ΦFlow is a research-oriented, differentiable, open-source physics simulation toolkit (with a particular focus on fluids). It is written mostly in Python and can use NumPy, TensorFlow and PyTorch for execution.
Having all functionality of a fluid simulation running in TensorFlow opens up the possibility of back-propagating gradients through the simulation as well as running the simulation on GPUs.
Contact us at i15ge@cs.tum.edu if you have comments or questions!
Download Φflow on github: https://github.com/tum-pbs/phiflow
Related papers:
– Learning to Control PDEs with Differentiable Physics
– Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solvers
Acknowledgements: This work is supported by the ERC Starting Grant realFlow (ERC-2015-StG-637014) and the Intel Intelligent Systems Lab.