Wind in the City
https://share.streamlit.io/nuflo-sarl/windload/main/dashboard/dashboard.py
Just finished another project involving a fluid simulation (CFD) of Luxembourg Kirchberg and including some custom programming (python, geodata). We’re pretty pleased with the outcome. As usual, here are some nice pictures to give you an impression of what is possible with the openfoam.com toolchain.
Microclimate and wind 💨 are important factors to take into account in city planning and building #construction.
Here are some pictures of wind #simulations (#CFD) in a simplified setup of a of a clean demo case and some pictures of a real world simulation of #Luxemburg city 🏙, taking into account terrain and existing buildings. For this simulation we have used custom tools written in #python to extract terrain data from publicly available raster data of Luxembourg city’s terrain, as well as the according #building 3D models. The simulation itself was created with the #openfoam toolchain. In this case simulating a small section of #Kirchberg was acceptable, this can be scaled to whole quarters or city with more computational power. In Luxembourg we rely on #LuxProvide and #meluxina to do so.
What are your experiences with scaling wind simulations to #HPC? Where do you see an application in the public sector? Joachim, Alban, Roger Lampach, Ramona G. Caulea, Robin, Astrid
In this case the results are used to evaluate #windload s on the buildings and to evaluate #windspeed s in the streets - an important factor for pedestrian comfort. Get into contact with us here: https://lnkd.in/eW7N3jBA
🌍 Useful python libraries: rasterio https://lnkd.in/eZqywzgJ, pyproj https://lnkd.in/equ26qfq, pyvista https://github.com/pyvista
Here you can get into contact to discuss an offer: https://lnkd.in/e-EZ-ycM
70 Meter Satelite Dish in a Breeze
Just a 70m satelite dish in a slight 25m/s breeze. “Colourful” fluid dynamics (CFD) with the openfoam.com toolchain of snappyHexMesh, simpleFoam and paraview.
Also needed to repair the model in blender and FreeCad. We really enjoy working with this opensource toolchain to provide automated solutions to our customers engineering teams. These screenshots were the result of a quick test to see how well the current version of snappyHexMesh performs with poor model quality and when spanning several orders of magnitude of relevant feature sizes.
Here is a view of the backside truss structure of the satelite dish. NASA has a repository of their 3-D models over at github which includes this model.
The dish diameter is 70m, the truss structure is massive, but compared to the dish diameter the individual trusses are relatively small. Depending on the angle of attack they may contribute significantly to the total drag. Ultimately, the important parameter is the dish shape, and the deformation under wind load determines the operational envelope. This is a nice test case because the relevant feature sizes span several orders of magnitute. Models like these are challenging to mesh, and to properly resolve.
