April 26th 2013
Commercial marketplace for CAE software is undoubtedly similar to any other marketplace where competitors vie for increased market share. Add to this landscape, organizations and individuals who believe in open source environments and the landscape becomes even more competitive for CAE software companies. This paradigm certainly holds true for CFD software as evident by “significant” number of available software and frequent new releases and upgrades. Discussing this landscape with a friend, he made an insightful comment that today’s CFD software (commercial or open source) seems to be about 10 to 15 years behind the currently available FEA tools in terms of maturity. Now, I don’t know enough about FEA software to gauge the accuracy of this comment. However, with some certainty, I can declare that the present CFD tools cannot be considered “mature” at this time. I can use RANS turbulence modeling as case in point. In my introductory turbulence modeling training course, I make sure that participants understand that there is no “one size fits all” when we consider turbulence modeling and a great deal of pre-work is required of the user if one is to apply turbulence modeling for a new application. For example, for new users, literature search is extremely helpful to find the most appropriate model for their class of problem. For more advanced users, the pre-work may include numerical experimentation and fine tuning of the chosen model along with validation against empirical data.
I also think that this “maturity” may not come anytime soon. This is evident as new features, techniques, and methodologies are required as CFD is increasingly being considered for new industries and more complex applications. These requirements are certainly having an effect on the rate of advances in CFD. So, what does this “lack of maturity” and rapid advances mean for CFD software makers and more importantly users? Well, I believe they make the present a very exciting time. The relative “youth” of CFD as a discipline as well as the competitive landscape for CFD software would ensure there would be plenty of choices for users in methodologies and software for some time to come. On the other hand, this foray into new territories results in much greater number of new users which can be somewhat disconcerting. Surely, having more new users can translate into greater growth and more rapid progress. However, significant number of new users can also mean that a major portion of these new users, some of them with no real engineering background, can have unrealistic expectation of CFD and wouldn’t really understand the applicability and challenges of CFD. Therefore, I believe it is imperative that these new users learn as much about the accuracy and limitations of CFD tools as soon as they begin to consider CFD.