Lessons in developing the JSF: sometimes bringing together physical and digital doesn’t always turn

By Christian Kelley

The opening presentation was a really good continuation of the first days theme on integrating the virtual and the physical. Paul Bevilaqua from Lockheed Martin talked about the development of the Joint Strike Fighter (aka the JSF or the F-35 Lightning II).  Maybe I was so in to this since it combined to things I can really get interested in: high performance fighter jets and CAE.  The JSF is a really interesting aircraft as it is intended to replace 3 different aircraft for 3 different services (the F-16 for the Air Force, the F/A-18 for Navy and the AV-8B Harrier for the Marines) the with a single model in 3 configurations.


Lockheed Martin has a long and storied history of developing cutting edge aircraft in their skunkworks.  They developed America’s first jet powered airplane, the first plane to pass Mach 2 and Mach 3 and the first stealth fighter (the F-117).  What I didn’t know about the skunkworks is that it is as much a design philosophy as it is a program or physical place.  Paul best captured the skunkworks philosophy when he said:

A good design tested this year is better than a better design tested next year

This philosophy drove LM to design new planes by getting them to an 80% complete state ‘on paper’ as quickly as possible and then figure out the last 20% using real world, physical prototypes.  This approach was successful for many years, but “you can’t live with that much risk today” so advanced digital tools that could more accurately predict real world performance were added to the program to get closer to 100% before putting someone at the stick.

At this point, most writers would go on to explain how LM was able to use computer simulations to eliminate all vestiges anything physical from their development process, making a production ready version of the JSF with the first metal that was formed.  But that’s not the part that I found most interesting about the story that Paul told.  While it’s true that computer models where used to do all sorts of complex predictions (from determining if the exhaust gasses would bounce off the ground and melt the aluminum skin of the JSF to whether the wings would flutter to the point of breaking) the interesting part to me was the new things they were able to do physically because of the extra time and precision that the use of the digital models added to the process.

For example, they were able to use digital representations of internal parts to create physical versions using stereo lithography.  Then they let their mechanics put the parts together and do physical routing of the piping and wiring to make sure they got it right.  It seems that simulating the routing is still a task that might be a little too complex for computer simulation, as the Airbus A-380 team can likely attest.  This is a story that often isn’t told: being able to do some evaluations digitally gives more time to do other things physically.  It’s this combination of physical and digital testing that was one of the prime reasons that LM had the edge in winning the JSF contract.

Proof of their success:

  • The JSF-B (configuration for the Marines) was the first single aircraft to fly supersonically, hover and then land vertically.

  • The JSF-A was converted to the JSF-B by removing a fuel tank and adding the forward vertical fan.  How’s that for modularity and reuse!

  • LM has received advanced funding to produce 465 combat ready aircraft over the next 10 years.

  • The team that developed the JSF received the Collier Trophy, the equivalent of the Pulitzer prize for aerospace and aeronautics.

The end of the presentation featured some great video of the various configurations of the JSF doing it’s thing.  Great eye candy, but it was much more meaningful knowing a lot more about the story behind what makes it all work.

Technorati Tags: plmconam08 jsf

Leave a Reply

This article first appeared on the Siemens Digital Industries Software blog at