Expressive Environments and Code Generation for High Performance Computing
|
|
- Elaine Estella Johnson
- 5 years ago
- Views:
Transcription
1 Expressive Environments and Code Generation for High Performance Computing Garth N. Wells University of Cambridge SIAM Parallel Processing, 20 February 2014
2 Collaborators Martin Alnæs, Johan Hake, Richard Katz, Anders Logg, Marie Rognes, Sander Rhebergen, Andy Wathen, Kristian Ølgaard
3 Scientific software SUBROUTINE DROTG(DA,DB,C,S) DOUBLE PRECISION C,DA,DB,S DOUBLE PRECISION R,ROE,SCALE,Z INTRINSIC DABS,DSIGN,DSQRT ROE = DB IF (DABS(DA).GT.DABS(DB)) ROE = DA SCALE = DABS(DA) + DABS(DB) IF (SCALE.EQ.0.0d0) THEN C = 1.0d0 S = 0.0d0 R = 0.0d0 Z = 0.0d0 ELSE R = SCALE*DSQRT((DA/SCALE)**2+ (DB/SCALE)**2) R = DSIGN(1.0d0,ROE)*R C = DA/R S = DB/R Z = 1.0d0 IF (DABS(DA).GT.DABS(DB)) Z = S IF (DABS(DB).GE.DABS(DA).AND. C.NE.0.0d0) Z = 1.0d0/C END IF DA = R DB = Z RETURN END
4 Be more expressive. Why? Simpler development Accelerated development Faster code with acceptable effort Detach mathematical intent from underlying algorithms, implementations and hardware User accessibility
5 Outline Domain-specific language for variational forms Code generator for variational forms Solver environment Illustrative application: block-preconditioned systems All presented tools are part of the FEniCS Project ( ) - all available under a free license Development site:
6 Unified Form Language (UFL) A domain-specific embedded language for variational forms Embedded in Python Highly expressive Mathematical error checking Implements various form manipulations Does not do any non-trivial computation Provides abstract representation (DAG) of a variational problem Requires a backend to generate concrete code Alnæs, Logg, Ølgaard, Rognes, Wells (2014) ACM TOMS,
7 Stokes equations Find where u, p V Q such that: a({u, p}, {v, q}) = L({v, q}) v, q V Q a L D D := u : v p v + ( u)q dx := f v dx
8 Stokes equations in the Unified Form Language V = VectorElement("Lagrange", triangle, 2) Q = FiniteElement("Lagrange", triangle, 1) TH = V * Q (u, p) = TrialFunctions(TH) (v, q) = TestFunctions(TH) f = Coefficient(V) a = (inner(grad(u), grad(v)) - p*div(v) + div(u)*q)*dx L = dot(f, v)*dx
9 Hyperelasticity Displacement field where u given by: u = argmin v V Π(v) Π := ψ(e(v)) B u dx T v ds ψ(e) E := ( F := Ω Ω is the strain energy density F T F I)/2 is the Green-Lagrange strain is the deformation gradient X v + I
10 Hyperelasticity in UFL V = VectorElement("Lagrange", "tetrahedron", 1) # Current displacement u = Coefficient(V).. # Kinematics (deformation gradient, right Cauchy-Green tensor) F, C = Identity(V.cell().d) + grad(u), F.T*F # Invariants of deformation tensors J, Ic = det(f), tr(c) # Stored strain energy density (compressible neo-hookean model) psi = (mu/2)*(ic - 3) - mu*ln(j) + (lmbda /2)*(ln(J))**2 # Total potential energy Pi = psi*dx - dot(b, u)*dx - dot(t, u)*ds # First variation of Pi v = TestFunction(V) F = derivative(pi, u, v) # Compute Jacobian of F (Hessian of Pi) du = TrialFunction(V) a = derivative(f, u, du)
11 Form to directed acyclic graph (1)
12 Form to directed acyclic graph (2)
13 Code generator (FEniCS Form Compiler) Code generator takes abstract representation and generates code in target language Kirby, Logg (2005) ACM TOMS; Ølgaard, Wells (2010) ACM TOMS; Logg, Ølgaard, Rognes, Wells (2012), FEniCS Book.
14 Domain-specific code optimisations Ølgaard & Wells (2010) ACM TOMS,
15 Problem solving environment DOLFIN is the FEniCS Project problem solving environment Synthesises domain-specific language, code generation, linear algebra, domain representation (mesh),... C++ and Python interfaces (implemented primarily in C++) Design reflects mathematical abstractions Logg and Wells (2010) ACM TOMS,
16 DOLFIN example from dolfin import * # Create mesh and define function space mesh = UnitSquareMesh(32, 32) V = FunctionSpace(mesh, "Lagrange", 1) # Define Dirichlet boundary (x = 0 or x = 1) def boundary(x): return x[0] < DOLFIN_EPS or x[0] > DOLFIN_EPS # Define boundary condition bc = DirichletBC(V, 0.0, boundary) # Define variational problem u, v = TrialFunction(V), TestFunction(V) f = Expression("10*exp(-(pow(x[0] - 0.5, 2) + pow(x[1] - 0.5, 2)) / 0.02)") a, L = inner(grad(u), grad(v))*dx, f*v*dx # Compute solution u = Function(V) solve(a == L, u, bc)
17 Mixed language development Language has large impact on productivity Low-level, performance necessary language will inevtitably not be suitable at the user/solver level Language has large impact on (potential) user base
18 Melt migration model Simplified model: ϵ(u) + p α u u k p where ϵ = ( u + u T )/2, α 0 and k 0 = 0 = 0
19 Block precondition a la Stokes: [ A 0 ] 0 Q + C 1 [ A ][ u] = [ A 0 ] [ f] B C p 0 Q + C g B T Code representing operator and preconditioner forms a = inner(sym(grad(u)), sym(grad(v)))*dx + alpha*div(u)*div(v)*dx \ - p*div(v)*dx - q*div(u)*dx - k*dot(grad(q), grad(p))*dx a_pc = inner(sym(grad(u)), sym(grad(v)))*dx + alpha*div(u)*div(v)*dx \ + p*q + k*dot(grad(q), grad(p))*dx 1 Identifying subspaces/matrix blocks TH = MixedFunctionSpace([V, Q]) V0 = TH.sub(0) # Extract first subspace V1 = TH.sub(1) # Extract second subspace
20 Solving the Stokes-like problem Iteration count using PETSc FieldSplit and MINRES N α = 1 α = 1 α = 100 α = 100 α = 1000 α = 100 LU AMG LU AMG LU AMG The devil is the grad - div term! Rhebergen, et al.
21 Modified problem formulation Introduce compaction pressure 1 ϵ(u) + p + + ( f p c u) 3 u + k p f u ξp c = 0 = 0 = 0 Code representing operator and preconditioner forms a = inner(sym(grad(u)), sym(grad(v)))*dx - (1.0/3.0)*div(u)*div(v)*dx \ - pf*div(v)*dx - pc*div(v)*dx - q*div(u)*dx \ - kappa*dot(grad(pf), grad(q))*dx - xi*pc*w*dx - w*div(u)*dx a_pc = inner(sym(grad(u)), sym(grad(v)))*dx \ + pf*q*dx + kappa*dot(grad(q), grad(pf))*dx \ + (xi + 1.0)*pc*w*dx
22 Solving the three-field problem Iteration count using AMG on the blocks + MINRES N α = 0 α = 1 α = New block preconditioners can created and tested in minutes
23 Summary Domain-specific approaches can accelerate development Detach mathematical intent from underlying algorithm/hardware Preserve mathematical abstractions Dynamic and configurable
FEniCS Course. Lecture 1: Introduction to FEniCS. Contributors Anders Logg André Massing
FEniCS Course Lecture 1: Introduction to FEniCS Contributors Anders Logg André Massing 1 / 15 What is FEniCS? 2 / 15 FEniCS is an automated programming environment for differential equations C++/Python
More informationFinite Element Code Generation: Simplicity, Generality, Efficiency
: Simplicity, Generality, Efficiency Simula Research Laboratory SCSE 07 Uppsala, August 15 2007 Acknowledgments: Martin Sandve Alnæs, Johan Hoffman, Johan Jansson, Claes Johnson, Robert C. Kirby, Matthew
More informationThe FEniCS Project. Anders Logg. Simula Research Laboratory / University of Oslo. Det Norske Veritas
The FEniCS Project Anders Logg Simula Research Laboratory / University of Oslo Det Norske Veritas 2011 05 11 The FEniCS Project Free Software for Automated Scientific Computing C++/Python library Initiated
More informationFEniCS Course. Lecture 0: Introduction to FEM. Contributors Anders Logg, Kent-Andre Mardal
FEniCS Course Lecture 0: Introduction to FEM Contributors Anders Logg, Kent-Andre Mardal 1 / 46 What is FEM? The finite element method is a framework and a recipe for discretization of mathematical problems
More informationThe FEniCS Project Philosophy, current status and future plans
The FEniCS Project Philosophy, current status and future plans Anders Logg logg@simula.no Simula Research Laboratory FEniCS 06 in Delft, November 8-9 2006 Outline Philosophy Recent updates Future plans
More informationFlexible Specification of Large Systems of Nonlinear PDEs
Flexible Specification of Large Systems of Nonlinear PDEs Hans Petter Langtangen 1,2 Mikael Mortensen 3,1 Center for Biomedical Computing Simula Research Laboratory 1 Dept. of Informatics, University of
More informationThe FEniCS Project. Anders Logg. Simula Research Laboratory University of Oslo NOTUR / 34
The FEniCS Project Anders Logg Simula Research Laboratory University of Oslo NOTUR 2011 2011 05 23 1 / 34 What is FEniCS? 2 / 34 FEniCS is an automated programming environment for differential equations
More informationFEniCS Course. Lecture 1: Introduction to FEniCS. Contributors Anders Logg André Massing
FEniCS Course Lecture 1: Introduction to FEniCS Contributors Anders Logg André Massing 1 / 30 What is FEniCS? 2 / 30 FEniCS is an automated programming environment for differential equations C++/Python
More informationGETTING STARTED WITH FENICS
GETTING STARTED WITH FENICS DOUGLAS N. ARNOLD 1. A first program in FEniCS 1.1. The boundary value problem. As a first problem we consider the Neumann boundary value problem: (1) u + u = f in, u = 0 on,
More informationSolving PDEs in Minutes - The FEniCS Tutorial Volume I
Hans Petter Langtangen, Anders Logg Solving PDEs in Minutes - The FEniCS Tutorial Volume I Sep 19, 2016 Springer Email: hpl@simula.no. Center for Biomedical Computing, Simula Research Laboratory and Department
More informationFEniCS, part IV: Generating a model
FEniCS, part IV: Generating a model M. M. Sussman sussmanm@math.pitt.edu Office Hours: 11:10AM-12:10PM, Thack 622 May 12 June 19, 2014 1 / 54 Topics Vortex shedding Background Implementation 2 / 54 Topics
More informationvon Kàrmàn vortex street FEniCS, part IV: Generating a model Clouds Discussed in FEniCS book Fluid flow past an obstacle is not steady
von Kàrmàn vortex street FEniCS, part IV: Generating a model Fluid flow past an obstacle is not steady M. M. Sussman sussmanm@math.pitt.edu Office Hours: 11:10AM-12:10PM, Thack 622 May 12 June 19, 2014
More informationFEniCS, part III M. M. Sussman sussmanm@math.pitt.edu Office Hours: 11:10AM-12:10PM, Thack 622 May 12 June 19, 2014 1 / 67 Topics Geometry Elements Nonlinear problems Transient problems Eigenvalue problems
More informationHow do I generate complicated meshes? FEniCS, part III. Mesh mapping. Mesh mapping cont d
How do I generate complicated meshes? FEniCS, part III M. M. Sussman sussmanm@math.pitt.edu Office Hours: 11:10AM-12:10PM, Thack 622 May 12 June 19, 2014 UnitIntervalMesh, UnitSquareMesh and UnitCubeMesh
More informationSolving PDEs in Python The FEniCS Tutorial I
SIMULA SPRINGER BRIEFS ON COMPUTING 3 Hans Petter Langtangen Anders Logg Solving PDEs in Python The FEniCS Tutorial I Simula SpringerBriefs on Computing Volume 3 Editor-in-chief Aslak Tveito, Fornebu,
More informationFEniCS Course. Lecture 8: A posteriori error estimates and adaptivity. Contributors André Massing Marie Rognes
FEniCS Course Lecture 8: A posteriori error estimates and adaptivity Contributors André Massing Marie Rognes 1 / 24 A priori estimates If u H k+1 (Ω) and V h = P k (T h ) then u u h Ch k u Ω,k+1 u u h
More informationIntroduction to Automated Modeling using FEniCS
Introduction to Automated Modeling using FEniCS L. Ridgway Scott University of Chicago Release 0.1 DO NOT DISTRIBUTE February 7, 2017 This page will contain copyright information in due course. Contents
More informationAutomated Modeling with FEniCS
Automated Modeling with FEniCS L. Ridgway Scott The Institute for Biophysical Dynamics, The Computation Institute, and the Departments of Computer Science and Mathematics, The University of Chicago UChicago
More informationShape Optimization Tutorial
Shape Optimization Tutorial By Stephan Schmidt Exercise 1. The first exercise is to familiarise with Python and FEniCS. We can use the built-in FEniCS tutorial to implement a small solver for the Laplacian
More informationSolving Non-linear Elasticity BVPs in FEniCS
Solving Non-linear Elasticity BVPs in FEniCS Bifurcations and Buckling of Beams Christiana Mavroyiakoumou University of Oxford A special topic report submitted for the degree of M.Sc. in Mathematical Modelling
More informationIntroduction to Automated Modeling using FEniCS
Introduction to Automated Modeling using FEniCS L. Ridgway Scott University of Chicago Release 0.3 DO NOT DISTRIBUTE October 2, 2017 This page will contain copyright information in due course. Contents
More informationLectures on the Finite Element Method
Anders Logg Kent-Andre Mardal Editors Lectures on the Finite Element Method Contents 1 The finite element method 1 1.1 A simple model problem..................................... 1 1.2 Solving Poisson
More informationAutomated Solution of Differential Equations by the Finite Element Method
Lecture Notes in Computational Science and Engineering 84 Automated Solution of Differential Equations by the Finite Element Method The FEniCS Book Bearbeitet von Anders Logg, Kent-Andre Mardal, Garth
More informationFEniCS Course Lecture 13: Introduction to dolfin-adjoint. Contributors Simon Funke Patrick Farrell Marie E. Rognes
FEniCS Course Lecture 13: Introduction to dolfin-adjoint Contributors Simon Funke Patrick Farrell Marie E. Rognes 1 / 25 dolfin adjoint automatic adjoint models for FEniCS P. E. Farrell, S. W. Funke, D.
More informationDiscontinuous Galerkin methods for nonlinear elasticity
Discontinuous Galerkin methods for nonlinear elasticity Preprint submitted to lsevier Science 8 January 2008 The goal of this paper is to introduce Discontinuous Galerkin (DG) methods for nonlinear elasticity
More informationIntroduction to Numerical Methods for Variational Problems
Introduction to Numerical Methods for Variational Problems Hans Petter Langtangen 1,2 Kent-Andre Mardal 3,1 1 Center for Biomedical Computing, Simula Research Laboratory 2 Department of Informatics, University
More informationIntroduction to Numerical Methods for Variational Problems
Introduction to Numerical Methods for Variational Problems Hans Petter Langtangen 1,2 Kent-Andre Mardal 3,1 1 Center for Biomedical Computing, Simula Research Laboratory 2 Department of Informatics, University
More informationSolving an elasto-plastic model using DOLFIN
Solving an elasto-plastic model using DOLFIN TTI 2005 Johan Jansson johanjan@math.chalmers.se Chalmers University of Technology Solving an elasto-plastic model using DOLFIN p. Overview Motivation Previous
More informationA FEniCS Tutorial (November 16, 2009)
A FEniCS Tutorial (November 16, 2009) H. P. Langtangen 1,2 1 Center for Biomedical Computing, Simula Research Laboratory 2 Department of Informatics, University of Oslo This document is in a preliminary
More informationEfficient multigrid solvers for mixed finite element discretisations in NWP models
1/20 Efficient multigrid solvers for mixed finite element discretisations in NWP models Colin Cotter, David Ham, Lawrence Mitchell, Eike Hermann Müller *, Robert Scheichl * * University of Bath, Imperial
More informationMechanics of materials Lecture 4 Strain and deformation
Mechanics of materials Lecture 4 Strain and deformation Reijo Kouhia Tampere University of Technology Department of Mechanical Engineering and Industrial Design Wednesday 3 rd February, 206 of a continuum
More informationSubdomains and boundary conditions
Chapter 4 Subdomains and boundary conditions So far, we have only looked briefly at how to specify boundary conditions. In this chapter, we look more closely at how to specify boundary conditions on specific
More informationTowards effective shell modelling with the FEniCS project
Towards effective shell modelling with the FEniCS project *, P. M. Baiz Department of Aeronautics 19th March 2013 Shells and FEniCS - FEniCS Workshop 13 1 Outline Introduction Shells: chart shear-membrane-bending
More informationAutomating the Finite Element Method
Automating the Finite Element Method Anders Logg logg@tti-c.org Toyota Technological Institute at Chicago Johan Hoffman, Johan Jansson, Claes Johnson, Robert C. Kirby, Matthew Knepley, Ridgway Scott Simula
More informationInverse Design (and a lightweight introduction to the Finite Element Method) Stelian Coros
Inverse Design (and a lightweight introduction to the Finite Element Method) Stelian Coros Computational Design Forward design: direct manipulation of design parameters Level of abstraction Exploration
More informationFEniCS Course. Lecture 6: Incompressible Navier Stokes. Contributors Anders Logg André Massing
FEniCS Course Lecture 6: Incompressible Navier Stokes Contributors Anders Logg André Massing 1 / 11 The incompressible Navier Stokes equations u + u u ν u + p = f in Ω (0, T ] u = 0 in Ω (0, T ] u = g
More informationLECTURE # 0 BASIC NOTATIONS AND CONCEPTS IN THE THEORY OF PARTIAL DIFFERENTIAL EQUATIONS (PDES)
LECTURE # 0 BASIC NOTATIONS AND CONCEPTS IN THE THEORY OF PARTIAL DIFFERENTIAL EQUATIONS (PDES) RAYTCHO LAZAROV 1 Notations and Basic Functional Spaces Scalar function in R d, d 1 will be denoted by u,
More informationin this web service Cambridge University Press
CONTINUUM MECHANICS This is a modern textbook for courses in continuum mechanics. It provides both the theoretical framework and the numerical methods required to model the behavior of continuous materials.
More informationFinite Elements with Symbolic Computations and Code Generation
Finite Elements with Symbolic Computations and Code Generation August 15, 2007 Outline We will focus on creating/assembling matrices based on FEM Manual implementation of the integrand/variational form
More informationINTRODUCTION TO THE FINITE ELEMENT SOFTWARE FEniCS ROLAND HERZOG
INTRODUCTION TO THE FINITE ELEMENT SOFTWARE FEniCS ROLAND HERZOG 1 Introduction FEniCS is free software for automated solution of differential equations. As a particular feature, it allows the formulation
More informationBridging the Gap: Harnessing advanced Computational Libraries for coupled multi-physics problems
Bridging the Gap: Harnessing advanced Computational Libraries for coupled multi-physics problems Marc Spiegelman 1,2 and Cian R Wilson 1 Plus enormous contributions from PETSc, FEniCS and AMCG groups 1
More informationc 2013 Society for Industrial and Applied Mathematics
SIAM J. SCI. COMPUT. Vol. 35, No. 4, pp. C369 C393 c 2013 Society for Industrial and Applied Mathematics AUTOMATED DERIVATION OF THE ADJOINT OF HIGH-LEVEL TRANSIENT FINITE ELEMENT PROGRAMS P. E. FARRELL,
More informationFast solvers for steady incompressible flow
ICFD 25 p.1/21 Fast solvers for steady incompressible flow Andy Wathen Oxford University wathen@comlab.ox.ac.uk http://web.comlab.ox.ac.uk/~wathen/ Joint work with: Howard Elman (University of Maryland,
More informationMultigrid and Iterative Strategies for Optimal Control Problems
Multigrid and Iterative Strategies for Optimal Control Problems John Pearson 1, Stefan Takacs 1 1 Mathematical Institute, 24 29 St. Giles, Oxford, OX1 3LB e-mail: john.pearson@worc.ox.ac.uk, takacs@maths.ox.ac.uk
More informationFETI Methods for the Simulation of Biological Tissues
SpezialForschungsBereich F 32 Karl Franzens Universita t Graz Technische Universita t Graz Medizinische Universita t Graz FETI Methods for the Simulation of Biological Tissues Ch. Augustin O. Steinbach
More informationAn abstract Hodge Dirac operator and its stable discretization
An abstract Hodge Dirac operator and its stable discretization Paul Leopardi Mathematical Sciences Institute, Australian National University. For presentation at ICCA, Tartu,. Joint work with Ari Stern,
More informationAdvancing Teaching and Research with Python
Advancing Teaching and Research with Python Hans Petter Langtangen Center for Biomedical Computing (CBC) at Simula Research Laboratory Dept. of Informatics, University of Oslo EuroSciPy 2010 Teaching Research
More informationContinuum Mechanics and the Finite Element Method
Continuum Mechanics and the Finite Element Method 1 Assignment 2 Due on March 2 nd @ midnight 2 Suppose you want to simulate this The familiar mass-spring system l 0 l y i X y i x Spring length before/after
More informationFast Iterative Solution of Saddle Point Problems
Michele Benzi Department of Mathematics and Computer Science Emory University Atlanta, GA Acknowledgments NSF (Computational Mathematics) Maxim Olshanskii (Mech-Math, Moscow State U.) Zhen Wang (PhD student,
More informationCourse Outline. 2. Vectors in V 3.
1. Vectors in V 2. Course Outline a. Vectors and scalars. The magnitude and direction of a vector. The zero vector. b. Graphical vector algebra. c. Vectors in component form. Vector algebra with components.
More informationA High-Order Discontinuous Galerkin Method for the Unsteady Incompressible Navier-Stokes Equations
A High-Order Discontinuous Galerkin Method for the Unsteady Incompressible Navier-Stokes Equations Khosro Shahbazi 1, Paul F. Fischer 2 and C. Ross Ethier 1 1 University of Toronto and 2 Argonne National
More informationDeformation of bovine eye fluid structure interaction between viscoelastic vitreous, non-linear elastic lens and sclera
Karel October Tůma 24, Simulation 2018 of a bovine eye 1/19 Deformation of bovine eye fluid structure interaction between viscoelastic vitreous, non-linear elastic lens and sclera Karel Tůma 1 joint work
More informationIntegrating Finite Element Analysis with Systems Engineering Models
Integrating Finite Element Analysis with Systems Engineering Models KONEKSYS Jerome Szarazi, Axel Reichwein July 26, 2016 This work was performed under the following financial assistance award NIST Grant
More informationIncomplete Cholesky preconditioners that exploit the low-rank property
anapov@ulb.ac.be ; http://homepages.ulb.ac.be/ anapov/ 1 / 35 Incomplete Cholesky preconditioners that exploit the low-rank property (theory and practice) Artem Napov Service de Métrologie Nucléaire, Université
More informationConstitutive models. Constitutive model: determines P in terms of deformation
Constitutive models Constitutive model: determines P in terms of deformation Elastic material: P depends only on current F Hyperelastic material: work is independent of path strain energy density function
More informationEfficient FEM-multigrid solver for granular material
Efficient FEM-multigrid solver for granular material S. Mandal, A. Ouazzi, S. Turek Chair for Applied Mathematics and Numerics (LSIII), TU Dortmund STW user committee meeting Enschede, 25th September,
More informationA Robust Preconditioned Iterative Method for the Navier-Stokes Equations with High Reynolds Numbers
Applied and Computational Mathematics 2017; 6(4): 202-207 http://www.sciencepublishinggroup.com/j/acm doi: 10.11648/j.acm.20170604.18 ISSN: 2328-5605 (Print); ISSN: 2328-5613 (Online) A Robust Preconditioned
More information7.4 The Saddle Point Stokes Problem
346 CHAPTER 7. APPLIED FOURIER ANALYSIS 7.4 The Saddle Point Stokes Problem So far the matrix C has been diagonal no trouble to invert. This section jumps to a fluid flow problem that is still linear (simpler
More informationAutomated Scientific Computing Logg, Mardal, Wells (editors)
Automated Scientific Computing Logg, Mardal, Wells (editors) This is a preliminary draft, June 9, 2009. Send comments and suggestions to fenics-book-dev@fenics.org. Automated Scientific Computing The FEniCS
More informationSolution Sheet 3. Solution Consider. with the metric. We also define a subset. and thus for any x, y X 0
Solution Sheet Throughout this sheet denotes a domain of R n with sufficiently smooth boundary. 1. Let 1 p
More informationGypsilab : a MATLAB toolbox for FEM-BEM coupling
Gypsilab : a MATLAB toolbox for FEM-BEM coupling François Alouges, (joint work with Matthieu Aussal) Workshop on Numerical methods for wave propagation and applications Sept. 1st 2017 Facts New numerical
More informationCOMPUTATIONAL MODELING OF SHAPE MEMORY MATERIALS
COMPUTATIONAL MODELING OF SHAPE MEMORY MATERIALS Jan Valdman Institute of Information Theory and Automation, Czech Academy of Sciences (Prague) based on joint works with Martin Kružík and Miroslav Frost
More informationIdealized Simulations of Vasoconstriction and Vasodilation of Cerebral Arteries based on Mechanotransduction
Idealized Simulations of Vasoconstriction and Vasodilation of Cerebral Arteries based on Mechanotransduction Shifteh Sherafat Master s Thesis, Spring 2016 Cover design by Martin Helsø The front page depicts
More informationFinite Element Decompositions for Stable Time Integration of Flow Equations
MAX PLANCK INSTITUT August 13, 2015 Finite Element Decompositions for Stable Time Integration of Flow Equations Jan Heiland, Robert Altmann (TU Berlin) ICIAM 2015 Beijing DYNAMIK KOMPLEXER TECHNISCHER
More informationDiscretization of PDEs and Tools for the Parallel Solution of the Resulting Systems
Discretization of PDEs and Tools for the Parallel Solution of the Resulting Systems Stan Tomov Innovative Computing Laboratory Computer Science Department The University of Tennessee Wednesday April 4,
More informationA monolithic FEM solver for fluid structure
A monolithic FEM solver for fluid structure interaction p. 1/1 A monolithic FEM solver for fluid structure interaction Stefan Turek, Jaroslav Hron jaroslav.hron@mathematik.uni-dortmund.de Department of
More informationPreconditioners for the incompressible Navier Stokes equations
Preconditioners for the incompressible Navier Stokes equations C. Vuik M. ur Rehman A. Segal Delft Institute of Applied Mathematics, TU Delft, The Netherlands SIAM Conference on Computational Science and
More informationSolving PDE s with FEniCS. L. Ridgway Scott. Laplace and Poisson
Solving PDE s with FEniCS Laplace and Poisson L. Ridgway Scott The Institute for Biophysical Dynamics, The Computation Institute, and the Departments of Computer Science and Mathematics, The University
More informationAdjoints for sensitivity, optimisation and stability: I
Adjoints for sensitivity, optimisation and stability: I Patrick Farrell, Simon Funke ANADE Summer School 2014 September 23, 2014 P. E. Farrell (Oxford) dolfin-adjoint I September 23, 2014 1 / 10 dolfin-adjoint
More informationAn Introduction to the Discontinuous Galerkin Method
An Introduction to the Discontinuous Galerkin Method Krzysztof J. Fidkowski Aerospace Computational Design Lab Massachusetts Institute of Technology March 16, 2005 Computational Prototyping Group Seminar
More informationEfficient Augmented Lagrangian-type Preconditioning for the Oseen Problem using Grad-Div Stabilization
Efficient Augmented Lagrangian-type Preconditioning for the Oseen Problem using Grad-Div Stabilization Timo Heister, Texas A&M University 2013-02-28 SIAM CSE 2 Setting Stationary, incompressible flow problems
More informationarxiv: v1 [math.na] 14 May 2016
NONLINEAR PRECONDITIONING: HOW TO USE A NONLINEAR SCHWARZ METHOD TO PRECONDITION NEWTON S METHOD V. DOLEAN, M.J. GANDER, W. KHERIJI, F. KWOK, AND R. MASSON arxiv:65.449v [math.na] 4 May 6 Abstract. For
More information(Multiscale) Modelling With SfePy
(Multiscale) Modelling With SfePy Random Remarks... Robert Cimrman & Eduard Rohan & others Department of Mechanics & New Technologies Research Centre University of West Bohemia Plzeň, Czech Republic PANM
More informationOverlapping Schwarz preconditioners for Fekete spectral elements
Overlapping Schwarz preconditioners for Fekete spectral elements R. Pasquetti 1, L. F. Pavarino 2, F. Rapetti 1, and E. Zampieri 2 1 Laboratoire J.-A. Dieudonné, CNRS & Université de Nice et Sophia-Antipolis,
More informationSTCE. Adjoint Code Design Patterns. Uwe Naumann. RWTH Aachen University, Germany. QuanTech Conference, London, April 2016
Adjoint Code Design Patterns Uwe Naumann RWTH Aachen University, Germany QuanTech Conference, London, April 2016 Outline Why Adjoints? What Are Adjoints? Software Tool Support: dco/c++ Adjoint Code Design
More informationAn Efficient Low Memory Implicit DG Algorithm for Time Dependent Problems
An Efficient Low Memory Implicit DG Algorithm for Time Dependent Problems P.-O. Persson and J. Peraire Massachusetts Institute of Technology 2006 AIAA Aerospace Sciences Meeting, Reno, Nevada January 9,
More informationUn problème inverse: l identification du profil de courant dans un Tokamak
Un problème inverse: l identification du profil de courant dans un Tokamak Blaise Faugeras en collaboration avec J. Blum et C. Boulbe Université de Nice Sophia Antipolis Laboratoire J.-A. Dieudonné Nice,
More informationANALYSIS OF AN INTERFACE STABILISED FINITE ELEMENT METHOD: THE ADVECTION-DIFFUSION-REACTION EQUATION
ANALYSIS OF AN INTERFACE STABILISED FINITE ELEMENT METHOD: THE ADVECTION-DIFFUSION-REACTION EQUATION GARTH N. WELLS Abstract. Analysis of an interface stabilised finite element method for the scalar advectiondiffusion-reaction
More information2.29 Numerical Fluid Mechanics Spring 2015 Lecture 9
Spring 2015 Lecture 9 REVIEW Lecture 8: Direct Methods for solving (linear) algebraic equations Gauss Elimination LU decomposition/factorization Error Analysis for Linear Systems and Condition Numbers
More informationDiscrete Projection Methods for Incompressible Fluid Flow Problems and Application to a Fluid-Structure Interaction
Discrete Projection Methods for Incompressible Fluid Flow Problems and Application to a Fluid-Structure Interaction Problem Jörg-M. Sautter Mathematisches Institut, Universität Düsseldorf, Germany, sautter@am.uni-duesseldorf.de
More informationThe Plane Stress Problem
The Plane Stress Problem Martin Kronbichler Applied Scientific Computing (Tillämpad beräkningsvetenskap) February 2, 2010 Martin Kronbichler (TDB) The Plane Stress Problem February 2, 2010 1 / 24 Outline
More informationSolving PDEs with Multigrid Methods p.1
Solving PDEs with Multigrid Methods Scott MacLachlan maclachl@colorado.edu Department of Applied Mathematics, University of Colorado at Boulder Solving PDEs with Multigrid Methods p.1 Support and Collaboration
More informationResources. Introduction to the finite element method. History. Topics
Resources Introduction to the finite element method M. M. Sussman sussmanm@math.pitt.edu Office Hours: 11:1AM-12:1PM, Thack 622 May 12 June 19, 214 Strang, G., Fix, G., An Analysis of the Finite Element
More informationANALYSIS OF THE FEM AND DGM FOR AN ELLIPTIC PROBLEM WITH A NONLINEAR NEWTON BOUNDARY CONDITION
Proceedings of EQUADIFF 2017 pp. 127 136 ANALYSIS OF THE FEM AND DGM FOR AN ELLIPTIC PROBLEM WITH A NONLINEAR NEWTON BOUNDARY CONDITION MILOSLAV FEISTAUER, ONDŘEJ BARTOŠ, FILIP ROSKOVEC, AND ANNA-MARGARETE
More informationProgress in Parallel Implicit Methods For Tokamak Edge Plasma Modeling
Progress in Parallel Implicit Methods For Tokamak Edge Plasma Modeling Michael McCourt 1,2,Lois Curfman McInnes 1 Hong Zhang 1,Ben Dudson 3,Sean Farley 1,4 Tom Rognlien 5, Maxim Umansky 5 Argonne National
More informationDesign and Analysis of Algorithms
Design and Analysis of Algorithms CSE 5311 Lecture 21 Single-Source Shortest Paths Junzhou Huang, Ph.D. Department of Computer Science and Engineering CSE5311 Design and Analysis of Algorithms 1 Single-Source
More informationA Sobolev trust-region method for numerical solution of the Ginz
A Sobolev trust-region method for numerical solution of the Ginzburg-Landau equations Robert J. Renka Parimah Kazemi Department of Computer Science & Engineering University of North Texas June 6, 2012
More informationLarge Deformation of Hydrogels Coupled with Solvent Diffusion Rui Huang
Large Deformation of Hydrogels Coupled with Solvent Diffusion Rui Huang Center for Mechanics of Solids, Structures and Materials Department of Aerospace Engineering and Engineering Mechanics The University
More informationLecture Notes #10. The "paradox" of finite strain and preferred orientation pure vs. simple shear
12.520 Lecture Notes #10 Finite Strain The "paradox" of finite strain and preferred orientation pure vs. simple shear Suppose: 1) Anisotropy develops as mineral grains grow such that they are preferentially
More informationComputational Models of Film and Bulk Superconductors in. Time Varying Magnetic Fields. Alden R. Pack
Computational Models of Film and Bulk Superconductors in Time Varying Magnetic Fields Alden R. Pack A senior thesis submitted to the faculty of Brigham Young University in partial fulfillment of the requirements
More informationDivergence Formulation of Source Term
Preprint accepted for publication in Journal of Computational Physics, 2012 http://dx.doi.org/10.1016/j.jcp.2012.05.032 Divergence Formulation of Source Term Hiroaki Nishikawa National Institute of Aerospace,
More informationKey words. finite element method, partial differential equations, embedded algorithms
UNIFIED EMBEDDED PARALLEL FINITE ELEMENT COMPUTATIONS VIA SOFTWARE-BASED FRÉCHET DIFFERENTIATION KEVIN LONG, ROBERT KIRBY, AND BART VAN BLOEMEN WAANDERS Abstract. Computational analysis of systems governed
More informationGame Physics. Game and Media Technology Master Program - Utrecht University. Dr. Nicolas Pronost
Game and Media Technology Master Program - Utrecht University Dr. Nicolas Pronost Soft body physics Soft bodies In reality, objects are not purely rigid for some it is a good approximation but if you hit
More informationAMS 529: Finite Element Methods: Fundamentals, Applications, and New Trends
AMS 529: Finite Element Methods: Fundamentals, Applications, and New Trends Lecture 3: Finite Elements in 2-D Xiangmin Jiao SUNY Stony Brook Xiangmin Jiao Finite Element Methods 1 / 18 Outline 1 Boundary
More information1 Useful Definitions or Concepts
1 Useful Definitions or Concepts 1.1 Elastic constitutive laws One general type of elastic material model is the one called Cauchy elastic material, which depend on only the current local deformation of
More informationConstitutive Relations
Constitutive Relations Andri Andriyana, Ph.D. Centre de Mise en Forme des Matériaux, CEMEF UMR CNRS 7635 École des Mines de Paris, 06904 Sophia Antipolis, France Spring, 2008 Outline Outline 1 Review of
More informationarxiv: v1 [math.oc] 26 Jun 2016
AN ITERATION COUNT ESTIMATE FOR A MESH-DEPENDENT STEEPEST DESCENT METHOD BASED ON FINITE ELEMENTS AND RIESZ INNER PRODUCT REPRESENTATION TOBIAS SCHWEDES, SIMON W. FUNKE, AND DAVID A. HAM arxiv:1606.08069v1
More informationFREE BOUNDARY PROBLEMS IN FLUID MECHANICS
FREE BOUNDARY PROBLEMS IN FLUID MECHANICS ANA MARIA SOANE AND ROUBEN ROSTAMIAN We consider a class of free boundary problems governed by the incompressible Navier-Stokes equations. Our objective is to
More informationNewton-Multigrid Least-Squares FEM for S-V-P Formulation of the Navier-Stokes Equations
Newton-Multigrid Least-Squares FEM for S-V-P Formulation of the Navier-Stokes Equations A. Ouazzi, M. Nickaeen, S. Turek, and M. Waseem Institut für Angewandte Mathematik, LSIII, TU Dortmund, Vogelpothsweg
More informationNewton s Method and Efficient, Robust Variants
Newton s Method and Efficient, Robust Variants Philipp Birken University of Kassel (SFB/TRR 30) Soon: University of Lund October 7th 2013 Efficient solution of large systems of non-linear PDEs in science
More information