President and CEO
MSC Software Corporation
Newport Beach, CA
The foundation stones of CAE were laid in the 1960s during the race for space. Because of the obvious defense implications, this was a national imperative. It was a time of virtually unlimited government funding for advancing engineering methods — long nights and weekends, with country first and family second.
NASA used to test the giant Saturn V rocket by putting it in a 111-meter-tall building and shaking it. The engineers would go into the building and inspect the rocket for cracks. This was structural analysis in the 60s. The Dynamic Test Stand Building that was built in 1964 still stands today.
With this backdrop, Tom Butler, a visionary at Goddard Space Flight Center, had a crazy idea. Why not do the same Saturn V test on this new device called a computer! Perhaps this event, more than any other, catalyzed the CAE industry.
No matter how much on-the-ground physical testing you did, nothing could completely mimic the ultimate physical test of actually launching the rocket. Nothing can add more to the sense of urgency and responsibility than the fact that humans would be sitting on top of the rocket. Butler’s vision and NASA’s contract led to the creation of a finite element analysis (FEA) program called Nastran that delivered answers from which engineering decisions could be made in a relatively timely manner.
Following the development of Nastran, other CAE codes came into the market including Ansys (1970), Marc (1971), Dyna3D (1976), Moldflow and Abaqus (1978), Adams (1979), Fluent (1983), StarCD (1987), and others.
An amazing phenomenon of the CAE industry, like no other field of software, is that all these codes are still the core of how simulation is done in the automotive, aerospace, machinery, and electronics industries, 30 to 50 years later! In a conversation a few years ago with Dr. Richard MacNeal, founder of MSC Software, he was surprised to hear that all of these codes were still around. He told me that originally he was not interested in becoming a software company because software in the 60s only lasted one to two years before something new replaced it.
Over the years, setting up a model and getting it to run has become easier. Also, performance has greatly improved by distributing simulation jobs to tens and hundreds of cores. More and more simulation capabilities are provided within the CAD system, but this has not been a panacea. For a designer at a CAD seat, getting the simulation to run and interpreting the answers properly has been far from easy. For an analyst, CAD includes too much capability that is not needed for simulation. So, still today, simulation is limited to relatively few experts.
According to the CIMdata 2015 Market Analysis Report Series, “a significant issue with simulation is that it is generally the province of elite groups of domain experts at large companies. Democratization is about simplifying the application of the tools and making them more widely available.”
Going forward, new kinds of simulation environments are emerging. I like to call it “engineering abstraction” — environments that are abstracted from hundreds of low-level items like nodes, elements, node IDs, node numbering, and more techniques that are just a pain and disconnected from engineering concepts like parts and assemblies. Engineering abstraction will be the major driver of change over the next five years. The result — simulation that is easier, faster, and dramatically more pleasurable. There will finally be a democratization of CAE that has been 50 years in the making.
Read more Executive Perspectives.