Choosing a CFD Solver
If you are new to computational fluid dynamics, selecting a CFD solver can sometimes be a daunting task. There are a bewildering variety of choices.
Commercial CFD Packages
Even though I make
developing non-commercial CFD codes and helping people use them, it must be said that commercial packages are a good solution for many organizations. So before we dive into the discussion of the various non-commercial CFD solvers, we need to make a quick side trip to the world of commercial software.
If you are looking at commercial CFD packages,
evaluate the different alternatives carefully
to be sure you're getting the right solution for your organization. There are many reputable companies that offer different CFD packages which cover the whole spectrum from highly customized expert systems, to stand-alone solvers, to all-in-one packages that attempt to provide everything you need to do CFD at your company. Follow the link for
a short discussion of just a few of them.
One size does not fit all, however, and sometimes businesses find that their needs require a different approach, either for a single project, or for all of their CFD simulations. For example, your projects might require the inclusion of terms in the governing equations which are normally neglected (because they're very small in most cases). Or, you might require a higher order scheme than is commonly available (for an acoustics application, perhaps). Or maybe your budget just won't stretch for the licensing costs.
Whatever your reason, if you need something besides the commercial packages, there are many high quality CFD solvers which are freely available from one source or another. Here are some thumbnail reviews of some of them. Where present, you can follow the links to more detailed descriptions of the software you're interested in.
What about the many packages I haven't covered here? If you are a developer or user of another package, you can
contribute your own tip or review of a free cfd solver.
You can also see and comment on the reviews that other users have submitted.
This is the code
I use for
a lot of my CFD work,
whether development or applications. It solves the compressible Euler or Navier-Stokes equations, but beyond that, it is about as flexible a code as any I've encountered. There are
numerous options for the handling of flow physics
and many different
computational grid topologies
can be handled. Also, Wind-US can be
run in parallel
in many different ways.
Along with the CFD solver there are
numerous other utilities
that come with it to make your life a bit easier. In summary, Wind-US is not the best code for everything, but it can be applied to more different areas than many other codes. It is freely available (including source code) to U.S. organizations (see the license for restrictions). If that sounds like what you need, check it out.
In November 2011, the NPARC Alliance released Wind-US version 3.0. I have a review of the new features of
this free cfd analysis software.
This “code” is really a library of C++ routines which facilitate the numerical solution of partial differential equations. Using this library, many different solvers (included with the software) have been built to address many classes of problems in fluid dynamics (and other fields as well). Applications range from laminar incompressible flow to fully turbulent reacting compressible flow to solving the Black-Scholes equation for pricing commodities options. OpenFOAM is freely available worldwide under the GNU Public License. To learn more, see my
review of this free CFD solver.
This is a general purpose structured grid compressible CFD solver maintained primarily by Chris Rumsey at NASA Langley Research Center. One of the code's many strengths is the diversity of available turbulence models. Like Wind-US, access to CFL3D is restricted to “U.S. persons”.
The development of this widely used code is led by Pieter Buning at NASA Langley Research Center. This compressible CFD solver is specialized for use on overset structured meshes (aka CHIMERA grids). The recent release of OVERFLOW 2.0 (and now 2.1) saw significant additions which make it a leader in its ability to compute large-scale unsteady problems such as store separation from a full aircraft configuration. For additional details, see
my review of this capable computational fluid dynamics software package.
This code is the solver portion of the TETRUSS (TETRahedral Unstructured Software System) system developed by NASA Langley. USM3D solves the compressible Navier-Stokes equations on unstructured grids made up of tetrahedral elements. Like the other NASA codes, USM3D is free to U.S. organizations, and in this case, training and support for USM3D (and the whole TETRUSS system) is available, for a fee, from ViGYAN, Inc.
Development of this CFD solver is led by Stéphane Popinet, with support from the New Zealand National Institute of Water and Atmospheric Research. Gerris solves the incompressible Euler, Stokes, or Navier-Stokes equations using unstructured Cartesian meshes with immersed boundaries. This allows the use of dynamic mesh refinement to automatically generate quality grids around complex geometries.
At the present time, it is best suited for unsteady flows, but no turbulence models are available as of this writing. As a result, turbulence must either be completely resolved (a direct simulation) or the numerical dissipation of the discretization algorithms must be relied on to act as a model (a la MILES approach). Gerris is freely available under the GNU Public License.
A World of Options
From the above, you can see that there is quite a variety to choose from, and this is just scratching the surface. To any readers from outside the United States, I apologize for the U.S. slant to the reviews here; I will try to balance things out eventually with additional reviews of more internationally available packages.
If you have your own favorite code, you can
tell us about it here.
You can also see other users' submissions, and add your own comments.
When you're ready,
return to the Innovative CFD home page