United Kingdom, September 8, 2014

The course addresses analysis of modern approaches to
turbulence modelling in light of HPC and their applications
to the solution of industrial problems. The course consists
of two parts: theoretical course on advanced CFD approaches
and hands-on training on Code_Saturne using STFC Daresbury
Laboratory super computers. The course will cover
fundamentals of Direct and Large Eddy simulations,
industrial applications, inlet and artificial boundary
conditions and best practice guidelines for CFD.

Description:

  

Computational Fluid Dynamics (CFD) is required in many  

industrial applications including external and internal  

aerodynamics, heat transfer, combustion, aeroacoustics,  

renewable energy systems etc. It is therefore, if not  

more, at least equally important as experimental Fluid  

Dynamics. Direct Numerical Simulation (DNS) for high  

Reynolds number flows is highly time consuming and only  

possible for relatively simple geometries. There are many  

existing turbulence modelling techniques which allow a  

significant reduction in this computational cost. However,  

different models can provide very different results for the  

same problem. The appropriate choice of a turbulence model,  

thus requires a deep enough understanding, significant  

experience and computational skills in the field of  

turbulence modelling. The course will address the analysis  

of modern approaches to turbulence modelling in light of the  

ever in- creasing High Performance Computing (HPC)  

facilities and their applications to the solution of  

industrial problems.  

  

The course consists of two parts: theoretical course on  

advanced CFD approaches and tutorials with open-source  

Code_Saturne for solving the Navier-Stokes equations. The  

main focus of the first part will be on Direct and Large  

Eddy simulations for turbulent flows, artificial boundary  

conditions and best practice guideline for CFD.  

  

The second (practical) part of the course will make use of  

the Science & Technology Facilities Council (STFC) Daresbury  

Laboratory super computers. This part is devoted to learning  

how to use the open- source CFD solver Code_Saturne, which  

is developed by EDF R&D for solving the Navier-Stokes  

equations. Code_Saturne is a co-located finite volume solver  

handling any type of mesh, structured or un- structured. It  

includes a large range of RANS/LES models with additional  

modules for compressible flows, combustion, radiative heat  

transfer, magneto-hydrodynamics, two-phase flows and  

atmospheric flows. Through various projects, STFC Daresbury  

Laboratory have tested Code_Saturne's HPC capabilities. They  

also have excellent training and HPC facilities; the Hartree  

Centre currently hosts 19th fastest supercomputers in the  

world (Top500 supercomputers, Nov 2013).  

  

The course is delivered by the academic staff from the  

School of Mechanical, Aerospace and Civil Engineering  

(MACE), The University of Manchester and experienced  

researchers from STFC Daresbury Laboratory. The School of  

MACE is a world-class Centre in CFD with the long term  

experience in the development of turbulence models and their  

implementation to the solution of real-life industrial problems.  

  

  

Learning Objectives:  

  

Main objectives are the following:  

-fundamentals of Direct and Large Eddy simulations  

-industrial applications  

-Inlet and artificial boundary conditions for LES/DNS  

-best practice guidelines for CFD