The Wind-US CFD Analysis Software Package Gets Better

When I last reviewed the NPARC Alliance's CFD analysis software, the production Wind-US flow solver stood at version 2.216, and development of version 3 was up to 3.25. Since then, a lot has happened, and version 3 has finally been released for production use (v3.161). So, this seemed like a good time to revisit the solver and look at what has changed.

NPARC Alliance logo

One thing that has not changed is that Wind-US is still a highly versatile general-purpose CFD code, albeit with a strong aerospace slant. So, if your business is located in the United States, and you're looking for maximum versatility, the Wind-US solver is definitely one option you should consider.

As always, this CFD analysis software is available, free of charge, to any U.S. organization. As a U.S. government code, Wind-US is export-controlled, so those of you located in other countries will need to find alternative solutions. Within the U.S., the solver and related utilities are freely available (including source code) to any U.S. organization (see the license for restrictions). I should also note that, while I am both a developer and user of Wind-US, I am not permitted to distribute the code to new users. If you are interested in obtaining the code, you must follow the instructions found on the NPARC Alliance home page

Major New and Improved Features

As you might expect in an update that was 4+ years in the making, almost every part of this CFD analysis software was changed. Here are some items that I consider to be highlights. For more details, check out the official overview of Wind-US 2.0--> Wind-US 3.0 changes.

Unstructured Solver Improvements

Wind-US has had a built-in unstructured grid flow solver since version 1.0, but earlier versions suffered from a number of deficiencies which, effectively, kept users from being able to use it. The new version is vastly improved, and (finally) delivers on the promise of a production capability. Here are some of the major elements of this improved capability:

  • The unstructured solver is more accurate and much more robust.
    • The extreme sensitivity of the unstructured solver to near-wall grid quality has been greatly improved by the addition of Diskin's face-tangent augmentation approach to the viscous terms.
    • Gradients are now computed using a least squares method, which better maintains accuracy in regions of poor grid quality.
    • Robustness and convergence rate have also been improved by the addition of a new Gauss-Seidel line solver.

  • All the chemistry capabilities of the structured solver are now available on unstructured meshes

  • Additional k-epsilon turbulence models are now available: Shih's non-linear low Reynolds number model and Goldberg's realizable non-linear low Reynolds number model.
  • Actuator disk and screen boundary conditions now work for unstructured meshes (as well as structured)
  • Moving body calculations can now be performed on unstructured grids.
Structured Solver Improvements

While the changes to the structured solver are perhaps not as dramatic as those of the unstructured, there are several noteworthy developments for this part of the CFD analysis software.

  • The new “average” mode of block coupling has been found to dramatically improve stability for many cases...the more face-coupled points in your mesh, the more likely you are to benefit from this feature.
  • A version of the SST turbulence model that simulates bypass transition has been added.
  • The Bender, Anderson, and Yagle (BAY) model for vane-type vortex generators has been added.
  • New flux calculation methods have been added for better time-accurate simulations.

Iso-surface of vorticity magnitude in an impinging jet calculation

Improved Chemistry Capabilities

  • Total conditions may now be specified for any of the gas models supported by Wind-US, both for the freestream (reference) conditions, and also at inflow boundaries.
  • A Damkohler rate limiter has been added for finite-rate chemistry.
  • The chemistry files shipped with this CFD analysis software have been updated to fix some errors and also use the NASA 3287 format for thermo quantities (which allows Wind-US to accurately compute total conditions when using them).
Other Code Improvements

  • “Surface Groups” can now be specified to simplify specification of loads calculations in Wind-US, and also make post-processing with the included CFPOST utility easier. Some commercial visualization tools can also make use of this information.
  • The calculation of wall distance (needed for most turbulence models) has been dramatically sped up. The days of waiting for hours every time you start a new run are over. Compared to the old algorithm, the calculation is 10x faster (or better).
  • Freestream reference conditions can now be specified in terms of Reynolds number instead of pressure (for users more accustomed to OVEFLOW-style inputs).
  • Additional bleed region models have been implemented, and a synthetic jet model is now available as well.
  • Support for rotating frames of reference and zones in relative rotation has been improved. New methods for coupling zones with different azimuthal extents have been added. The interface can also now be radially oriented (e.g. an outer ring rotating about a fixed inner cylinder), instead of only axially (e.g. a rotor-stator interface).
  • Conjugate heat transfer can now be computed using Wind-US. In this mode, Wind-US is linked in a “loosely-coupled” manner to an external heat transfer code (called HTX—also available from NPARC).

For More Information...

To learn more about this new version of Wind-US (and related utilities), see the documentation on the NPARC web site . If you are interested having me support your organization's use of the Wind-US CFD analysis software or would like custom capabilities implemented in it, let me know.

Or, when you are finished here, you can return to the free CFD analysis software page.

Alternatively, head back to the Innovative CFD home page and browse one of the other topics.

New! Comments

Have your say about what you just read! Leave a comment in the box below.