Upfront CFD really means CFD for Product Design

Just read a post by a buddy of mine that addresses the issue that many companies are truly never "upfront" in their design processes. They, in fact, are always somewhere "downstream" in the design process. He makes some really important points on the fact that "upfront" is multi-faceted. Click the link above to get his perspective.

I think I may take for granted that people see the value in Upfront CFD. Sometimes it seems very simple and straightforward, other times not so much. I think almost everyone would agree that if we started with a clean sheet of CAD and begin sketching up conceptual designs, it would be great to study the fluid flow and thermal aspects of our designs. We could quickly and inexpensively get great feedback on our desktop from an Upfront analysis. Take it further, we can change the CAD in an effort to improve performance, repeat and compare. It would save a great deal of time, reduce the number of prototypes and save $$$ in the long run, not to mention reduce the time to market. The classic "faster, cheaper, better"!!! Unfortunately, this scenario is a reality for some and an unrealistic dream for others. To give you a taste, check out the redesign of a Power Inverter below. Notice the two different flow patterns based on two separate vent designs.

In fact, many of us are on the go constantly, old designs come back to haunt us, new designs are constantly evolving and we are constantly put under the gun to bang out a quick fluid flow/thermal simulation. What I want out of an Upfront CFD tool is..

1. A tool that is easy to use that leverages my parametric CAD data. It should speak my language, not CFD jargon.
2. Meshing is crucial. I want it to be easy, fast and reliable. The less decisions I have to make, the better. Essentially, the more automatic, the better. But it has to be reliable.
3. Upfront CFD doesn't mean entry level CFD. I need a wide-variety of physics. Nothing that requires a PhD, but those that I consider the fundamentals - incompressible/compressible, newtonian/non-newtonian, thermal (conduction, forced & natural convection + radiation), solid body motion (poppet valves, rotating components) etc..
4. Accuracy is important. I want the software to have built-in intelligence. Select the appropriate scheme/solver etc. Tell me, when it is converged etc.
5. Really, really slick results visualization. Contours, vectors, isosurfaces, traces, animation etc. Hardcore data - XY plots, Wall Calculators.
6. Ability to share and collaborate my data with technical and non-technical folks.

Shameless plug - but all of the above exists in only a few tools out there. Arguably, a "few" is kind of stretching it. Some are better than others. I have driven most of them myself. Happy to give my opinion, so feel free to ask. Over the next few weeks, I plan to give my opinion anyway. It doesn't mean that that traditional CFD tools are wrong, just means that each has their place. Much, much more to come on this front...