Make CAE Mainstream

By John F. Mannisto, Whirlpool Corporation

I recently asked a couple college interns a question: “How do they teach you CAE?” Their answers: “What is CAE?” “One of my professors offers a class in FEA every other year but it’s not very popular.”

Not very popular? Yikes! (and these were good Michigan engineering colleges).

A couple years back I did a webinar with engineering professors. They wanted to know what they could do to better prepare their students for industry. I gave them my decidedly biased opinion: make computer-aided engineering mainstream.

Part of the discussion centered around how CAE material is taught. It is my belief that there is too much emphasis on the inner workings of a computer program and not enough on the practical application.

A calculator is a well-accepted engineering tool. We were taught: “Push 1, then +, then 1, and you get 2.”

Imagine if I taught the use of a calculator the same way CAE is taught: “The main processor is connected to the input device which is a series of keyboard sensors. The liquid crystal display (LCD) is often mounted to the main circuit board. More on the LCD in chapter 12.”

Crazy, right? Yet grad students struggle with stiffness matrices and shape functions while undergrads run away from any class that has three letters in it: CAE, FEA, CFD. This is a big miss.

Build a Foundation

I have to admit that in my grad classes I had to struggle through a few of these things. Euler, Bernoulli, Fourier, Laplace all inflicted their pain on me — it was how they measured engineering students. But the “real” education came in my many years as a CAE analyst, staring at stress contours, animations and flow fields. I can’t derive the beam equation anymore, but I understand how to identify bending stress and membrane stress. I don’t remember how upwind differencing schemes work, but I understand flow recirculations and separations.

I would expect my story to be familiar to those who grew up with punch cards, 9-track tape readers and spaghetti FORTRAN. CAE was for the professors who huddled around the “super” computer and could interpret those reams of green and white printer paper. This was not for the faint of heart.

Today, products like ANSYS Workbench and CDadapco’s STAR-CCM+ have minimized the need to understand every nuance of the model building process. They are moving to the point where the engineer need only understand the physics of his/her problem, and let the software do the rest. Fourier, Dirchlet, Navier and Stokes can be casual acquaintances instead of daily collaborators.

It was once a badge of honor to write your 10,000 line FORTRAN code with time-step integration, but now the 1-D codes like LMS Amesim or Dymola/Modelica require only a good definition of the operating domains. Runge and Kutta can remain behind the scenes.

Take a Holistic Approach

So I repeat my plea to the universities: make CAE mainstream.

Teach CAE not as a course, but instead build it into the entire curriculum. Teach the beam equation, teach Mc/I, then apply it using finite element analysis. Teach continuity, momentum and Navier-Stokes, then show how to define an entrance condition and a pressure boundary in a computational fluid dynamics program. To be fair, my brief survey of university offerings revealed some clear movement in this direction, but I found it to be the exception, not the rule.

I get feedback that professors do not want to teach software usage, they want to teach fundamentals. I agree. If you teach poetry, teach poetry, not Microsoft Word. If you teach accounting, teach accounting, not Excel. Teach the fundamentals, but make the tools available. A $5 memory stick with CAELinux (an array of open-source codes), and a pile of YouTube videos can easily solve this problem. I also suspect many of the software vendors are anxious to provide free/low cost student versions of their commercial codes. They have a vested interest.

While CAE is growing, the rate is limited by the market — by the number of “true” CAE analysts. Democratizing CAE can expand this market 10-fold. The vendors know this, and have been working to make their products as mainstream as possible. Now they just need the market to arrive, and that will come from the universities.

John F. Mannisto is engineering director, simulation based design at Whirlpool Corporation. Send e-mail about this commentary to [email protected]. 

More Info

• Whirlpool Corporation