Reducing the Mesh-burden and Computational Expense in Multi-scale Free Boundary Engineering Problems

Presentation (2014, May 12)

We present recent results aiming at affording faster and error-controlled simulations of multi scale phenomena including fracture of heterogeneous materials and cutting of biological tissue. In a second ... [more ▼]

We present recent results aiming at affording faster and error-controlled simulations of multi scale phenomena including fracture of heterogeneous materials and cutting of biological tissue. In a second part, we describe methodologies to isolate the user from the burden of mesh generation and regeneration as moving boundaries evolve. Results include advances in implicit boundary finite elements, (enriched) isogeometric boundary elements and extended finite element methods for multi-crack propagation. ABOUT THE PRESENTER In 1999, Stéphane Bordas joined a joint graduate programme of the French Institute of Technology (Ecole Spéciale des Travaux Publics) and the American Northwestern University. In 2003, he graduated in Theoretical and Applied Mechanics with a PhD from Northwestern University. Between 2003 and 2006, he was at the Laboratory of Structural and Continuum Mechanics at the Swiss Federal Institute of Technology in Lausanne, Switzerland. In 2006, he became permanent lecturer at Glasgow University’s Civil Engineering Department. Stéphane joined the Computational Mechanics team at Cardiff University in September 2009, as a Professor in Computational Mechanics and directed the institute of Mechanics and Advanced Materials from October 2010 to November 2013. He is the Editor of the book series “Advances in Applied Mechanics” since July 2013. In November 2013, he joined the University of Luxembourg as a Professor in Computational Mechanics. The main axes of his research team include (1) free boundary problems and problems involving complex geometries, in particular moving boundaries and (2) ‘a posteriori’ discretisation and model error control, rationalisation of the computational expense. Stéphane’s keen interest is to actively participate in innovation, technological transfer as well as software tool generation. This has been done through a number of joint ventures with various industrial partners (Bosch GmbH, Cenaero, inuTech GmbH, Siemens-LMS, Soitec SA) and the release of open-source software. In 2012, Stéphane was awarded an ERC Starting Independent Research Grant (RealTcut), to address the need for surgical simulators with a computational mechanics angle with a focus on the multi-scale simulation of cutting of heterogeneous materials in real-time. [less ▲]

From image to analysis: an extended finite element method to simulate the mechanical response of soft-tissue

Presentation (2014, April 10)

In this seminar we consider the problem of constructing a numerical method particularly well suited to modelling domains described by segmented image data of the human body. Instead of constructing a ... [more ▼]

In this seminar we consider the problem of constructing a numerical method particularly well suited to modelling domains described by segmented image data of the human body. Instead of constructing a conforming mesh of the problem domain, we use implicitly defined domains described using the level-set method. We then include information about the implicitly defined domains by enriching the usual finite element basis functions defined on a cartesian quadtree or octree mesh with hanging nodes. This approach introduces significant complexities compared with classical finite element methods. We will discuss difficulties with the treatment of hanging nodes, numerical integration and imposing Dirichlet boundary conditions. We will discuss the possible future of extensions of this work, including cutting of soft tissue, multiscale problems with complex microstructure, and model order reduction problems. [less ▲]

An enriched quadtree/octree implicit boundary finite element method for the simulation of incompressible hyperelastic materials

Scientific Conference (2014, April 03)

Octree (and quadtree) representations of computational geometry are particularly well suited to modelling domains that are defined implicitly, such as those generated by image segmentation algorithms ... [more ▼]

Octree (and quadtree) representations of computational geometry are particularly well suited to modelling domains that are defined implicitly, such as those generated by image segmentation algorithms applied to medical scans. In this work we consider the simulation of soft-tissue which can be modelled with a hyperelastic constitutive law. We include the effects of both non-linear geometry and material properties. Similarly to Legrain et al. (10.1002/nme.3070) and Moumnassi et al. (10.1016/j.cma.2010.10.002) we use the implicitly designed level set functions as the basis for a partition of unity enrichment to more accurately represent the domain boundary. Furthermore we use traditional extended finite element (XFEM) ideas to introduce arbitrary cuts and discontinuities in the domain. We explore the use of a two-field u-p mixed approach as well as a smoothed finite element method (SFEM) to deal with the problem of volumetric-locking in the incompressible limit. We will discuss the extension of our method towards both traditional parallel and GPU implementation. We aim to solve extremely large problems as well as produce snapshots to feed into model order reduction methods for real-time surgical simulations. [less ▲]

A Meshless Method for the Reissner-Mindlin Plate Equations based on a Stabilized Mixed Weak Form

Scientific Conference (2013, September)

Towards Effective Shell Modelling with the FEniCS Project

Scientific Conference (2013, March)

Fast and efficient simulations of shell structures are required in a wide range of engineering fields such as fluid-structure interaction and structural optimisation. Because of its expressive high-level ... [more ▼]

Fast and efficient simulations of shell structures are required in a wide range of engineering fields such as fluid-structure interaction and structural optimisation. Because of its expressive high-level form language UFL the FEniCS project is in an ideal position to tackle tough problems such as large deformations of non-isotropic shells. In this talk we will discuss some aspects of achieving this goal; generalised mixed formulations, reduction and projection operators for eliminating shear and membrane locking, the general shell model vs classical models and the recent work by Rognes et al. on manifolds. [less ▲]

Maximum-Entropy Meshfree Method for the Reissner-Mindlin Plate Problem based on a Stabilised Mixed Weak Form

Scientific Conference (2012)

Meshless methods, such as the Element Free Galerkin (EFG) method, hold various advantages over mesh-based techniques such as robustness in large-deformation problems and high continuity. The Reissner ... [more ▼]

Meshless methods, such as the Element Free Galerkin (EFG) method, hold various advantages over mesh-based techniques such as robustness in large-deformation problems and high continuity. The Reissner-Mindlin plate model is a particularly popular choice for simulating thin structures. It is well known in the Finite Element and Meshless literature that the simplest numerical treatments of the Reissner-Mindlin model lead to shear-locking which in turn produces erroneous results. This is due to the inability of the approximation functions to satisfy the Kirchoff constraint in the thin-plate limit. A recent advance in the area of meshless approximation schemes are Maximum-Entropy (MaxEnt) approximants. MaxEnt schemes provide a weak Kronecker-delta property on convex node sets which allows the direct imposition of Dirichlet (essential) boundary conditions. In this work, we derive a shear-locking free meshless method using MaxEnt approximants by consider- ing a stabilised mixed weak form. We include a scalar parameter which splits the energy from the shear bilinear form into two parts; the first is formed from the displacement fields only and the second from the independently interpolated shear strain field and the displacement fields. This splitting greatly eases the satisfaction of the LBB stability condition. We then eliminate the independently interpolated shear strain field using a localised projection operator, related to the “volume-averaged pressure” technique, which produces a final system of equations in the original displacement unknowns only. We show the good performance of the method for a variety of test problems. [less ▲]

Simulation of Shear Deformable Plates using Meshless Maximum Entropy Basis Functions

Scientific Conference (2011, June)

First-order Shear Deformable Plate Theory (FSDT) is widely used throughout engineering practice to simulate structures with planar dimensions much larger than their thickness. Meshless methods have seen ... [more ▼]

First-order Shear Deformable Plate Theory (FSDT) is widely used throughout engineering practice to simulate structures with planar dimensions much larger than their thickness. Meshless methods have seen use in the literature as a method for discretising the FSDT equations and hold numerous advantages over traditional mesh based techniques. A recent advance in the area of meshless methods are Maximum Entropy approximants (MaxEnt). MaxEnt combines many properties of various prior meshless approximants such as a weak Kronecker-delta property, seamless blending with Delaunay triangulations, high continuity, and convexity. In this work MaxEnt along with other meshless approximants have been implemented in a hybrid object-oriented Python/C++/Fortran computer simulation for the simulation of static deflection, free vibration and linear buckling of FSDT plates. The relative performance and ease of implementation of each of the methods will be discussed. The causes of shear locking along with the merits of various alleviation techniques will be covered, including matching fields method, mixed-variational formulations and construction of higher order polynomial basis via both intrinsic and extrinsic (partition of unity) methods. Convergence results show that MaxEnt provides in most cases similar and in some cases superior behaviour to MLS and RPIM approximants when used to discretise the FSDT equations. [less ▲]

Mathematical Modelling of Flux Decline due to Concentration Polarisation and Cake Layer Formation in Crossflow Filtration Systems

Scientific Conference (2007, May)

Crossflow membrane filtration is an effective way of removing both colloidal and dissolved organic matter from contaminated water supplies. Two phenomena domimate solute flux in crossflow systems ... [more ▼]

Crossflow membrane filtration is an effective way of removing both colloidal and dissolved organic matter from contaminated water supplies. Two phenomena domimate solute flux in crossflow systems; concentration polarization and cake layer formation. Many innovative mathematical models for predicting both flux decline and quasi-steady state flux have been produced in the literature. However, limited regime applicability and conflicting physical predictions have made choosing optimal performance parameters in design a challenging process. An overview of the current field of models was undertaken, including an assessment of mathematical assumptions, numerical computation workload and number and complexity of system constants. A new model incorporating viscosity dependence on concentration is developed for the concentration polarization regime. Preliminary results will also be presented from a Monte Carlo based molecular dynamics simulation of volume packing fraction in the cake layer regime. The novel aspects of these models will be compared with existing models and the experimental results of our collaborators. Laboratory ultrafiltration tests were undertaken using varying concentrations of Dextran. An outline of future model directions and refinements will be presented. [less ▲]

Bayesian inference for the stochastic identification of elastoplastic material parameters: Introduction, misconceptions and insights

E-print/Working paper (n.d.)

We discuss Bayesian inference (BI) for the probabilistic identification of material parameters. This contribution aims to shed light on the use of BI for the identification of elastoplastic material ... [more ▼]

We discuss Bayesian inference (BI) for the probabilistic identification of material parameters. This contribution aims to shed light on the use of BI for the identification of elastoplastic material parameters. For this purpose a single spring is considered, for which the stress-strain curves are artificially created. Besides offering a didactic introduction to BI, this paper proposes an approach to incorporate statistical errors both in the measured stresses, and in the measured strains. It is assumed that the uncertainty is only due to measurement errors and the material is homogeneous. Furthermore, a number of possible misconceptions on BI are highlighted based on the purely elastic case. [less ▲]