There is one RANS and LES CFD solver that was not yet mentioned in this thread: GitHub - mberto79/XCALibre.jl · GitHub
Any experience with this code? Does it work well for high Reynolds numbers?
ExtendableFEM.jl can solve Stokes and Navier-Stokes problems. It can use diverngence-free, pressure robust mixed finite elements.
I recently came across [2605.16684] GPU Performance of an Entropy-Stable Discontinuous Galerkin Euler Solver with Non-Conservative Terms
It seems to have a Julia implementation, at least for GPU. But I can’t find it online.
channelflow is a C++ code for incompressible Navier-Stokes and related flows in a channel geometry: a 3d rectangular box bounded in one direction with rigid walls and no-slip boundary conditions, and periodic boundary conditions in the other two directions. Spatial discretization is Fourier x Fourier x Chebyshev, temporal discretization is various finite-difference formulae. The algorithm is outlined in Caunto, Hussaini, Queteroni, and Zhang’s “Spectral Methods” book, volume 2, section 3.4.1.
The channel geometry and spectral Fourier-Chebyshev discretization is good for theoretical studies of wall-bounded turbulence. I was the principle author of this code; now it is maintained by collaborators. I’ve long wanted to rewrite the code in Julia, to make it more accessible and maintainable, and to avail of libraries like BifurcationKit.jl.
Hi,
In the current version our compressible solver handles weakly compressible flows (~M=0.5-0.6) quite well, not beyond that.
I’m preparing a new release (coming out in 1 or 2 weeks) where we implement the transonic correction for our pressure-based solvers (for M=0.6-1.5 or so) and a new density-based solver (explicit for now) for higher Mach numbers.