CFD democratization is possible
My views on Casual CFD usage by non-specialists have changed dramatically since 1999. I still believe there is a ton of value in pushing CAE usage into the hands of multi-tasking Design Engineers. But, in the last 2 decades, I've repeatedly seen this approach result in shelfware and non-use for A LOT of the the upfront CFD and FEA tools out there. That's a lot of wasted software investment.
Casual users wear many hats and often do not reach an adequate comfort level with their CAE abilities. They would rather rely on a CAE expert or physical test (set up and run by a testing expert on their behalf) than put their stamp of approval on the Upfront CAE results.
The best way to overcome this hurdle is to remove some of the "responsibility for accuracy" for these front-line engineers. That can be done through tightly focused, single-purpose CAE apps created either by the CAE vendor or the CAE experts within the company.
Unfortunately, it is too costly for CAE vendors to create "apps" for the unlimited number of niche, long-tail needs out there. But, smart CAE companies can justify app-ification for some of the larger niches.
Admixtus from CD-adapco is a great example of this. Chemical and food process mixers have a well defined set of possible parameters and geometries. Admixtus is a slick "skin" atop the full power of STAR-CCM+ covering the needs of Casual mixer design with just a few button choices. There are some other "apps" for mixing on the market (built upon Open Source CFD solvers), and a couple of them look quite nice. But, when you hit their capability wall, there's simply no way to move forward.
If you hit that wall in Admixtus, you can simply toggle over to full blown STAR-CCM+ and tickle a few advanced options in the more general purpose environment. Or, ask your in-house CFD expert to assist in that common environment.
Take a gander at this short overview video. (Disclaimer: engineers should never be allowed to name products or companies... and they certainly should not pick background music for videos.)
Innovation Roadmap Manager at Siemens Digital Industries Software, Simulation and Test Solutions
8 年Interesting article Jeff. We took a similar approach with FloTHERM PCB. A 'vertical vertical' app based on FloTHERM.
Principal Director at Applied CCM Inc
8 年Nice discussion Jeff. I tend to agree with your "Upfront CFD" advocacy. I do however see a misconception about the efforts to democratize CFD so as to enable design engineers to carry out upfront CFD. I do not necessarily believe it is the democratization of CFD (or CAE), or lack thereof, that is/will be the barrier. It's more about the abilities of the engineer to interpret the results rather than the inabilities of the software to produce the results. This fact really has nothing to do with software. The same logic applies to an experiment. The engineer using the tool, be it software or hardware, needs to have a solid background in the underlying physics. I.e. fluid dynamics in and of itself is complicated. Given this is it realistic to expect a design engineer, who have to wear so many hats, to reliably and efficiently produce accurate answers, irrespective of the tool? This, in my opinion, is the challenge for democratization.
Help You Get the Full Value out of Your Simulation Data
8 年Can I launch Admixtus without launching the full STAR-CCM+? Can I post-process Admixtus results without launching the full STAR-CCM+?