with micro-structural deformation of muscle-fibre bundles as well as their motion relative to the bulk material. As such, the assumed hyperelastic material behaviour of myocardial tissue is represented by a non-linear strain energy function which includes contributions linked to the bulk material representing the cytoskeleton and the micromorphic-fibre continuum emulating the micro-kinematics of the interwoven muscle-fibre bundles.

Short Bio: Sebastian Skatulla is an Associate Professor of Structural Engineering and Mechanics at the University of Cape Town. He graduated  as Diplom Bau-Ingenieur (TH) from the Karlsruhe Institute of  Technology (KIT) in 2003. He was awarded his PhD degree in Mechanical Engineering from the University of Adelaide in 2007.

He is the Director of the Computational Continuum Mechanics Research Group (CCM)  which has its research activities centred in multiscale and multiphase continuum methods. Current activities comprise the poroelasticity of Antarctic sea-ice and biological tissue.

He is the President of the South African Association for Theoretical and Applied Mechanics (SAAM) and member of the Scientific Council of the International Centre for Mechanical Sciences (CISM).

References

[1] LeGrice, I.J., Smaill, B.H., Chai, L.Z., Edgar, S.G., Gavin, J.B. and Hunter, P.J., Laminar structure of the heart: ventricular myocyte arrangement and connective tissue architecture in the dog. Am. J. Physiol. Heart Circ. Physiol., 269:H571–H582, 1995.

[2] Robinson, T.F., Geraci, M.A., Sonnenblick, E.H. and Factor, S.M., Coiled perimysial fibers of papillary muscle in rat heart: Morphology, distribution, and changes in configuration. Circulation Research, 63:577–592, 1988.

[3] LeGrice, I.J., Takayama, Y. and Covell, J.W., Transverse shear along myocardial cleavage planes provides a mechanism for normal systolic wall thickening. Circulation research, 77(1):182–193, 1995.

[4] Spotnitz, H.M., Spotnitz, W.D., Cottrell, T.S., Spiro, D. and Sonnenblick, E.H., Cellular basis for volume related wall thickness changes in the rat left ventricle. Journal of molecular and cellular cardiology, 6(4):317–331, 1974.

[5] von Hoegen, M., Skatulla, S. and Schrder, J., , ”A generalized micromorphic approach accounting for variation and dispersion of preferred material directions”, Computers and Structures, 232: 105888, 2020

GCEC seminar: Characterising and modelling fracture in soft solids

Group: Glasgow Computational Engineering Centre
Speaker: Prof Maria Charalambides, Imperial College London
Date: 21 October, 2020
Time: 14:00 - 15:00
Location: Zoom meeting

This talk will summarise studies towards developing predictive numerical models of deformation and fracture within soft solid materials. Computational modelling of the in-vivo mechanical response of various biological materials within the human organism, such as brain tissue, bone, arteries and ingested food, is an increasingly cost-effective design tool for bio-medical, bio-engineering and surgical applications. Hydrated food is one such system produced naturally after swallowing and its mechanical response is a crucial factor during digestion which impacts on the rate of release of nutrients to the human body. We show that a viscoplastic-damage constitutive law calibrated through compression tests on hydrated biscuit particles, can be utilised in Eulerian Finite Element (FE) analysis to predict complex localised deformation-fracture material behaviour during wire cutting at two length scales with high fidelity. We demonstrate that in such materials a fracture term is not always necessary to predict ultimate separation and that the Eulerian FE analysis is a versatile approach based on which largely different material cutting behaviours can be modelled. The use of these methods to develop in silico models of the critical initial digestion stages will be presented.

Maria will also give an overview of the research within the Mechanics of Materials Division of the Department as well as her own other research conducted at the Soft Solids research group.

Zoom link: 

https://uofglasgow.zoom.us/j/99798134060?pwd=MTZhV28rZmdNTzZzVVlUYk1pcG5Pdz09

GCEC Seminar: Finite element and virtual element approximations for problems of isotropic and anisotropic elasticity

Group: Glasgow Computational Engineering Centre
Speaker: Prof Daya Reddy, Department of Mathematics and Applied Mathematics at the University of Cape Town
Date: 15 October, 2020
Time: 15:00 - 16:00
Location: Online

Zoom Link: https://uofglasgow.zoom.us/j/96885164003?pwd=SkRpeExBUVQ2b2JETHlLdStNVFR6UT09

Title: "Finite element and virtual element approximations for problems of isotropic and anisotropic elasticity”

Abstract: This presentation is concerned with the behaviour of elastic bodies under limiting conditions of near-incompressibility and, for transversely isotropic materials, near-inextensibility. The concept of volumetric locking – that is, a lack of convergence - in the context of low-order finite element approximations is well understood, and a variety of effective remedies exist. Corresponding studies in relation to near-inextensibility have on the other hand been somewhat limited. Both theoretical and computational approaches to these classes of problems are explored. The virtual element method (VEM) is a relatively recent extension of the finite element method (FEM) that allows for the use of polygonal or polyhedral elements having any number of sides or surfaces. Near-incompressibility and near-inextensibility are explored using both the FEM and VEM. In particular, the excellent convergence properties of low-order Virtual Element approximations is demonstrated, for small and large deformations.

Biography: Daya Reddy completed a bachelor’s degree in civil engineering at the University of Cape Town and a Ph.D. degree at Cambridge University in the UK. He currently holds the South African Research Chair in Computational Mechanics, in the department of mathematics and applied mathematics at the University of Cape Town.

His teaching and research activities reflect his multidisciplinary perspectives, which he pursues largely through the Centre for Research in Computational and Applied Mechanics, a centre comprising academic staff and postgraduate students in five different departments. Much of his work is concerned with mathematical and numerical analysis of problems arising in solid and fluid mechanics. His many publications include two graduate-level texts and a research monograph, now in its second edition, on plasticity theory.

Daya Reddy is actively involved in bodies that work towards strengthening the scientific enterprise and providing science advice to policymakers. He served a term as president of the Academy of Science of South Africa, and is currently president of the International Science Council, the largest representative non-governmental global scientific organization.

He is a recipient of the Award for Distinguished Service from the South African Association for Computational and Applied Mechanics,the SAMS Award for Research Distinction, and the Order of Mapungubwe from the President of South Africa. He has held numerous visiting positions, including those of Visiting Faculty Fellow at the University of Texas at Austin and the Timoshenko Lecturer at Stanford University. He is a recipient of the Georg Forster Research Award from the Alexander von Humboldt Foundation of Germany.

GCEC Seminar: Computational Fluid-Structure Interaction Modeling for Aerospace and Biomedical Applications

Group: Glasgow Computational Engineering Centre
Speaker: Prof Ming-Chen Hsu, Department of Mechanical Engineering, Iowa State University
Date: 03 September, 2020
Time: 15:30 - 16:30
Location: Zoom meeting

Zoom link: https://uofglasgow.zoom.us/j/96821163575?pwd=ZE11UXcxc3ora3Jld3RtNGJBMHI5QT09

Title: Computational Fluid-Structure Interaction Modeling for Aerospace and Biomedical Applications

Speaker: Prof. Ming-Chen Hsu, Department of Mechanical Engineering, Iowa State University

Abstract: Fluid-Structure Interaction (FSI) is a multiphysics phenomenon that occurs when moving or deformable structures interact with internal or surrounding fluid flows. The coupling between the dynamics of the fluid and mechanics of the structure often gives rise to unexpected behaviors vital to many science and engineering problems. In this presentation, I will discuss a new computational FSI framework developed based on isogeometric and immersogeometric analysis with application to the modeling and simulation of aerospace and biomedical problems. The fully-coupled FSI formulation is derived using the augmented Lagrangian approach to enforce kinematic and traction constraints and naturally accommodates nonmatching and non-boundary-fitted fluid-structure interfaces. This novel method can make direct use of the CAD boundary representation of a complex design structure and effectively deal with FSI problems involving large deformations of the fluid domain, including changes of topology. The key ingredients to achieving high simulation accuracy will be reviewed. The proposed FSI framework is applied to engineering and science applications at different scales, ranging from studying complex military aircraft tail buffeting due to different angles of attack to understanding prosthetic heart valve leaflet flutter under physiological conditions. The findings and challenges will be shown and discussed in detail.

Biography: Ming-Chen Hsu is an Associate Professor in the Department of Mechanical Engineering at Iowa State University. He received his MS degree in Engineering Mechanics from UT Austin in 2008 and PhD degree in Structural Engineering from UC San Diego in 2012. From 2012 to 2013, he was a postdoctoral fellow at the Institute for Computational Engineering and Sciences at UT Austin before joining Iowa State University. He is the recipient of the 2019 USACM Gallagher Young Investigator Award and is listed as a Web of Science Highly Cited Researcher from 2016 to 2019. He has published over 70 peer-reviewed journal papers and serves on several national and international professional society committees on computational methods and applications. His research focuses on computational mechanics, engineering, and sciences with an emphasis on fluid-structure interaction problems.

GCEC Seminar: Manifold Learning Based Data-Driven Modeling for Biological Tissues

Group: Glasgow Computational Engineering Centre
Speaker: Prof J. S. Chen, University of California, San Diego
Date: 30 July, 2020
Time: 16:00 - 17:00
Location: Zoom meeting

Zoom link: https://uofglasgow.zoom.us/j/96664866429?pwd=ekJmek94WDJWdGZrdUlWVzNQZE1vQT09

Abstract: Data-driven modeling is a hybrid approach that integrates universal physical laws with experimental material data directly to circumvent the necessity of using phenomenological constitutive models. A data-driven simulation approach based on manifold learning technique, termed locally convex data-driven (LCDD) computing, is formulated under the Galerkin meshfree framework to simulate heart valve tissues under finite deformation, where the material data from biaxial and pure shear tests were employed. The proposed approach reconstructs a local material manifold with the convex hull based on the nearest experimental data to the given state, and seeks for the optimum solution via the projection onto the associated local manifold. This learning process of local data structure leads to less sensitivity to noisy data and convergence enhancement. A penalty relaxation is also introduced to recast the local learning solver in the context of non-negative least squares that can be solved effectively. Due to the inherent manifold learning properties, LCDD performs well for high-dimensional data sets that are relatively sparse in real-world engineering applications. The reproducing kernel approximation with stabilized nodal integration is employed for the solution of the physical manifold to allow reduced stress–strain data at the discrete points for enhanced effectiveness in the LCDD learning solver. The employment of meshfree approximation and discretization allows for a smooth transition of material properties at the interface of different biological material components and a better approximation of stress/strain fields in the data-driven solver due to the enhanced smoothness. The numerical results demonstrate the effectiveness of the proposed data-driven approach for modelling complex biological materials.

Short-Bio: J. S. Chen is the William Prager Chair Professor of Structural Engineering Department, Professor of Mechanical and Aerospace Engineering Department, and the Director of Center for Extreme Events Research at UC San Diego. Before joining UCSD in October 2013, he was the Chancellor’s Professor of UCLA Civil & Environmental Engineering Department where he served as the Department Chair during 2007-2012. J. S. Chen’s research is in computational mechanics and multiscale materials modeling with specialization in the development of meshfree methods. He is the Past President of US Association for Computational Mechanics (USACM) and the Past Present of ASCE Engineering Mechanics Institute (EMI). He has received numerous awards, including the Computational Mechanics Award from International Association for Computational Mechanics (IACM), ICACE Award from International Chinese Association for Computational Mechanics (ICACM), the Ted Belytschko Applied Mechanics Award from ASME Applied Mechanics Division, the Belytschko Medal, U.S. Association for Computational Mechanics (USACM), among others. He is the Fellow of USACM, IACM, ASME, EMI, SES, ICACM, and ICCEES.

GCEC Seminar: Phase-field modeling of brittle fracture: an overview and a new paradigm to address multiple solutions

Group: Glasgow Computational Engineering Centre
Speaker: Laura De Lorenzis, ETH Zürich
Date: 18 June, 2020
Time: 14:30 - 15:30
Location: https://uofglasgow.zoom.us/j/98252849021

Abstract: The phase-field modeling approach to fracture has recently attracted a lot of attention due to its remarkable capability to naturally handle fracture phenomena with arbitrarily complex crack topologies in three dimensions. On one side, the approach can be obtained through the regularization of the variational approach to fracture introduced by Francfort and Marigo in 1998, which is conceptually related to Griffith's view of fracture; on the other side, it can be constructed as a gradient damage model with some specific properties. The functional to be minimized is not convex, so that the necessary stationarity conditions of the functional may admit multiple solutions. The solution obtained in an actual computation is typically one out of several local minimizers. Evidence of multiple solutions induced by small perturbations of numerical or physical parameters was occasionally recorded but not explicitly investigated in the literature.

In the first part of this talk, the speaker gives a brief overview of the phase-field approach to fracture and of recent related research carried out in her group. In the second part of the talk, the focus is placed on the issue of multiple solutions. Here a paradigm shift is advocated, away from the search for one particular solution towards the simultaneous description of all possible solutions (local minimizers), along with the probabilities of their occurrence. We propose the stochastic relaxation of the variational brittle fracture problem through random perturbations of the functional and introduce the concept of stochastic solution represented by random fields. In the numerical experiments, we use a simple Monte Carlo approach to compute approximations to such stochastic solutions. The final result of the computation is not a single crack pattern, but rather several possible crack patterns and their probabilities. The stochastic solution framework using evolving random fields allows additionally the interesting possibility of conditioning the probabilities of further crack paths on intermediate crack patterns.

Bio-sketch: Laura De Lorenzis received her Engineering Degree and her PhD from the University of her hometown Lecce, in southern Italy, where she then became Assistant and later Associate Professor of Mechanics. In 2013 she moved to the TU Braunschweig, Germany, as Professor and Director of the Institute of Applied Mechanics. Since February 2020 she is Professor of Computational Mechanics at the ETH Zürich. She was visiting scholar in several renowned institutions, including Chalmers University of Technology, the Hong Kong Polytechnic University, the Massachusetts Institute of Technology (Fulbright Fellowship), the Leibniz University of Hannover (Alexander von Humboldt Fellowship), the University of Texas at Austin and the University of Cape Town. She is the recipient of several prizes, including the RILEM L’Hermite Medal 2011, the AIMETA Junior Prize 2011, the IIFC Young Investigator Award 2012, two best paper awards and two student teaching prizes. In 2011 she was awarded a European Research Council Starting Researcher Grant. She authored or co- authored more than 120 papers on international journals on different topics of computational and applied mechanics.

GCEC Seminar: Numerical Modeling of Rock Mass in Finite Element Simulations of Deep Tunneling

Group: Glasgow Computational Engineering Centre
Speaker: Magdalena Schreter, Universität Innsbruck
Date: 12 March, 2020
Time: 11:00 - 12:00
Location: Rankine Building, Room 816

Title: Numerical Modeling of Rock Mass in Finite Element Simulations of Deep Tunneling

Abstract: Analyzing the mechanical behavior of the rock-support system during the construction of a deep tunnel poses a complex three-dimensional time-dependent problem. To cope with these complexities, finite element simulations are a powerful tool for predicting the deformations occurring during the excavation process, for dimensioning the required tunnel support and for determining critical situations close to  collapse. However, in terms  of their predictive capacity, they strongly rely on the employed constitutive model for the surrounding rock mass, being the major lead-bearing part of the tunnel structure. In this lecture, challenges and methods for modelling the mechanical behavior of rock mass are discussed, and the application of an advanced constitutive model for rock mass in finite element simulations of the construction of a section of the Brenner Base Tunnel is presented.

Biosketch: Magdalena Schreter was born in Austria in 1991. From 2009 to 2015 she studied Civil Engineering at the University of Innsbruck, completing her Master’s thesis in the field of concrete structures. In 2015, Magdalena was employed as a university assistant at the Unit for Strength of Materials and Structural Analysis at the University of Innsbruck. In December 2018, she received her PhD for her thesis “A Gradient-Enhanced Rock Damage-Plasticity Model for Numerical Analyses of Deep Tunneling”, supervised by Prof. Günter Hofstetter. Since March 2019, she has been working as a postdoctoral researcher at the Unit of Strength of Materials and Structural Analysis at the University of Innsbruck.

GCEC Seminar: Want to start a CAE consultancy company?

Group: Glasgow Computational Engineering Centre
Speaker: Mark Yeoman, Continuum Blue
Date: 13 February, 2020
Time: 14:00 - 15:00
Location: Rankine Building, Room 601

Title: Want to start a CAE consultancy company?

Abstract: Come and hear from Mark Yeoman, the founder of Continuum Blue, a COMSOL Certified Consultancy that, 10 years ago was providing high end multiphysics simulation services to a number of blue chip companies and growing rapidly. At the time, Continuum Blue had received Technology Strategy Board (now Innovate UK) funding under a consortium to develop orthopaedic implants and predict bone remodelling in patients, and it was making head way into new industries such as renewables, aerospace, automotive and chemical processing, as well as running specialist analysis services for companies like GSK, Jaguar Land Rover and BAE Systems. Fast forward 5 years, and after much growing pains, and a number of client related problems, he exited early, selling part of the company’s intellectual property, and set up a number of exclusivity agreements with clients to ensure long term, continued revenue. Come and hear about the issues he faced while starting Continuum Blue, and get a quick guide on the pitfalls of starting up a CAE consultancy company, or working for one.

Biography:Mark Yeoman, is the founder of Continuum Blue, he has a degree in engineering and a PhD in computational modelling and applied mathematics, where his postgraduate studies focused on the development of cardiovascular implants using numerical techniques and genetic algorithms for Medtronic Inc. Before starting Continuum Blue, he lectured in applied dynamics and engineering. With over 20 years of experience in the field of multiphysics modelling, he has worked in, and continues to work in a variety of industries, including medical, renewables, aerospace and chemical processing. In 2015 he sold some intellectual property developed at Continuum Blue and has since taken an extended sabbatical to explore other interests and business opportunities.

GCEC Seminar: Likely instabilities in stochastic elasticity

Group: Glasgow Computational Engineering Centre
Speaker: Angela Mihai, Cardiff University
Date: 07 February, 2020
Time: 11:00 - 12:00
Location: Rankine Building, Room 816

GCEC Seminar: Computational Inverse Mechanics to Facilitate Smart Structural Systems

Group: Glasgow Computational Engineering Centre
Speaker: John Brigham, Durham University
Date: 31 January, 2020
Time: 15:00 - 16:00
Location: Rankine Building, Room 816

Title: Computational Inverse Mechanics to Facilitate Smart Structural Systems

Abstract: Opportunities for leveraging computational mechanics are highly prevalent for the optimisation, control, and/or characterisation of increasingly complex systems through the development and utilization of sophisticated computational inverse solution strategies.  Moreover, the application of computational inverse techniques cuts across several disciplines, and is particularly relevant for a wide range of smart structure/structural system applications.  In terms of smart systems, capabilities range from developing a self-aware system (e.g., characterization of the current state, including potential damage and degradation) to facilitating adaptability (e.g., control or design optimisation to adapt based on the current state and environment).

This talk will present an overview of the activities of the Computational Diagnostics and Inverse Mechanics Research Group led by John Brigham, and then focus on two specific efforts to utilize computational mechanics and inverse solution strategies: (1) for the efficient solution of problems in characterisation of material properties in solids and structures and (2) optimal design of a smart material morphing building surface tile. These examples will show how many of the same tools, or at least the same overall framework, can be used to address a wide range of inverse problem applications.  Yet, the particular challenges of each application will often require unique attention, especially if given additional practical objectives, such as a need for computational efficiency.

Biosketch:  John Brigham received a BE from Vanderbilt University and a MS and PhD from Cornell University.  Following his PhD in 2008 he joined the University of Pittsburgh as an Assistant Professor, and was later promoted to Associate Professor in the Department of Civil and Environmental Engineering, with a secondary appointment in the Department of Bioengineering.  John joined the Department of Engineering at Durham University as an Associate Professor in Applied Mechanics in 2016 and was promoted to a Chair in Engineering in 2019.  Focusing on computational mechanics and inverse problems, John’s research group is actively involved in a number of diverse projects, including kinematic analysis of the heart for improved diagnosis of cardiovascular disease, novel design concepts and optimal design strategies for smart material morphing structures, and efficient and accurate quantitative nondestructive evaluation algorithms.

Applications of computational modelling to tackle pathophysiological problems in aortic and coronary circulations

Group: Glasgow Computational Engineering Centre
Speaker: Ryo Torii, UCL
Date: 27 November, 2019
Time: 11:00 - 12:00
Location: Rankine Building, Room 816

Bio: Ryo Torii is an associate professor at UCL Mechanical Engineering. His research interests have been in both developing new techniques of computational modelling and application to clinically-relevant problems, which are working hand-in-hand. Recent applications are primarily in the area of coronary artery atherosclerosis, especially to predict disease progression and to evaluate performance of various types of stents using computational modelling. He has been promoting use of computational modelling in clinical studies and has recently published an expert recommendation paper for clinical community (Gijsen et al. European Heart Journal 2019). He also has activities outside coronary artery pathophysiological analyses, including biomechanical analysis of aortic valve disease, growth prediction of tissue-engineered muscular constructs and process optimisation of biofabrication systems.

Diagnosis and treatment of thoracic aortic aneurysms: mechanical and computational aspects

Group: Glasgow Computational Engineering Centre
Speaker: Heleen Fehervary
Date: 24 July, 2019
Time: 16:00 - 17:00
Location: Rankine Building, Room 816

Abstract

Aortic aneurysms are a treat affecting approximately 4% of men aged between 65 and 74 in England. Current diagnostic tools only take the geometry of the aneurysm into account, while material properties of the arterial wall are neglected. Current aneurysm treatment is an invasive procedure where the diseased tissue is replaced with a synthetic vascular graft. A new promising treatment is the use of a personalized external aortic root support, but a mechanical evaluation of this treatment has not yet been conducted. In this talk I will present the research conducted in the Soft Tissue Biomechanics group, led by prof. Nele Famaey, to improve both diagnosis and treatment of thoracic aortic aneurysms. This includes the use of medical images, mechanical testing, parameter fitting and finite element modelling with algorithms that predict growth and remodeling of soft biological tissues.

GCEC Seminar

Group: Glasgow Computational Engineering Centre
Speaker: Heleen Fehervary
Date: 24 July, 2019
Time: 16:00 - 17:00
Location: Rankine Building, Room 816

Unravelling the interplay of structural and material properties of skin using mechanistic modelling

Group: Glasgow Computational Engineering Centre
Speaker: Georges Limbert, University of Southampton (UK) and University of Cape Town (South Africa)
Date: 04 June, 2019
Time: 12:00 - 13:00
Location: Rankine Building, Room 816

Considering the place of the skin in our life and its multiple physiological functions, understanding its complex physiology and biophysics in health, disease and ageing has become, particularly in the last two decades, a broad and very active multidisciplinary research arena. To unravel some of the secrets of such a complex organ new experimental, imaging and computational techniques are needed and novel mechanistic theories explaining particular mechanobiological processes need to be formulated and put to the test. Developing and exploiting such an integrated framework underpin many aspects of our research which aims to understand the interplay between the microstructural and material properties of the skin, particularly as they evolve over the life course. The skin microstructure can play a critical role in how macroscopic deformations are modulated at the microscopic level. These structural mechanisms are also at the heart of skin tribology by being part of, and conditioning mechanical load transmission and the nature of surface physics interactions. Skin biophysics is therefore fundamental to many industrial sectors from biomedical devices, personal care and cosmetic products to vehicle safety, sport equipment, wearable electronics and tactile surfaces.

In this talk, I will present some of the modelling approaches we have been developing to gain a mechanistic understanding of the interplay between material and structural properties of the skin, and ultimately, to exploit this knowledge for a variety of clinical and industrial applications. Examples will include computational contact homogenisation procedures to study skin friction, constitutive modelling of skin ageing and analysis of skin surface instabilities to understand mechanisms of wrinkle formation.

1. Limbert, G., Masen, M. A., Pond, D. Graham, H. K., Sherratt, M. J., Jobanputra, R. and McBride, A. (2019) Biotribology of the ageing skin-Why we should care. Biotribology, 17:75-90.
2. Graham, H. K., McConnell, J. C., Limbert, G., Sherratt, M. J. (2019) How stiff is skin? Experimental Dermatology, 28(S1):4-9.
3. Pond, D., McBride, A., Davids, L., Reddy, B.D., Limbert, G. (2018) Microstructurally-based constitutive modelling of the skin-Linking intrinsic ageing to microstructural parameters. Journal of Theoretical Biology, 444:108-123.
4. Limbert, G., Kuhl, E. (2018) On skin microrelief and the emergence of expression micro-wrinkles. Soft Matter, 14(8):1292-1300. 
5. Limbert, G., 2017. Mathematical and computational modelling of skin biophysics-A review. Proceedings of the Royal Society Part A, 473:1-39. 
6. Leyva-Mendivil, M.F., Lengiewicz, J., Page, A., Bressloff, N.W., Limbert, G., 2017. Skin microstructure is a key contributor to its friction behaviour. Tribology Letters 65, 12.

Georges Limbert (GL) is an Associate Professor at the national Centre for Advanced Tribology in the Faculty of Engineering and Physical Sciences at the University of Southampton. He also holds an Honorary Associate Professor position in the Department of Human Biology within the Faculty of Health Sciences at the University of Cape Town, South Africa. His current research focus is on the constitutive modelling of soft tissues with special interest on skin biophysics and the non-linear mechanics and tribology of biological structures His research is supported by world-leading organisations operating in the consumer goods, cosmetic, pharmaceutics and military sectors. GL acts as a consultant for Fortune 500/FTSE 100 and SME companies. Some of his constitutive models are used in industry, academia and the US Army. He has a Master in Engineering Mechanics (Toulouse, France) and a Master in Solid Mechanics (Bordeaux, France). He obtained a PhD in Computational Biomechanics from the University of Southampton in 2002 and is a Chartered Engineer and Fellow of the Institution of Mechanical Engineers (IMechE).

Nanoscale simulations and digital design of construction materials

Group: Glasgow Computational Engineering Centre
Speaker: Enrico Masoero, School of Engineering, Newcastle University
Date: 17 May, 2019
Time: 14:00 - 15:00
Location: Rankine Building, Room 816

Abstract

Concrete and other composite construction materials owe their properties to the collective behaviour of multiple interacting phases. Optimising the chemical composition of these complex materials is a challenge that often requires long and costly experimental campaigns. There is hope that the increasing predictive abilities of models and simulations will soon provide an in-silico route to complement the design of new composite materials, enabling a preliminary shortlisting of promising solutions that can later be tested using fewer but more focussed experiments. Within this picture, a crucial role is played by modelling and simulation at the nanoscale, where the linkage between molecular chemistry, structure, and properties starts to emerge. This talk will discuss the pathway to exploit the predictive abilities of molecular simulations to obtain constitutive laws for larger scale models, e.g. homogenisation theory and Finite Element simulations. Special attention is paid to the mesoscale between the nanometre and the micrometre, where unique structural features (e.g. mesopores) and properties emerge, which can significantly affect the macroscopic behaviour of the material. Two examples are discussed in some details: one is about molecular modelling of geopolymer cements, and the other one is about mesoscale simulations of calcium silicate hydrate precipitation and calcium silicates dissolution. The results are analysed in terms of ability to predict relationships between structure, chemistry, and properties of individual phases, which is the starting point to then combine different phases, e.g. using appropriate homogenisation schemes.

Bio

Enrico obtained his PhD in structural engineering at Politecnico di Torino, Italy, in 2010, with a dissertation on structural collapse. From 2010 until 2013 he was postdoc at the Concrete Sustainability Hub at MIT, USA, working on nanoparticle simulation of cement hydration. Since 2013, Enrico is Lecturer in Structural Engineering at Newcastle University, in the UK. His core research is in the hygro-chemo-mechanics of cement and concrete, from the nanoscale up to the macroscale of engineering application. Enrico has authored 23 articles in international journals. He has been a member of the TU1404 COST Action on concrete durability, and is member of the ASCE EMI Materials Properties committee. At Newcastle University, Enrico teaches Structural Mechanics, Engineering Materials, Structural Collapse, and Multiscale Modelling.

Fluid-structure interaction and contact problems in patient-specific cardiovascular modeling

Group: Glasgow Computational Engineering Centre
Speaker: Dr. Miquel Aguirre, Mines Saint-Etienne, Univ Lyon
Date: 15 May, 2019
Time: 15:00 - 16:00
Location: Rankine Building, Room 816

CRIMSON (CardiovasculaR Integrated Modelling and SimulatiON) is a three-dimensional modelling and simulation software environment for patient specific computational hemodynamics. The computational core of CRIMSON relies on PHASTA [1], a massively parallel flow solver for both compressible and incompressible flow. To account for the compliance of the vessel wall and, consequently, capture the wave propagation phenomena within the human vasculature, CRIMSON uses the Coupled Momentum Method algorithm, presented by Figueroa and coauthors [1], which represents the vessel wall as a linear membrane working under the small displacement assumption. Overall, this results into a computationally efficient algorithm that has proven to be very robust and has provided numerous clinically relevant results. The first part of the talk will present the extension of the CMM to the nonlinear regime by using a fully coupled Arbitrary-Lagrangian-Eulerian (ALE) formulation. The vessel wall is modeled by means of a Total Lagrangian membrane model. For the fluid part, the Navier-Stokes equations under their ALE form are solved using a stabilized finite element scheme. As in the CMM, the membrane model avoids additional degrees of freedom. The mesh motion is handled using a 3D linear elastic solid and the time integration is performed using the generalized alpha method.

Endovascular procedures, a part of minimally invasive surgery (MIS), are designed to minimize the patient’s trauma and have dramatically changed the outcome of cardiovascular interventions. Endovascular procedures are widely used to repair aortic aneurysm, aortic dissection, or to prevent embolization of a deep vein thrombus. Yet, the design of patient-specific endovascular prosthesis, such as stents, is a complex modeling task, as its performance will heavily depend on the positioning within the patient’s vessel. This in turn depends on the surgeon’s precision and in the contact mechanics of the prosthesis against the vessel. On top of this, the surgeon performs the intervention without seeing the patient’s tissue. The surgeon is guided by X-ray, which shows only hard tissue and medical instruments. Contrast agent can be injected in order to see the soft tissue, but it is kept at a minimum level to avoid secondary effects. Because all of the above, there is a strong need in bringing information on the mechanics of stent-graft deployment in order to 1) aid in the design of patient-specific prosthesis and 2) provide computational tools for real-time visualization in the clinic. Based on extensive experience of the group in the topic [2], the second part of this talk will present the development of corotational beam models and contact mechanics to help in the design of endovascular prosthesis and towards real-time simulation of their deployment.

References

[1] C. A. Figueroa, I. E. Vignon-Clementel, K. E. Jansen, T. J. R. Hughes, and C. A. Taylor, “A coupled momentum method for modeling blood flow in three-dimensional deformable arteries,” Comput. Methods Appl. Mech. Eng., vol. 195, no. 41–43, pp. 5685–5706, 2006.

[2] D. Perrin ,  P. Badel, L. Orgéas, C. Geindreau, A. Dumenil, J. N. Albertini and S. Avril. “Patient specific numerical simulation of stent-graft deployment: Validation on three clinical cases”. Journal of biomechanics, 48(10), pp. 1868-1875.

GCEC Seminar: On analysis and applications of discrete scattering theory involving edges

Group: Glasgow Computational Engineering Centre
Speaker: Dr. Basant Lal Sharma, Indian Institute of Technology Kanpur
Date: 02 May, 2019
Time: 15:00 - 16:00
Location: Rankine Building, Room 816

In the presence of length scales, the elastic wave propagation problem, as well as those involving other kinds of dynamical phenomena (lattice waves or phonons, electronic transport, etc), involve an interplay between wave dispersion and structural features.

The present talk concerns an exposure to the analysis of discrete scattering effects in certain simple structures. As a prototype example of diffraction phenomenon, a discrete analogue of Sommerfeld diffraction by a half plane has been recently analysed for three different lattice structures: square, triangular, and hexagonal. The two simple cases that have been recently studied correspond to semi-infinite row with discrete Neumann condition and discrete Dirichlet condition. Both kinds of half-row `defects’ have been analyzed in infinite lattices as well as in waveguides using Fourier analysis.

The talk will give a flavour of the mathematical formulation and some techniques which are useful to derive physically relevant entities; for the infinite lattice, an example of the latter is far-field wavefunction and for the waveguides, the reflectance and transmittance. The talk will also include some open problems associated with these newly emerging developments as well as some glimpses of the ongoing work. An application of the developed framework to electronic transport in nanotubes will be also briefly discussed.

Multiscale Modelling of Reinforced Concrete (Structures)

Group: Glasgow Computational Engineering Centre
Speaker: Adam Sciegaj, Chalmers University of Technology, Sweden
Date: 05 April, 2019
Time: 11:00 - 12:00
Location: Rankine Building, Room 816

Abstract:

Since concrete cracks at relatively low tensile stresses, the durability of reinforced concrete structures is highly influenced by its brittle nature. Cracks allow for ingress of harmful substances, e.g. chlorides, which in turn cause corrosion of the reinforcement. Therefore, crack widths are limited in the design codes. For larger structures, precise crack width prediction requires detailed models, which can become computationally expensive. Hence, multiscale modelling methods are of interest.

In this presentation, a multiscale model for reinforced concrete based on Variationally Consistent Homogenisation (VCH) is outlined. An effective reinforce concrete solid is considered at the macroscale, while the subscale modelling comprises plain concrete, reinforcement bars and the bond-slip interaction between the materials. The effective macroscopic response is obtained upon computational homogenisation of the response of individual Representative Volume Elements (RVEs), in a nested FE^2 (Finite Element squared) algorithm. Furthermore, enrichment of the multiscale model by a macroscopic reinforcement slip variable is addressed. Several application examples of modelling uniformly and non-uniformly reinforced concrete deep beams with this multiscale method are given, and results in terms of force-deflection relations, crack width and crack patterns are presented. 

About the speaker:

Adam Sciegaj is a visiting PhD student from Chalmers University of Technology in Gothenburg, Sweden. His research is focused on developing a multiscale modelling framework for reinforced concrete structures, with the aim to make it possible to study the crack growth in detail for large-scale structures, like e.g. bridges.

From dry quicksand to unsinkable suspensions - exploring the underlying links between dry granular flow and wet suspensions

Group: Glasgow Computational Engineering Centre
Speaker: Dr Jin Sun, University of Edinburgh
Date: 19 February, 2019
Time: 14:00 - 15:00
Location: Rankine Building, Room 514

The flow behaviour of dry granular materials, e.g. free flowing sand, is commonly observed to be very different from wet suspensions, e.g. sticky cornstarch-water mixtures. We show, through particle simulations and rheology experiments, that the different behaviour can be put in a unifying rheological framework, where the interplay between particle-size scaling and the observation window gives rise to the perception. More interestingly, the underlying microstructure and force networks bear surprising similarities, which can be used to understand many sometimes puzzling non-linear flow phenomena.

A particular example is shear thickening, where suspension viscosity increases with shear stress or shear rate. It is a ubiquitous feature of many different particle systems and flow processes in nature and industry, particularly for particles of intermediate sizes (diameter 1 μm ≤  d ≤ 50 μm) at high concentrations. This phenomenon has been traditionally explained as being driven purely by hydrodynamic interactions. However, recent theoretical, modelling and experimental work has shown the inadequacy of this mechanism, by elucidating the important role of frictional particle contact.

 In this talk, I will present experimental and simulation evidence for shear thickening as a transition from a typical colloidal to a granular behaviour, in which the formation of frictional non-hydrodynamic contacts is key. We have directly quantified the contact contribution to the suspension viscosity during shear thickening by means of shear-reversal rheological measurements, providing new and unambiguous evidence. I will finally describe the ‘tuning’ of shear thickening and other rheological properties based on the understanding of particle contacts, by means of active control or suspension formulation.

A First Order Conservation Law Framework For Large Strain Thermo-elasticity

Group: Glasgow Computational Engineering Centre
Speaker: Chun Hean Lee, University of Glasgow
Date: 18 February, 2019
Time: 16:00 - 17:00
Location: Rankine Building, Room 816

Current commercial codes (e.g. PAM-CRASH, ANSYS Explicit, LS-DYNA, ABAQUS, HyperCrash) used in industry for the simulation of large scale solid mechanics problems (e.g. crash, contact, impact, fracture) are typically based on the use of second order displacement based Finite Element (FE) formulations. However, these traditional formulations suffer from well known short- comings. Some of these include (1) numerical artefacts due to shear and/or volumetric locking, (2) reduced convergence for strains and stresses in comparison with displacements, (3) high frequency noise in the vicinity of shocks and (4) pressure checker-boarding in near incompressibility. There exist in literature several (ad-hoc) approaches that are now implementable in existing FE solvers  to alleviate, if not eliminate, these shortcomings, without resorting to increased number of degrees of freedom for the problem or use of higher-order interpolation elements.

In this talk, a hyperbolic system of first order conservation laws for solid dynamics will be presented [1,2]. The linear momentum p conservation equation will be solved in conjunction  with three geometric conservation equations for the minors of the deformation tensor, namely the deformation gradient tensor F (fibre map), its co-factor H (area map) and its Jacobian J (volume map).  In order to account for thermal effects,  the total energy conservation law  E  (also known  as first law of thermodynamics) will beincorporated to the set of physical laws described above. For closure of the system,  a polyconvex model will be employed guaranteeing the existence of  real wave speeds, and thus material stability.The formulation has an eye on bridging the gap between Computational Fluid Dynamics (CFD) and large strain solid dynamics, with its tailor-made implementation from scratch within the modern CFD code  “OpenFOAM”.

From the spatial discretisation viewpoint, a vertex centred finite volume method will be utilised. Naturally, discontinuity of the conservation variables across (dual) control volume interfaces leads to a Riemann problem, whose approximate solution will be derived by means of an acoustic Riemann solver. Taking advantage of the conservation formulation for (thermo-elastic) solids, well-established shock capturing  algorithm will also be used to greatly enhance the performance  of  the algorithm in the vicinity of severe (thermal) shocks. A number of benchmark tests will be provided in order to demonstrate the robustness and applicability of the proposed methodology.  In this case, both stresses and temperature converge at the same rate as velocities and displace- ments. This is in clear contrast to the classical displacement based formulations where derived variables (e.g. stresses and strains) converge at one order below the rate of displacements. For completeness, comparisons with other in-house finite volume [3] or particle based   methodologies [4] will be provided.

References

[1]  C. H. Lee, A. J. Gil, J. Bonet, Development of a stabilised Petrov-Galerkin formulation for conservation laws in Lagrangian fast solid dynamics, CMAME 268 (2014)40–64.

[2]  A. J. Gil, C. H. Lee, J. Bonet, R. Ortigosa, A first order hyperbolic framework for large strain computational solid dynamics. Part II: Total Lagrangian compressible, nearly incompressible and truly incompressible elasticity, CMAME 300 (2016)146–181.

[3]  J. Haider, C. H. Lee, A. J. Gil,  A. Huerta,  J.  Bonet, An  upwind  cell centred  Total  Lagrangian finite volume algorithm for nearly incompressible explicit fast solid dynamic applica- tions,  CMAME 340 (2018)  684–727.

[4]  C. H. Lee, A. J. Gil, A. Ghavamian, J. Bonet, A Total Lagrangian upwind SPH algorithm for large strain explicit solid dynamics, CMAME 344 (2019)209–250.

Recursive Bayesian methods for parameter estimation and aspects of identifiability

Group: Glasgow Computational Engineering Centre
Speaker: Sanjay Pant, Swansea University
Date: 08 February, 2019
Time: 11:00 - 12:00
Location: Rankine Building, Room 630

Patient-specific computational modelling of biomedical systems requires estimation of a large number of model parameters. This estimation is typically performed through clinical measurements, which are inherently uncertain, acquired in the patient. In recent years Kalman-filtering type methods, which can be viewed as recursive Bayesian methods or data assimilation techniques, have gained popularity for this purpose. This talk will present the formulation and application of such methods. Particular focus will on the the widely used Unscented Kalman Filter, which provides computational efficiency. Applications will be presented on both lumped-parameter and geometric multi-scale models of haemodynamics. Two pathophysiologies of congenital heart disease will be of concern: i) coarctation of the aorta; and ii) hypoplastic left heart syndrome (single-ventricle physiology). Finally, some aspects of parameter identifiability, the question of whether parameters can be identified/estimated given a set of measurements, will be presented.

Translating computer simulation towards a clinical centre: Patient-Specific models for Planning Cardiovascular procedures

Group: Glasgow Computational Engineering Centre
Speaker: Claudio Capelli, University College London & Great Ormond Street Hospital
Date: 07 February, 2019
Time: 16:00 - 17:00
Location: Rankine Building, Room 630

Patient-specific computational models have been extensively developed over the last decade and applied to investigate a wide range of the cardiovascular mechanics. Modelling can also offer support to personalized and predictive medicine. Such vision could be particularly suitable to face the wide variety of heart disease. The translation of these technologies into clinical applications, however, is still far from becoming a standard of care in clinical practice and currently limited to few single cases. This talk reports the experience of a single clinical and engineering centre, based in the main UK children hospital, which has been involved in the development of a modelling framework that allows the use of realistic simulations to prospectively support clinical decisions.

After introducing the numerical methods used in this study including finite element analyses (FEA), computational fluid-dynamics (CFD) and fluid-structure interaction (FSI), I will focus on a cohort of patients who were referred for various intervention and how the planning was supported by integrating computational simulations within a clinical workflow. Image data routinely acquired for clinical assessment (MRI, CT, echocardiography, x-ray) were postprocessed to set up patient-specific models. FEA and CFD were performed to predict structural and haemodynamic changes following the procedures. Simulations were carried out to: select the best-matching device for each anatomy; address the risks of spatial interference with surrounding structures; optimize size and positions of device; design a surgical patch. The results were presented during clinical unit’s multidisciplinary meeting. Measurable clinical outcomes from the real procedures were compared with the computer model predictions.

The numerical results of FEA and CFD analyses were in accordance with the delivered treatment in all cases except in one case. Post-procedural images were also used to confirm correct prediction of sizing and positioning of the stent. Pressure and velocity data acquired by transthoracic echocardiography showed agreement with the results calculated with CFD analyses with a max error less than 3 mmHg. Each computational framework process was completed within a week with no requirements for additional clinical data.

The early results of using computer simulations in clinics seem to be promising in terms of reliability of the simulations, response time, and usefulness in clinical practice. The translation of these technologies is crucial as it can limit the procedural risks for treatment of CHD cases.

A fresh look at nonlinear computation

Group: Glasgow Computational Engineering Centre
Speaker: Ankush Aggarwal
Date: 04 December, 2018
Time: 16:00 - 17:00
Location: Rankine Building, Room 816

The initial developments in the field of computational engineering were based on linear systems. As the scope of computational methods expanded, their application to highly nonlinear problems, such as biomechanical, became more commonplace. The common approach of linearizing the nonlinear equations locally and solving incrementally via continuation can be extremely slow.
In this talk, I will start with an overview of existing computational solvers. I will explore the possibility of taking steps before linearization with the aim of improving the simulation speed and accuracy. Based on this central idea, I will describe some new techniques to solve forward and inverse nonlinear problems and compare the results for simple cases. I will end with some speculative ideas on extending this approach to other scenarios.

A matrix-free approach for finite-strain hyperelastic problems using geometric multigrid

Group: Glasgow Computational Engineering Centre
Speaker: Dr Denis Davydov, University of Erlangen-Nuremberg
Date: 07 November, 2018
Time: 15:00 - 16:00
Location: Rankine Building, Room 816

The performance of finite element solvers on modern computer architectures is typically memory bound. The main cause for this is that loading matrix elements from RAM into CPU cache is significantly slower than performing the arithmetic operations when solving the problem. In order to improve the performance of iterative solvers within the high-performance computing context, so-called matrix-free methods are widely adopted in the fluid mechanics community, where matrix-vector products are computed on-the-fly.

In this work, we extend the application of matrix-free approaches to problems in solid mechanics and investigate different implementations of the finite-strain hyperelastic tangent operator numerically. In order to improve the convergence behavior of iterative solvers, we also propose a method by which to construct level tangent operators and employ them to define a geometric multigrid preconditioner.

Our implementation employs MPI and Intel Threading Building Blocks parallelization, and SIMD vectorization.
The performance of the matrix-free operator and the geometric multigrid preconditioner is compared to the matrix-based implementation with an algebraic multigrid preconditioner for a representative numerical example of a heterogeneous hyperelastic material in two and three dimensions.

Experimental and numerical investigations of cell reorientation

Group: Glasgow Computational Engineering Centre
Speaker: Pradeep Keshavanarayana, GCEC
Date: 17 October, 2018
Time: 14:00 - 15:00
Location: Rankine Building, Room 816

Cells, the fundamental unit of living organisms experience a variety of forces within a living body. It has been observed that their responses depend on the mechanical stimuli they experience. Chemical reactions occurring within the cell lead to the growth of actin stress fibres. They usually terminate at the focal adhesions, which are present on the cell membrane and can sense the properties of the extracellular matrix. Experiments have shown that cells reorient away from the direction of loading when a uniaxial cyclic strain is applied. In this seminar, the mathematical model developed to explain this phenomenon, coupling focal adhesion and stress fibre growth, along with the solution schemes followed will be discussed. The device that can apply the required cyclic loading to the cell, cell stretcher, which has been built using the 3D printing technology will be demonstrated. Experimental observations of cell reorientation will be compared with results from numerical modelling.

Enriched simulations in computational mechanics

Group: Glasgow Computational Engineering Centre
Speaker: Prof. Jon Trevelyan, Durham University
Date: 26 September, 2018
Time: 15:00 - 16:00
Location: Rankine Building, Room 514

Classical finite element and boundary element methods seek the solution to a boundary-value problem in a space of piecewise polynomial functions. There are some classes of problems for which this approach leads to inefficiencies, sometimes so severe that the problem becomes infeasible. In these cases, we can make progress by using functions other than the conventional polynomials, the choice of these functions being informed by our knowledge of the form of the solution. In this seminar we will look at the solution to a range of problems in fracture mechanics, wave propagation and heat transfer, showing how accurate solutions may be obtained on coarse meshes.