|
   |
 |
                 
|
UPDATE #57 - November 2, 1998 PART 1: It never gets old! IT NEVER GETS OLD!
Did you share the excitement with us on Thursday as Discovery launched amid a media frenzy? Were you able to be one of the thousands of people who viewed the Quest webcast with Brandt Secosh, Karen Dodson, and Mike Ciannilli bringing to you the sights and people that were present at the Press Viewing Site? I hope so. Even here in our humble trailer on the back lot at NASA Ames Research Center, we drew the attention of the press. They curiously watched as we brought you coverage over Learning Technologies Channel, allowing you to interact by feeding your questions to Florida through the QuestChat room. Would you like to see their coverage? http://quest.arc.nasa.gov/ltc/sto/launch/sts95/kron.ram or if you have a really fast machine: http://quest.arc.nasa.gov/ltc/sto/launch/sts95/kron200.ram Below, Karen shares her first impressions of Kennedy Space Center. I'm anxious to have her write a new journal now, after the launch. As I heard a newsman say, "Seeing a launch in person is something everyone should put on their list of must-do's." This was Karen's first, but I hear from the local residents, it never gets old. 'Til next Update, Linda HOW TO GET YOUR QUESTIONS ANSWERED
Your students and you can have your questions answered by the experts. But first... be sure to read carefully the overview and directions on: http://quest.arc.nasa.gov/space/ask/question.html Sometimes the answer to a question is already available on the web and I can see the makings of a terrific scavenger hunt online. One source of existing answers listed in categories is the Space Team Online Q&A archive at: http://quest.arc.nasa.gov/space/ask/index.html From the questions we received most frequently during the launch program, I would like to post a few more existing sources of answers to frequently asked questions: For questions about how to become an astronaut, see http://shuttle.nasa.gov/factsheets/asseltrn.html For questions about Senator John Glenn, see his own website at http://www.senate.gov/~glenn/ If you have a question about the experiments to take place on STS-95, see: http://www.lifesciences.nasa.gov To follow this flight and get information regarding day-to-day account with pictures, see: http://www.shuttle.nasa.gov or http://www.ksc.nasa.gov/shuttle/missions/sts-95/countdown.html UPCOMING CHATS
->Wednesday, Nov. 4, 9 AM Pacific (Noon Eastern) Sten Oldenwald is an astronomer from Hughes/STX in Washington, D.C. and a new team member for our Space Scientists Online project. For the month of November, he will chat and answer questions about how solar storms effect our lives, how the sun will die, and prospects for life on other planets. See Sten's biography at: http://quest.arc.nasa.gov/sso/team/oldenwald.html ->Thursday, November 5, 1998, 9 AM Pacific Time (Noon Eastern): Sherri Carlson, electrical engineer at Kennedy Space Center, is responsible for checking the designs of the systems to make sure they will work together with the equipment they control. See her profile at: http://quest.arc.nasa.gov/space/team/carlson.html ->Thursday, November 12, 1998, 10 AM Pacific Time (1 PM Eastern): Ric Adams, project engineer at Kennedy Space Center, is involved in inspecting the space shuttle for any minor damage before, during, and after a mission. See his profile at: http://quest.arc.nasa.gov/space/team/adams.html ->Thursday, November 19, 1998, 10 AM Pacific Time (1 PM Eastern): Rick Pettegrew works with a team that analyzes the characteristics and behavior of fire. Rick and his team try to better understand the science by performing experiments in reduced gravity environments. Read Rick's profile at: http://quest.arc.nasa.gov/space/team/pettegrew.html [Editor's note: Karen works with NASA Life Sciences Education Outreach and was with us for the Challenge Project in Key Largo, coordinating crew members. Most recently, Karen co-hosted our "And now from the launch" webcast.] ARRIVAL, KENNEDY SPACE CENTER October 26, l998 I flew into Orlando last night and got to my hotel around 8:00 p.m. Flying into Florida is like time traveling into summer, you leave in the fall and arrive in the summer. It's 80 degrees out today, beautiful blue sky with white puffy clouds, and the land is as flat as can be. As I drove into KSC this morning, I went over a bridge spanning the Banana River. The best views around here are from bridges. When I was at the top, I happened to glance over (I was driving parallel to KSC) and shining in the sunlight was the VAB [Vertical Assembly Building]. It looked huge, the largest thing in the entire landscape. It dominated everything around it, trees looked like blades of grass compared to its mighty size. You never realize how different something feels in person as compared to a picture until you are there. I had seen pictures of the VAB before, but when I actually saw it live, I gasped. I said out loud (and there was no one in the car with me), "Hey look, it's the VAB!" It was then that the first wave of excitement washed over me. Shortly after, I was pulling into Kennedy. The first badging area I came to was the press pass area. The parking lot was full of TV vans and the line into the building was 30 people long. Reporters and camera people were all over, reporters in perfectly tailored suits with every hair in place, and the camera people looking the yin to their yang. When you see one reporter on TV it doesn't look that strange, but when you see about 50 of them in power suits, it's kind of scary. After learning the lay of the land through a series of wrong turns, Brandt, my KSC cohort, came to the rescue and led me to his office. As we passed through the various gates, cars were backed up waiting to get in. Tour buses are all over, and there is a definite energy in the air. Driving across the causeway, Brandt pointed to the left. Off in the distance was the space shuttle, live and in person, waiting for its journey to space three days from now (today is Monday, Oct. 26th). It really makes you feel proud and part of something greater when you see it out there. It is absolutely amazing to think of the engineering and years of learning that have allowed us to get to such magnificent capabilities. And that's saying a lot coming from a generation X-er! [Editor's note: Robert uses his math abilities to, as he says, "move as much simple, boring work to the computers as we can and allow the people to spend their time on jobs that increase safety, capability and efficiency of the shuttle mission. Below he describes a process used to test for leaks.] IS THERE A LEAK? October 16, l998 Hi again. It has been a long time since my last journal, which is because only a small part of my work involves supporting Shuttle operations. My last journal was about a project that I came into as a consultant. The idea was to come up with computer programs that could monitor the fuel system of the Shuttle Auxiliary Power Units (APUs) to make sure that none of the propellant was leaking. This is how the project ended up. My specialty is mathematics, so I'll leave the detailed description of the APU system to the experts. The part of the system that we had to understand was the fuel tank. Because the APU system must operate with or without gravity, and since the fuel is a liquid, the tank has a special design. Unlike tanks on earth that use gravity to hold their contents (water, gasoline, etc.) in place at the bottom, the APU fuel tank is fitted with an elastic "balloon" that holds the liquid fuel. The metal tank holds nitrogen gas under pressure, which squeezes the balloon and forces the liquid into the APU. This peculiar arrangement works very well, but has one big problem. There is nothing in the system that would correspond to a fuel gauge so you would know how much propellant is left in the balloon. Your gas gauge on a car works by sensing the position of a float in your gas tank. Remember, no gravity, so no float. Well, we still need to know how fast the APUs are burning their fuel, so instead of a fuel gauge, there are sensors that measure the pressure of the nitrogen in the tank. As the fuel is consumed, the balloon gets smaller and leaves more room for the nitrogen. As you allow a gas under pressure to expand, it becomes cooler and the pressure drops. By monitoring the pressure of the nitrogen and the temperature of the tank, the flight controllers can see how fast the fuel is leaving the tank. The problem we had was to try to write a computer program that could do what the flight controllers do with their eyes and experience. The reason this is hard is because there is a very complex relationship between the tank pressure and temperature and the amount of liquid remaining in the balloon. Furthermore, there are many other factors (sloshing, temperature variations and inaccuracies of the pressure sensors) that are very hard to anticipate in computer code, but which flight controllers learn to handle. I consulted on this project because one of my specialties is adaptive modeling. A "model" is a computer simulation that (hopefully) simulates a physical system and is used to predict how something will behave. Models are built by people who are quite familiar with the system being modeled and computer programmers who actually write the code. Because some systems are so complicated, it is hard to anticipate everything, so variables are put in the code. The program fills in these variables as the system runs, and this process allows the program to adapt or learn how better to model the system. Most computer models have some adaptive parts. You may have heard of one kind of adaptive model called "neural networks" that learns to reproduce a complicated system in a way that is at least a little bit like the way brain cells work. We tried lots of different approaches to tracking the fuel consumption of the APUs. We even tried neural networks, but the approach that gave the best results was a combination of several different methods that included some pieces based on neural networks and others that used our knowledge of mathematics and the physics of the system. We "teach" the computer program to reproduce the behavior of the system by running the model and comparing the predictions of the model with what the system really did. During the beginning of the training, the model does not agree with the real system, and this difference (called the error) is used to adjust the model. We know how various pieces of the model change its predictions. We keep adjusting those variable parts of the model until it gives us predictions that are in reasonable agreement with the real system. Once we train our models, then it's time to test them against new situations. It does us no good to make a "parrot" that will tell us back what we trained it to do but does not work correctly when faced with a new situation. This process of training and testing is repeated until either the model passes all its tests or we figure out some way to make it work better. Once we finish training and testing the model, we can use it to predict (from those pressure and temperature measurements) how much fuel has been used and how fast it is being consumed. After a lot of hard work, we were not able to fully match the skill of the flight controllers with a computer program. What was possible and practical was to give the controllers a "fuel gauge" for the APU. The computer program monitors the pressure and temperature sensors on the fuel tank and converts those measurements into an estimate of how quickly the fuel is leaving the tank. We still rely on the flight controllers to determine the health of the machinery, but the program does a lot of the complex math that converts the raw data into an estimate of the fuel consumption. I've been working on a book, in my spare time, on computer programming. I just finished another three chapters. The "book" was originally for my son who's taking a computer course, but I've tried to cover topics that most people have trouble understanding, and I hope it may be useful to others as well. Most of the examples I use are in the computer programming languages of C and C++. I've also started walking my dog, Arrow, again. He's been so bored lately. I get some of my best thinking done, especially related to work, when we take these walks. Arrow, a black lab and Doberman mix, does most of the navigating and seems to lead me, though he has no formal training as a seeing eye dog. I wonder if he could find our way home by himself? So far I haven't tested him because I'm worried that he would just keep going forever - he enjoys walking that much.
|
||
|
|
|
