Tutorials
Tutorial 1: Continuum Mechanics of Multi-Component and Multiphasic Materials
Arndt Wagner, Wolfgang Ehlers, Institute of Applied Mechanics (Civil Engineering), Universität Stuttgart
Venue: Room 0.124, Universitätsstraße 38
Short Description:
This tutorial basically aims to provide a fundamental introduction to continuum mechanics and material theory of multiphasic materials within the framework of the Theory of Porous Media (TPM). For this purpose, a sound theoretical part is given by means of a hands-on lecture which covers the topics of kinematics, stress concept, balance relations, constitutive settings and, conclusively, its application to a fluid-saturated, elastically deformable solid skeleton as a basic example. In this regard, accompanying lecture notes are available for the participants, which are invited to treat selected tasks during the course and to discuss the results within the group. A concluding part shows the suited computational usage of the TPM for engineering applications. For the numerical investigation of the arising coupled problems of porous materials within a finite-element solution procedure, the capabilities of the research code PANDAS (Porous media Adaptive Nonlinear finite element solver based on Differential Algebraic Systems, cf. http://www.get-pandas.com, developed at the Chair of Continuum Mechanics of the Institute of Applied Mechanics (CE) at the University of Stuttgart) are of immense benefit for the implementation of such models.
Tutorial 2: Isogeometric Analysis
Manfred Bischoff, Institute for Structural Mechanics (Civil Engineering), Universität Stuttgart
Venue: Room 1.155, Pfaffenwaldring 7
Short Description:
The tutorial provides an introduction into isogeometric analysis. This is a specific version of the finite element method combining the isoparametric concept with a representation of the geometry which is based on CAD data and which is independent of the discretization. It is therefore also often referred to as a method representing the "exact" geometry. The fundamental idea relies on using parameterizations of the function spaces (shape functions) for the unknowns (e.g. displacements in problems of solid and structural mechanics) which are typical for CAD systems, namely B-splines and NURBS (non-uniform rational B-splines). An important feature of shape functions created by splines is an easy control of continuity. This facilitates a much easier treatment of problems with variational index higher than one (e.g. Bernoulli beams and Kirchhoff-Love shells). It is shown in the tutorial how isogeometric finite elements are formulated and what are there distinctive features. Implementations in the computer algebra system Maple demonstrate the features of this technology in numerical examples. Both advantages and disadvantages of the approach are highlighted. The tutorial will take place in the multi-media lab of the Institute of Structural Mechanics, allowing for hands-on computations of numerical examples by the attendees.
Tutorial 3: Introduction to the parallel solution of partial differential equations with DUNE-PDELab
Olaf Ippisch, Steffen Müthing, Interdisciplinary Center for Scientific Computing (IWR), Universität Heidelberg
Venue: Room 0.124 (morning session) and room 0.452 (afternoon), Universitätsstraße 38
Short Description:
The Distributed and Unified Numerics Environment (DUNE) is a software framework for the numerical solution of partial differential equations with grid-based methods. Using generic programming techniques in C++ it strives for both: high flexibility (efficiency of the programmer) and high performance (efficiency of the program). DUNE provides, among other things, a large variety of local mesh refinement techniques, a scalable parallel programming model, an ample collection of finite element methods and efficient linear solvers.
DUNE-PDELab is a powerful tool for implementing discretisations of partial-differential equations. It helps to substantially reduce the time to implement discretisations and solvers for (systems of) PDEs based on DUNE. It is not only suitable for rapid prototyping but also for building high-performance simulation software and is used by a variety of projects already.
This one day course will provide an introduction to model implementation with DUNE-PDELab. At the end the attendees will have an insight into the implementation of finite element and finite volume methods with PDELab for high-performance computers. Topics covered are the solution of stationary and time-dependent problems, and the use of parallel computers. The course also includes practical exercises. The exercises require a solid knowledge of C++ including the usage of templates.
Tutorial 4: GPU Programming
Michael Burger, Scientific Computing, Technische Universität Darmstadt.
Venue: Room 0.452, Universitätsstraße 38
Short Description:
This beginners tutorial will provide a basic introtuction into programming GPGPUs with the OpenCL framework managed by Khronos group. It starts with a brief review of the hardware-evolution from simple graphic cards, producing only 2D images, to fully programmable and massively parallel units. A view into the hardware-architecture of modern Southern Island (AMD) and Kepler (NVidia) cards will reveal the strengths, weaknesses and limitations of GPGPU programming. The OpenCL execution- and memory- model will be presented in detail and compared to NVidias widespread CUDA framework. The following introduction into the syntax of OpenCL C is coupled with practical programming exercises and some information about best practices.
3rd International Workshop on Computational Engineering
Registration Deadline
September 1, 2014Date
October 6-10, 2014Venue
Universität StuttgartComputer Science Building
Universitätsstraße 38
Campus Map
70569 Stuttgart
Germany