Myth-Busting CFD

The best way to get more people to use CFD is to eliminate the FUD – fear, uncertainty and doubt – that people have about CFD and what it can do for them. The folks at Mentor Graphics Mechanical Analysis Division, where FloEFD comes from, have debunked several CFD myths in their whitepaper The Five Myths of Computational Fluid Dynamics.

If I told you that three of the five myths involve mesh generation or preprocessing, would you be surprised?

Myth #1: CFD is Too Difficult

If 58.1% of engineers said they don’t have knowledge or skills to use CFD it’s apparently because they think their mesh will require “a substantial amount of tuning and tweaking such as manually modifying cells to improve the mesh quality” because low mesh quality is “one of the biggest reasons for run divergence.”

As the paper points out, however, that may have been true ten years ago but the latest meshers have automated a lot of that effort. Pointwise’s T-Rex technique for hybrid mesh generation of viscous flowfields includes automated techniques for collision detection and quality optimization for each and every cell. These measures can be customized for the ones important for your solver’s numerics.

Myth #2: CFD Takes Too Long

The authors don’t beat around the bush: “The greatest time sink is the meshing process.”

Well, one engineer’s time sink is another engineer’s last, best opportunity to influence the convergence and accuracy of the solution. By taking the time here to ensure you have a mesh that matches the CAD and has good quality cells you’re investing in the resulting CFD solutions. Pointwise provides both Fault Tolerant Meshing (healing the mesh over bad CAD) and Solid Meshing (automatic assembly of CAD into a closed solid) for minimizing geometry interoperability problems. And as noted above, the meshing algorithms are highly automated to minimize your time.

If you really want to minimize your meshing time, scripting with the Tcl-based Glyph scripting language is a great way to implement templates and macros that automate some or all of the meshing process.

Myth #4: You Can’t Directly Use CAD

While previous generations of CFD codes required a “considerable amount of manual intervention” to translate the CAD model into the CFD program, that’s certainly not the case anymore.

Pointwise has built-in readers for IGES, STEP, and seven native CAD formats so you can go directly from CAD to CFD. And as noted above, the Solid Meshing feature suite can assemble your CAD data into a watertight solid model and do so automatically on import freeing you to concentrate on meshing and CFD.

More Myth Busting

I won’t give away all of the myths in Mentor Graphics’ whitepaper. It’s available for free (with registration) from their website so get yourself a copy of The Five Myths of Computational Fluid Dynamics today.

After you read it, come back here and let us know if there are some myths they missed.

And if you want to do a little hands-on myth-busting, now might be the time to try Pointwise for your CFD mesh generation.

This entry was posted in Software and tagged , , , , , . Bookmark the permalink.

6 Responses to Myth-Busting CFD

  1. We are developing and probably will release a real online CFD platform. And I still want to say, some of the Myths is still true. The unsteady CFD simulation is still long in order to get a meaning turbulence statistics for high Re.

    PS: John, this is lakeat, we met on the forum. 🙂

  2. John Chawner says:

    Hello Daniel. An online platform for CFD sounds interesting. How will it address the myths in Mentor Graphics’ paper? How do you expect it to help the required expertise, the long times required, other issues they cite?

  3. Kevin says:

    What’s Myth #3 ?

  4. Kevin says:

    What’s the third myth?

  5. John Chawner says:

    Kevin, I didn’t think the Mentor Graphics people would like it if I gave away all five of their myths which is why I included the link to their whitepaper at the bottom of the post. I just touched on the ones that had to do with meshing.

  6. Kevin says:

    Ah, missed that tidbit – just couldn’t make sense of counting from 1, 2, to 4. 🙂 Sorry for the double-post; my browser didn’t refresh after the first!

Leave a Reply to KevinCancel reply