How we run 4 Real-life CFD Simulations in the Cloud in 20 Minutes for 5.6 USD
CFD ventilation duct case, simulated with

How we run 4 Real-life CFD Simulations in the Cloud in 20 Minutes for 5.6 USD


Without a doubt, and based on my personal experience: Kaleidosim is one of the easiest ways to run a CFD simulation in the cloud. In this short review, I would like to demonstrate, how one could run multiple parameter studies. Not only fast but also cost-efficiency.

ventilation duct 3D geometry with partition walls and 2 figures | black and white picture | transparent
Mockup for CFD-Simulation | ventilation duct task

Kaleidosim CFD computation cloud functionality to be checked by running a mid-size performance test.

My test-case specification:

  • 4 different parametric geometries have been created and meshed
  • 1 inlet and 2 outlets
  • You will find a complete tutorial with 2 cases on our GitLab repository
  • Air flow distribution in several apartment configurations to be investigated by using CFD methods
  • Simplified steady-state study
  • Impact on people in the area should be investigated for different geometries

Apart from this, in case you are looking for some open-source 3D design tools: here is our extensive list.

PLATFORM DETAILS (Status 04/2020)

In short, find some interesting details about the platform:

  • CFD computation cloud with all available OpenFOAM (Foundation and ESI-OpenCFD)
  • Running directly in your browser
  • Possibility to host your own solver
  • Virtual machines with up to 96 cores (status on 04/2020, scalable)
  • Managed by a Swiss startup based in the beautiful city of Zurich
  • 24 CPU Hours for new registrations
  • Pay as you go price is 0.12 USD / 1 CPU Hour, (find more details here)
  • No extra costs, everything is included
  • Highest data security standard
  • Step by step tutorial available at the homepage


  • Full automated “Ventilation Duct” parametric 3D model creation with Salome Shaper. We share this module among many others with you on GrabCAD. Let us know if you need some modifications on it.
  • 4 models have been defined
  • Automated meshing with snappyHexMesh
  • OpenFOAM setup and short test on local machine
  • Config based on OF motorbike tutorial, simpleFoam solver
  • Upload zip-archives to Kaleidosim platform
  • Setup and start the simulation runs
  • Download the data
  • Postprocessing with ParaView


Mesh for CFD simulation: created with open-source snappyHexMesh tool | iso-view
Hexadominant mesh for CFD simulation: created with open-source snappyHexMesh tool | slice view
  • All meshes < 2 mio. cells
  • Boundary layer on walls, ca. 90 % coverage
  • Inlet and outlets extended
  • Some cells with bad skew quality, but the  focus here is on to follow the physics, not the exact geometry


  • Lenovo ThinkPad T530i

  • Architecture: x86_64

  • Linux Ubuntu LTS 18.04

  • Thread(s) per core: 2

  • Core(s) per socket: 2

  • Intel(R) i5-3320M CPU

  • CPU MHz: 2620.722

  • BogoMIPS: 5188.22

  • 1000 Iterations

  • simpleFoam solver

  • 20 data exports

  • 80 minutes!



To summarise:

  • Kaleidosim platform is easy to use, reliable and self-explanatory
  • Free available Open Source CFD solver, already installed on the cloud and ready-to-go.
  • Sufficient capacity to run small and mid-size real-life projects
  • Scalable and unlimited
  • Run time on the cloud took ca. 5.3 CPU Hours for 1 config => 0.64 USD
    (mean value for all runs)
  • Download time with standard internet connection took to local machine ca. 15 min for all test cases (each 20 time steps)


Create ZIP Archive

$ zip -vr run-01/

Create Animation in GIF or MOV Format From Single Pictures

$ convert -loop 0 -delay 20 -quality 100% anim1.*png
$ convert -loop 0 -delay 20 -quality 100% anim2.*bmp movie2.gif

Merge multiple GIF Animations

$ gifsicle movie1.gif movie2.gif > movie_merged.gif 

This offering is not approved or endorsed by OpenCFD Limited, producer and distributor of the OpenFOAM software via, and owner of the OPENFOAM® and OpenCFD® trade marks.