#Barbagroup News—preprint on #arXiv: "Predictive Limitations of Physics-Informed Neural Networks in Vortex Shedding"

We look at the flow around a 2D cylinder and find that data-free PINNs are unable to predict vortex shedding. Data-driven PINN exhibits vortex shedding only while the training data (from a traditional CFD solver) is available, but reverts to the steady state solution when the data flow stops.

https://arxiv.org/abs/2306.00230

#Barbagroup News—preprint on #arXiv: "Predictive Limitations of Physics-Informed Neural Networks in Vortex Shedding"

We look at the flow around a 2D cylinder and find that data-free PINNs are unable to predict vortex shedding. Data-driven PINN exhibits vortex shedding only while the training data (from a traditional CFD solver) is available, but reverts to the steady state solution when the data flow stops.

#Barbagroup News—preprint on #arXiv: "PyExaFMM: an exercise in designing high-performance software with #Python and #Numba"
https://arxiv.org/abs/2303.08394

We find that designing performant Numba code for complex algorithms can be as challenging as writing in a compiled language.