March 29, 2000
Heidelberg graduate looking towards the outer reaches of space
By Cathy Willoughby
Staff Writer
The use of models and simulations can be useful in demonstrations. Especially when the actual item is in earth orbit, looking out toward the outer reaches of our universe.
This is one lesson learned during a presentation by Larry Dunham, vice president of the Aerospace Division of Jackson and Tull, at Heidelberg College Tuesday morning.
He helped the audience at Herbster Chapel visualize what it was like for him to be part of the construction and continued maintenance of the space telescope.
A 1980 graduate of Heidelberg, Dunham began working right out of college on F-117As, or Stealth fighters, as a program analyst. He was responsible for developing simulations and performing analysis for weapon-delivery systems.
In 1982, he began his work with the Hubble program when he was employed by Lockheed-Martin in California. In 1989, he moved to Maryland to Jackson and Tull, which had started an aerospace division to support Goddard Space Flight Center. He was part of the group that supported the launch of the Hubble in 1990.
During astronaut servicing missions in 1993, 1997 and 1999, he served as part of the Anomaly Response Manager task force supporting the efforts at Goddard.
He brought along models of the Hubble on Tuesday, to better explain the parts of the telescope.
"It's an optical telescope that is 42 feet long,'' he said. "It was deployed in 1990 by the space shuttle. One of the advantages of having a telescope in space is that we are not looking through the atmosphere.''
As the telescope orbits, 324 miles above the Earth, it receives energy for operation from two sources.
"It is in a 93-minute orbit around the earth,'' Dunham said. "It spends 63 of those minutes in sunlight, so it collects energy on its solar panels, and is solar powered. Batteries provide the power during the night.''
With a hand-held model, Dunham illustrated how the four reaction wheels help the telescope change direction, spinning in the direction they need to help gather the data astronomers need for their studies.
Dunham said the Hubble was unique because it was designed so that it can be repaired and upgraded by astronauts.
"The intent was for astronauts to go up every three to four years and upgrade as technology changes,'' he said. "These are scientific instruments that they have to pull out and put in, and some are the size of phone booths.''
One of the first problems occurred in 1990. When the telescope was going from day to night and back again, the solar panels were experiencing extremes of heat and cold. Every time the telescope moved to correct the problem, they would lose the fix on the stars. The panels were swapped out by a team of shuttle astronauts.
In order to fix the problems, and make sure new computer upgrades will work, Dunham said simulations are done to "test, test and retest.''
"The flight computer, or 'brain,' is manufactured by Jackson and Tull,'' Dunham said. "It has to have handles so the astronauts can carry it, and have rounded corners, so it won't snag on the astronauts' suits.''
The computer has to be able to work in an environment subjected to radiation, and with other instruments that may have an earlier computer chip. That is why rather than equipping the Hubble with a Pentium chip, a 486 was used, Dunham said.
To test these, a facility was built to copy the effects of the telescope in space, called the VEST, or Vehicle Electrical System Test.
"They built a copy of this part of the telescope,'' Dunham said. "We can send commands to this facility, and it will think it's running the real telescope. We send commands to the models that simulate what happens in space.''
To train the astronauts on how to handle repairs in the weightlessness of space, they are sent to huge water tanks in Houston. "There, they have a life-sized copy of the Hubble, the entire shuttle payload bay in an enormous swimming pool,'' he said.
The preparation is similar for the ground-control team, which runs simulations in six different rooms.
"There we are simulating a real mission,'' he said. "We are sending commands, such as an astronaut's suit has sprung a leak, or the computer won't power up. We work on it as if it's real, and you can get so into it, it seems real.''
Dunham first started serving the Hubble mission on the response management team in 1990. He said if there is a problem, he is asked to assemble a team of experts, with a broad knowledge of the whole system, to solve the problem in the least amount of time possible.
"We have a timeline,'' he added. "We have to have a status report in 20 or 30 minutes. We do brainstorming using logical thought processes with a time constraint.''
Dunham returned to his use of models, saying that often the easiest way to solve a problem is to simulate it.
"When you get into modeling, you learn the overall system,'' he said. "You need to know how it interfaces, and what is important and what is not.''
He recalled one problem involving a loose cable. One of the members of the response team ran to a toy store and purchased a set of Tinker toys to simulate the problem. It worked so well that they took it to NASA headquarters in Washington, D.C., to show others how the cable was interfering.
"In the old days, they had test pilots,'' Dunham said. "If something didn't work, they crashed and died. Now we use software models. They have a faster turn around time and we don't lose people.''