PRACE Autumn School 2018 - HPC for Engineering and Chemistry

CET
Lecture room IV/2 and computing rooms II/5, III/1, N17 (University of Ljubljana)

Lecture room IV/2 and computing rooms II/5, III/1, N17

University of Ljubljana

Faculty of Mechanical Engineering Aškerčeva 6 1000 Ljubljana Slovenia
Leon Kos (University of Ljubljana)
Description

In today's competitive product development, high performance computing (HPC) delivers outstanding value and investment return. Parallel computing increases understanding, productivity and accuracy of the simulation - a faster turnaround, reduced costs, systematic design variations and more complex models.
The PRACE Autumn School communicates and discusses issues and perspectives of HPC targeting engineering applications arising from evaluation performance and/or design of products, e.g. equipment and processes, with a particular emphasis on the life sciences, chemistry,  aerospace, and energy fields. The PRACE Autumn School places priority on algorithms, simulation strategies, and programming techniques for life sciences, chemistry, finite elements,  complex fluid flow simulations,  which usually have coupled multiphysics requiring intensive use of HPC resources. In case of expertise and software availability, there will be presentations of other issues, e.g. robustness and performance analysis, a brief introduction to pre- and post-processing techniques, e.g. CAD integration, mesh generation and visualization are presented.


The PRACE Autumn School targets at (40-50) industrial users, researchers and students. Case studies and hands-on tutorials will be carried on the ULFME cluster. The tutorials will be held in parallel and in subsequent sessions, depending on the applicants' interest indicated in the registration questionnaire.

The program is free of charge (not including travel and accommodation). For the hands-on sessions, participants are expected to bring their own laptops.

Applications are open to researchers, academics and industrial researchers residing in PRACE member countries, and European Union Member States and Associated Countries. All lectures and training sessions will be in English.

  Short URL: http://events.prace-ri.eu/event/as18
 

 

Cluster access instructions
    • 08:00 08:30
      Registration 30m
    • 08:30 08:45
      Welcome address 15m
      Speakers: Prof. Janez Povh (University of Ljubljana, ULFME), Prof. Jože Duhovnik (PRACE MB member for Slovenia)
      Slides
    • 08:45 09:30
      Introduction to HPC
      • Motivation to use HPC system,
      • HPC ecosystem - HPC facilities available to researchers in Europe,
      • HPC platforms architecture;
      • How to use an HPC system (access, user interface, writing and compiling simple parallel code, dealing with errors, batch schedulers, performance tuning...)
      Conveners: Dr Jan Jona Javoršek (SLING), Dr Janez Povh (ULFME), Dr Leon Kos (ULFME)
      slides
    • 09:30 10:30
      Parallel Computing 1h
      This course will treat - the merits and limits of parallel programming - different parallel programming models (e.g. task and data parallelism) - the differences between shared and distributed memory systems
      Speaker: Mr Caspar van Leeuwen (SURFsara)
      Slides
    • 10:30 11:00
      Coffee break 30m
    • 11:00 12:15
      OpenMP 1h 15m
      OpenMP is a widely used standard for parallel programming in C/C++/Fortran on shared memory systems. This course will introduce some basic parallelization and work sharing constructs available in OpenMP and provides hands-on examples to the participants. The goal is to teach participants how multithreading can be used for parallel computation and to provide them with a starting point from which they can develop their OpenMP knowledge.
      Speaker: Mr Caspar van Leeuwen (SURFSara)
      Slides
    • 12:15 13:15
      Lunch 1h
    • 13:15 15:30
      MPI 2h 15m
      This course will introduce the basic concepts of MPI. The most commonly used routines for sending, receiving, and aggregating data will be explained using hands-on examples. The goal is to provide participants with a starting point from which they can develop their MPI knowledge, and/or to provide background knowledge for participants using HPC applications that rely on MPI for their communication.
      Speaker: Mr Caspar van Leeuwen (SURFsara)
      Slides
    • 08:30 12:00
      An introduction to Finite Element, Boundary Element, and Meshless Methods - with applications in heat transfer and fluid flow 3h 30m
      This course presents the Finite Element Method (FEM), Boundary Element Method (BEM) and Meshless Methods (MEM) within a unified framework of the method of weighted residuals (MWR). The course begins by introducing the MWR and one dimensional applications of all three methods as well as finite difference and finite volume methods. The basic fundamentals of the finite element method are then developed using simple examples in heat transfer. Particular attention is given to the development of the discrete set of algebraic equations, beginning with simple one-dimensional problems, illustrating the principles of automated generation of finite elements via Gauss quadratures and body-fitted coordinates, the loading matrix, and continuing to two- and three-dimensional elements. Once these principles are established, the concept of boundary element methods are then introduced. The relation of the BEM to the Green’s function approach for the analytical solution of partial differential equations is presented. The advantage of the BEM in reducing the dimensionality of a problem is demonstrated along with applications to problems with infinite domain boundaries. The boundary element technique is a natural extension of the finite element method, and this becomes greatly appreciated by users. The BEM is developed to 2 and 3D problems, symmetric and non-symmetric solvers iterative solvers are presented along with applications in heat transfer and inverse problems. Finally, meshless methods are developed. Here, the focus is on radial basis function (RBF) based meshless formulation in strong form. The method is simple to grasp, and simple to implement. The power of the method is becoming more appreciated with time. The meshless method has been shown to yield solutions with accuracies comparable to finite element methods employing an extensive number of elements, yet requiring no mesh (or connectivity of nodes). We have used it for structural analysis, fluid flow, heat transfer, environmental transport, and various biomedical applications. Ref: Darrell W. Pepper (University of Nevada Las Vegas), Alain J. Kassab (University of Central Florida), Eduardo A. Divo (Embry-Riddle Aeronautical University), An introduction to finite element, boundary element, and meshless methods with applications to heat transfer and fluid, American Society of Mechanical Engineers (ASME), New York, New York, USA, 2014.
      Speaker: Dr Alain Kassab (University of Central Florida, Orlando, Florida)
    • 10:00 10:30
      Coffee break 30m
    • 12:00 13:15
      Lunch 1h 15m
    • 13:15 13:45
      Nanofludics 30m
      Speaker: Dr Jens H. Walther
      Slides
    • 13:45 14:45
      Hands-on tutorial on Molecular dynamics and Computational fluid dynamics for nanofluidics 1h
      Speaker: Dr Jens H. Walther
      Slides
    • 14:45 15:00
      Coffee break 15m
    • 15:00 15:45
      Conjugate heat transfer and proper orthogonal decomposition 45m
      Speaker: Prof. Alain Kassab (University of Central Florida, Orlando, Florida)
      Slides
    • 08:30 17:00
      HPC in Chemistry and Chemical engineering
      • 08:30
        Quantum methods 45m
        Speaker: Dr Jernej Stare (KI)
        Slides
      • 09:15
        Classical and ab-initio molecular dynamics 45m
        Speaker: Dr Franci Merzel (KI)
      • 10:00
        Coffe Break 30m
      • 10:30
        QM/MM 30m
        Speaker: Dr Milan Hodošček (KI)
        Slides
      • 11:00
        Multiscale modeling & simulation 30m
        Speaker: Dr Matej Praprotnik
        Slides
      • 11:30
        Hands-on tutorial on Molecular dynamics 45m
        Speaker: Dr Milan Hodošček (KI)
        Input generator for CHARMM
      • 12:15
        Lunch 1h
      • 13:15
        Hands-on tutorial on Quatum methods (Gaussian, cpmd...) 1h 20m
        Speaker: Dr Jernej Stare (KI)
      • 14:35
        Break 15m
      • 14:50
        Hands-on tutorial on Espresso++ 55m
        Speaker: Dr Jurij Sablić (KI)
        examples
    • 08:30 15:45
      HPC in Engineering
      • 08:30
        Multiscale CFD modeling for congential heart disease 45m
        Speaker: Alain Kassab (U. Florida)
      • 09:15
        Multiscale CFD modeling for left ventricular assist device 45m
        Speaker: Prof. Alain Kassab (U. Florida)
      • 10:00
        Coffe Break 25m
      • 10:25
        Introduction to OpenFoam 1h 45m
        Speaker: Dr Aleksander Grm (University of Ljubljana, FPP)
        Slides
      • 12:15
        Lunch 1h
      • 13:15
        Hands-on tutorial on solving engineering problems with OpenFoam 1h 45m
        - simpleFoam application to simulate steady-state, turbulent, incompressible flow around the vehicle. - Solving mixed fluid-dissolution problem
        Speaker: Dr Aleksander Grm (University of Ljubljana)
        Examples