Publication type: Course material
Title: Tutorial on OpenFOAM & kaleidosim : introducing the kaleidoscope feature
Authors: Boiger, Gernot Kurt
Sharman, Darren
Michel Rivero, Jhimy
Lienhard, Viktor
et. al: No
Issue Date: Jul-2021
Publisher / Ed. Institution: ZHAW School of Engineering, ICP Institute of Computational Physics
Language: English
Subjects: OpenFOAM; Kaleidosim; Kaleidoscope; Parameter study; Multiphysics; Cloud computing; Simulation; CFD
Subject (DDC): 005: Computer programming, programs and data
530: Physics
Abstract: The Kaleidoscope Feature is introduced within the Kaleidosim cloud platform: Using a couple of relatively simple Python utilities, OpenFoam v1912, simpleFoam solver and the Motorbike tutorial case, Prof. G. Boiger of ZHAW_ICP does a 19min live-demo of a work-flow that would have taken four full working days only six months ago: The famous OpenFoam 'Motorbike' tutorial case is modified such that relative onset flow velocity as well as the entire wind channel are being rotated in 360 steps (one step per degree and 360° in total) around the 'Motorbike'. One single base-case is prepared introducing variable parameters within 'initialConditions' and 'blockMeshDict' dictionaries such that: @Variable-Parameter-Name@. Here the variable parameter is the #Angle-of-Attack. The 'eval' function of OpenFoam v1912 is used as well as since wind-channel coordinates are modified with respect to variable angles and using 'sin' and 'cos' functions. The thus prepared base-case is uploaded to Kaleidosim cloud platform. Then the 'Kaleidoscope Feature' comes into play: 360 individual turbulent steady-state OpenFoam simulation cases are created automatically and run simultaneously in the cloud using Kaleidosim (MSCC Massive Simultaneous Cloud Computing). Drag- and lift- coefficients are being calculated and evaluated for each case from parsing terminal output. One Paraview image is being automatically created per simulation run in the cloud using Paraview-Batch-Mode via a prepared Python script that was uploaded along with the OpenFoam case. Results are selectively downloaded from the cloud using the #Katana File Downloader function and Paraview images are automatically forged into one movie, rotating the view around the 'Motorbike' along with resulting turbulent flow-field calculation. It is shown that the whole immense workflow, comprising 360 individual simulation runs on a 300k cell mesh, is completed in just 28min.
Further description: YouTube
URI: https://youtu.be/6T08Li7gVqE
https://digitalcollection.zhaw.ch/handle/11475/23829
License (according to publishing contract): Not specified
Departement: School of Engineering
Organisational Unit: Institute of Computational Physics (ICP)
Published as part of the ZHAW project: GeoCloud – Simulation Software for Cloud-based Digital Microstructure Design of New Fuel Cell Materials
Appears in collections:Publikationen School of Engineering

Files in This Item:
There are no files associated with this item.


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.