Mark asked:

>The parachute and the air bags were problems. Shows the value of testing
>which was short changed on several failed missions. I was surprised however
>that the parachute fix seemed to be more of a trial and error nature than
>design. The air bag fix seemed that way too. I'm curious if Gary or anyone
>else familiar with NASA can shed some light on how much of mission
>development is trial and error based on educated guesses and how much is
>detailed design calculations. Gary's comments usually indicate a large
>amount of calculation in his area.

The area I work in is mainly involved with heat shields and hypersonics. Hypersonic flow is a bit unusual because it is very susceptible to engineering analysis (pre-computer, 1950s era type methods). Its tractibility is the main reason why I like this field. Computers can take a lot of the fun out of engineering because everything becomes a "black box". The basic method in my area is "conservative simplification". The question I have to ask: Can I simplify the problem such that it becomes tractible to yielding an inaccurate but conservative solution that I know is safer than the insolvable exact solution? This design philosophy works with hypersonic aeroshells but doesn't apply to parachutes and airbags. The flow associated with parachutes and airbags is so complicated that Computational Fluid Dynamics (CFD) is not trustworthy. People do CFD solutions anyway but the results are basically eye candy for the managers. For parachutes and airbags you've got no choice but to run experiments, typically in wind tunnels. There were some last minute issues with the parachutes for MER (people were afraid the chutes would shred). Experiments were done at NASA Ames 40x80 wind tunnels with new parachute designs that demonstrated the designs were safe. The 40x80 wind tunnel has since been decommisioned and its crew reassigned or laid off (there is no other facility like it in the world). Along the lines of this happy note, the parachutes for Pioneer Venus also shredded but were fixed at the last minute using experimental methods. However most of the lessons learned from Pioneer Venus were not passed down to the MER parachute design crew because the engineers involved had long since retired or died. During Dan Goldin's tenure as NASA's head, many old NASA guys were forced into retirement for purposes of budget reduction. Most of those old guys took decades of engineering knowledge with them. Dan Goldin did a real good job of giving NASA a lobotomy.

>I also agree that NASA should use expertise from other areas of
>government. I wonder if Larry, Gary, or some other insiders from the air
>force or NASA can shed some light on whether there are barriers to this
>kind of exchange or not.

We have information exchange with the guys at Sandia who design "baby killers" (ICBM RVs). We envy the guys at Sandia because they have five times the resources that we do. The guys at Sandia envy us because our problems are often times more interesting and we don't have to kill anyone. Most of the technology that I work with was derived from ballistic missile research and development. General Electric was paid hundreds of millions of dollars to develop the Mark 6 RV for the Titan II ICBM. The W53 warhead housed inside the Mark 6 could vaporize a city the size of San Francisco in a single flash. Most of GE's RV technology developed for the Mark 6 was directly applied towards the design and fabrication of the Pioneer Large Probe and the Galileo Probe aeroshells. Almost everything in my trade is dual use technology.

Gary