Hello, CFD Meshing World!

For the third time*, I find myself writing the obligatory, "Hello, World!" first blog post. So "Hello" to all you readers of Cadence's Computational Fluid Dynamics blog. (Or as someone else put it, Cadence Fluid Dynamics.)

Who am I? I'm an aerospace engineer by education, a programmer by practice, and now I find myself mostly meeting and learning from interesting people. That's certainly been the case since joining Cadence last month.

Who am I not? Despite having passed all my undergraduate electrical engineering classes, all I know about electronics is how to plug them in to the wall outlet. And sometimes I even get that wrong.

Why am I here? The Cadence Intelligent System Design strategy is broadening to include CFD as evidenced by the recent acquisition of CFD software provider NUMECA. And I have to admit that it gave me a warm fuzzy feeling when Cadence built upon their in-house geometry modeling and mesh generation expertise, and the fine capabilities in those same areas from NUMECA by adding Pointwise, where all we did for over 26 years was geometry modeling and mesh generation, mostly for CFD and primarily in the aerospace and defense market. Cadence is a company that gets it—mesh generation is the foundation upon which computational simulation is built.

For those of you who may not be intimately familiar with mesh generation, it is the process of discretizing the volumetric space of a computational simulation. That volumetric space would be the air or water flowing through or around a body in the case of CFD. Discretizing just means approximating the continuous volume with a finite number of simply connected mesh cells such as bricks or tetrahedra. The CFD simulation (i.e., the iteratively obtained solution to the Navier-Stokes equations) is computed across the entire mesh. For example, the image below shows a mesh for the air region around an airfoil with ice build-up at its leading edge.  

This image also illustrates some important meshing concepts. It shows a hybrid mesh, so called because it uses a mix of cell types—in this case quads and triangles. Second, it illustrates the need to use smaller mesh cells where the physics of the flow require higher-resolution—in this case, adjacent to the airfoil and ice shape.

I'll leave the details for a future blog post but let's just say that what makes a suitable mesh for any given simulation is a bit complex. The truly curious can watch our webinar, The Influence of Meshing Strategies on Simulation Efficiency. 

To bring my first blog post to an end, I'll point you at a recent article from our friends at engineering.com for which they interviewed Nick Wyman, a meshing expert from the Pointwise team who's now with us at Cadence. The article asks why it's difficult to prepare geometry models for mesh generation. The article includes a link to a white paper we recently published on "Geometry Modeling for Mesh Generation and CFD". We hope you find it interesting.

*The first two times were my personal blog and my former company's blog, Another Fine Mesh.