The Ghosts of CFD: Past, Present & Future

I think we are all familiar with the story of The Christmas Carol. Good ole' Ebenezer takes a trip through the past, present and future to see the impact he has on his own destiny. Got me thinking of how it relates to where we are today in the engineering software world, specifically the CFD world.

If we take a quick look at the PAST, we see that CAD was changing the way that engineers designed products. In the 80s and into the 90s 3D CAD was considered to be on the high end of the design spectrum. It wasn't mainstream, but by only a select few industries. There was an even smaller sector of very specialized engineers that were focused on CFD. It was used primarily in the automotive and aerospace industries. There were numerous independent codes that were developed to solve a particular problem. Some of these specialized codes evolved and morphed into various portions of commercially available software. Some of which is still available today.

The typical user was a very specialized engineer, usually a PhD, that was focused on driving and in many cases customizing CFD software to solve a very specific problem. The use of USR (user sub-routines) was common. As was spending a tremendous amount of time building a mesh manually to hopefully solve a very simplified version of a particular problem. The end result was often a very anti-climactic simple vector and iso-line plots Not to mention, the computer power required to solve these models was either super expensive to buy or needed to be shared/rented.

Over the years, huge strides have been made in geometric handling, automatic mesh generation, infinite list of numerical (solver) methods and true real-life visualization. Many of the commercial vendors made improvements in making it "easier to use". But let's face it, anything was easier than what was available in the beginning.

That brings us to the PRESENT. Dare I say, that CFD is "beginning" to become mainstream in a number of industries. It is used by engineers with various backgrounds. Safe to say that you can't turn your head and identify a product, piece of equipment, or device that a CFD model was run at some point.

Many of the older, "traditional" CFD vendors are finding themselves in a difficult position. They are still serving a very specialized community of "R&D" engineers that are solving very sophisticated problems. The vendors are trying to make their software easier to use so that they can touch a wider audience. Some are remaining very general purpose, others are creating specific verticals and some are creating "CFD-lite" programs. It is very challenging for a company to have multiple products that serve vastly different markets. But the challenge drives deeper as many of the vendors do not realize that the markets are vastly different. Many feel that if the product is easier to use, then the market that is serves grows. This isn't exactly true.

Some of us, work for the new generation CFD companies that are focused on mainstream engineers that happen to have a need to optimize their designs from a fluid flow and heat transfer perspective. I'll spare you the sales pitch. But think it is only fair to acknowledge that there are a variety of challenges involved in taking something that is technically very complex, and make it simply work - all the time, for every situation, the way that "Joe Engineer" expects it to work.

The crunch that the world is finding itself into lately is only going to make things tougher on the vendors but should result in some pretty amazing products for engineers everywhere.

So, what does the FUTURE hold for the CFD market?

• The monster CAE vendors that are gobbling up company after company have their hands full to deliver quality products that serve a unique market of design engineers that need to make CFD a part of their product design process.
• Expectations of design engineers is growing everyday. Users with minimal experience expect that the CFD software will help solve their problem with minimal input (requiring automation), extremely fast (requiring sophisticated software) and help bring products to market faster (requiring design insight beyond traditional visualization).
• New-generation engineers expect instant gratification. CFD software should work like all of the other tools that engineers use. We will begin to see MCAD, CFD, Office and Web Applications looking and working alike.
• High performance computing, distributed computing, maybe even cloud computing will become a normal part of an engineers day to day.
  
• Communication and collaboration tools (including Social Media) will play a bigger part in the product development community.

Just a bit of reflection on where we've been and where we're going.