UPDATE # 3 - December 4, 1997
Here is a simple activity that will work on the communication skills of
WEB CHAT WITH BRENT WELLMAN
The best part of chats is being in touch LIVE with an expert, so JOIN US for an exciting hour of give and take about helicopters and related subjects. Our upcoming chat will be: Wednesday, December 10, 10:00-11:00 a.m. Pacific Time Brent Wellman, Deputy Project Manager, 2GCHAS Brent is a project manager for a computer program that makes it easier to design helicopters and similar aircraft. Please read Brent's biography and journal prior to joining this chat. http://quest.arc.nasa.gov/aero/team/wellman.html To participate (ask questions), you will need to pre-register for the chats. Go to: http://quest.arc.nasa.gov/aero/chats
A CATASTROPHIC STRUCTURAL FAILURE
Brent Wellman's Field Journal
October 29, 1997 This morning I called an engineer for one of the major helicopter companies (Maybe you have heard of them -- Sikorsky) to check on a problem he had been having with the new release of our computer program for helicopter analysis, 2GCHAS (pronounced "TOO-GEE-CHARLEY). His problem had evaporated, so I set out for more pressing matters... I wrote a section of a report on an incident that occurred in one of our wind tunnels. A major aerospace company (perhaps you've heard of them -- but you didn't hear it from me!) had been testing a stopped rotor concept (the rotor stops spinning to become a [fixed] wing!). Without warning, the rotor experienced a catastrophic structural failure and disintegrated, sending debris in every direction. I was named to an investigating board and had to use my knowledge of physics, and engineering materials to piece together what had occurred. This report section pretty well puts my participation to rest, but I remain on call (as long as we push the limits of technology, we will have accidents). Back to computer stuff...I worked on a persistent problem with an upgrade to 2GCHAS that I have been working on. It is a mainstay of modern physics that physical laws do not depend on whether the physicist stands on his feet or his head (we say the laws are independent of observer's reference frames). It is rather handy for engineers to use special reference frames (usually X-Y-Z coordinate systems, if you've had that much math) to describe the items he is working with. To do this in our program, one must define the reference frame with respect to a standard one, by defining a "direction cosine matrix." It is complicated to do (and frought with errors for all but the simplest transformations), but, once done, the computer-generated "physics" in one coordinate system immediately translate into [consistent] "physics" in the reference coordinate system. My improvement would allow an engineer to say to the program, "slide the reference coordinate system to *here,* then turn it so, twist it so, roll it so," and the computer figures it all out. It is provable that ANY transformation, no matter how cockeyed, can be specified this way. Trouble is, the computer keeps dropping out a step and improperly transforming some of the various reference frames. I figured out why the program was dropping out the extra step and cured the program sections I had been working on. A little more testing and I'll be able to put this to bed. A problem came up with one of the versions of the most recent software. The code was compiled with the DEBUG option turned on. As a result, the code worked fine, but the computer was spending so much time keeping track of what it was doing, that it only spent a little time actually doing it. So the program runs about four times slower. I notified the affected customers by phone and had the systems operator generate new code (with DEBUG off). After lunch and a workout (whew!), I looked into safety concerns over a new rotor test in one of our wind tunnels. This is an all-government test, so I can talk to the project persons directly. I contacted the principal investigator of the wind tunnel test following that one about my acting as backup pilot for him. Rotors in wind tunnels have to be "flown." They can't just spin in the wind. They need someone to control their collective (which governs the amount of lift a rotor develops) and cyclic (which governs the propulsive force generated) controls. Anyway, this project's pilot is nearing retirement, and he may need to be replaced. I hopped back to the computer and checked up on one of our programmer's work. Each of us on the team checks (we say "certifies") the programming changes that are made by the others. In this way, we look over each other's shoulders, so to speak. I actually found an error in her work (a change she said she made which was overlooked), brought it to her attention, and had her correct it. I documented my certification of the software units and went on to other things. I set up a visit by an engineering VP for a small technology company (no, you haven't heard of them) to Ames Research Center to see me and our chief scientist about a -- umm, ...device he is developing for us. I can't say a lot about that project, except that it is part of the "peace dividend" you may have heard about and it will REVOLUTIONIZE vertical flight if it works. I act as technical monitor for the contract, and conduct "technical surveilance" of the work. Incidentally, I was able to set up the clearance for his visit all on our new, secure Web site (no, you may NOT have the URL!). Cool. That's the day. I'm off to the day care center to pick up the tyke, and to have dinner, bye.
SUBSCRIBING & UNSUBSCRIBING: HOW TO DO IT!
If this is your first message from the updates-aero list, welcome! To catch up on back issues, please visit the following Internet URL: http://quest.arc.nasa.gov/aero/updates To subscribe to the updates-aero mailing list (where this message came from), send a message to: email@example.com In the message body, write these words: subscribe updates-aero CONVERSELY... To remove your name from the updates-aero mailing list, send a message to: firstname.lastname@example.org In the message body, write these words: unsubscribe updates-aero If you have Web access, please visit our site at http://quest.arc.nasa.gov/aero