Image of Asteroid 5535 Annefrank taken by the Stardust spacecraft just prior to closest approach.

November 4, 2002 - Late Friday evening Pacific time on November 2, 2002 at the Jet Propulsion Laboratory (JPL) in Pasadena, California, and at Lockheed Martin Space Systems - Astronautics (LMA) near Denver, Colorado, the NASA STARDUST flight team pulled off a tremendously successful close flyby of the main belt asteroid Annefrank. This flyby was used as an engineering test of the ground and spacecraft operations that will be implemented at the primary scientific target, Comet Wild 2 (pronounced "Vilt" 2) just over one year from now.

STARDUST is a low-cost Discovery Mission that continues to perform as expected after more than three and a half years into a planned seven-year mission to rendezvous with Comet Wild 2 in January 2004. STARDUST will collect cometary dust samples, flowing from the nucleus just hours before spacecraft flyby, and return the samples to Earth in a Sample Return Capsule in January 2006. The close flyby of Annefrank offered a unique opportunity to thoroughly test all planned operations on the spacecraft and ground support operations which will be used during the rendezvous with Comet Wild 2.

Stardust team waits for data to arrive from the Stardust spacecraft in the Mission Operations Room

"We performed a full dress rehearsal with the cometary dust collector deployed as we flew STARDUST within 3,300 kilometers of Annefrank," said Professor Donald Brownlee, the project's Principal Investigator from the University of Washington. "The spacecraft was poised in its flyby attitude with all the science instruments on. The flyby has exceeded all of our expectations and provided us with unexpected data about the asteroid," said Brownlee.

The approach geometry to Annefrank was much more difficult than will be the case for Comet Wild 2. The spacecraft was pointed over 60 degrees off of the normal Sun and Earth pointing attitude and was running on its battery in order to attempt to detect and capture images of Annefrank.

"The spacecraft performed every command perfectly and did everything asked of it," said Allan Cheuvront, Spacecraft Engineer at Lockheed Martin Space Systems near Denver. "We are thrilled with how well the entire operation went. We couldn't have asked for better performance from STARDUST and the images it captured of the asteroid exceeded everyone's expectations. The spacecraft's pointing, attitude and flight operations were excellent. This really adds to our level of confidence about how well the spacecraft will perform when we reach Wild 2," added Cheuvront. Cheuvront and a team of engineers at Lockheed Martin's spacecraft control center, known as the Mission Support Area, control the spacecraft in conjunction with JPL and the Deep Space Network.

The Navigation Camera was straining to see Annefrank during approach. "This camera was operating at its limit of performance and seeing very dim stars down to about 11th visual magnitude", said Ray Newburn, the Lead Scientist for the camera at JPL.

However, the brightness predicted by Drs. Stephen Synnott and Donald Yeomans of JPL was dimmer than 11th visual magnitude. "We tried everything we could think of including taking multiple long exposures and adding these on the ground", said Dr. T. S. Mike Wang, Optical Navigation Specialist at JPL, "but Annefrank was not cooperating. It was just too dim."

Because of the high probability of not seeing Annefrank during the approach, the flyby was designed to be successful without having to see it up to 20 minutes from encounter. "A flyby distance of 3,000 km (1,864 miles) was chosen so that there was no risk of the spacecraft flying near any possible dust environment or small satellites of Annefrank", said Ed Hirst, JPL Mission Design Manager. "We also wanted to ensure that Annefrank would be in the camera view at the start of the encounter sequence," added Hirst.

Since Annefrank was not seen in the approach images, the flight team felt that the asteroid was at least as dim as predicted and possibly even dimmer. The team decided to send up a new encounter configuration file and set the initial flyby exposures longer. "We had a planned uplink six hours before encounter for this very purpose," said Robert Ryan, Mission Manager at JPL. "We had some communications problems the day before that gave us some difficulty, but NASA's Deep Space Network gave us highest priority, and excellent communications on Friday, allowing us to play back earlier images we missed as well as sending our final encounter commands," added Ryan.

At 8:00 pm (PST) Friday evening, communications were established with the spacecraft to watch its pre-loaded sequence command turn the spacecraft away from the Sun and Earth into its flyby attitude. "We have built up over three years of flight experience and a tremendous amount of confidence and respect for our spacecraft to perform such operations routinely," said Joe Vellinga, STARDUST Program Manager at Lockheed Martin who led the development and manufacture of the spacecraft. "The spacecraft did not miss a beat during its flyby and it maintained all critical thermal, power, attitude, memory and reserves at or above design levels," added Vellinga.

The main function to be tested during flyby was a sophisticated flight computer program that would take over control of the spacecraft to keep the camera view locked on Annefrank during a 25-minute period around its closest encounter. "This software was a derivative of the nucleus tracking software successfully flown on the Deep Space 1 (DS1) flyby of Borrelly," said Dr. Shyam Bhaskaran, developer of the algorithms at JPL. "Based upon my previous experience on DS1, it performed up to my expectations with this encounter at Annefrank with over 60 successful images having Annefrank right in the middle of each image," added Bhaskaran. David Gingerich, Flight Software specialist at LMA who implemented and tested the nucleus tracking software said, "its performance was executed just like the coach drew it on the blackboard."

Over 70 encounter images were obtained that show a typical small solar system body, highly irregularly shaped and cratered. Annefrank is about twice as large as predicted, at least 6 kilometers in diameter, but darker than expected and therefore more difficult to detect in the early images. Not only did the camera perform well but the University of Chicago Dust Flux Measurement Instrument (DFMI) and the German Cometary and Interstellar Dust Analyzer (CIDA) performed as expected.

Professor Tom Economou, DFMI scientist from University of Chicago, stated "we ran for 28 minutes as we will at Wild 2 with DFMI performing all expected functions". Dr. Jochen Kissel, Lead Scientist for CIDA from Max Planck Institute in Garching, Germany, said "I will be able to put CIDA into an even better configuration at Wild 2 based upon the Annefrank experience." Both dust instrument teams are combing through their data to see if by chance they may have seen a dust particle.

Stardust Project Manager Tom Duxbury (left) discusses Annefrank data with Camera Lead Ray Newburn

"Performing such flight testing before the primary encounter is a critical part of reducing risks and significantly increasing the probability of success when we reach Wild 2", said JPL Project Manager, Thomas Duxbury. "We have performed exhaustive testing and training with LMA at their spacecraft test laboratory and through flight simulations, but these cannot totally replace actual flight operations testing. We learned a lot that will improve our operations at Wild 2 based upon the lessons learned at Annefrank. The bottom line is that if Annefrank had been Wild 2, we would have succeeded in every respect," added Duxbury.

"I applaud the entire flight team," said Don Brownlee. "We could not have asked for more, except possibly for Annefrank to be a little brighter. However, for everything that we could control with the spacecraft, we were nearly perfect.

Even though this was an engineering test, the flyby with Annefrank provided new information previously unknown about the asteroid about its size, shape, spin state and brightness as a function of viewing angle.