From The Project Manager

Galileo's arrival at Jupiter on December 7th was a tremendous success. Eighteen years of dogged tenacity and imaginative engineering solutions to some of the toughest technical and political problems ever faced by a project had finally paid off. Galileo's success is truly a triumph of the human spirit and creativity. Entering the Jupiter atmosphere is by far the most difficult planetary entry in our solar system. Galileo did it flawlessly. The Orbiter released the Probe last July for its five-month, solo, unguided ballistic flight to the Jupiter entry corridor. The Orbiter aimed the Probe so accurately that the Probe used only about 15% of its entry corridor margin. The heat shield protected the descent module from outside entry temperatures that reached 25,000 deg F at the stagnation point in front of the Probe and all elements of the Probe withstood a 230 g entry structural load. All seven of the Probe's scientific instruments worked. Our requirement was to reach a pressure depth of 10 bars---the Probe transmitted data to the Orbiter continuously for 57.6 minutes reaching a depth of 23 bars! The Relay Link began at four minutes after entry, so transmission ended 61.4 minutes after entry. The original mission design ended the Relay at 60 minutes after entry---the Orbiter propellant savings achieved by the VEEGA trajectory allowed us to extend the Orbiter support of the Link to 78 minutes after entry to ensure getting every bit (literally) of data we could from the Probe---we did! The Probe descent mission actually started about a minute late due apparently to a wiring problem with the g-switches that told the Probe when to deploy the parachute. The Probe stopped transmitting when the transmitters got too hot. The descent module's inside temperatures were much closer to the outside temperature during descent than we had expected.

The Orbiter has now sent to Earth essentially all of the Probe data. Only some mostly redundant, confirming data remains to be played back from the tape recorder. Beginning two days after arrival and until mid-March, playback of the Probe data, both from the abbreviated direct computer memory storage and from the Tape Recorder, has been the Orbiter's main task. With an average data rate of ~10 bps, it takes ten hours to playback one minute of Probe data! Our effective data rate for the Orbiter tour will increase over one-hundred-fold when we load the new flight software in May.

In the last Messenger we described our plan to take a color image of Jupiter two months before arrival---this would be our only image returned before the new software was on board. We were dealt a terrible blow on October 11th just after taking this image. The Tape Recorder would not rewind. Telemetry data showed that the Recorder was running but the tape wasn't moving. Every imaginable explanation was that the Recorder was broken and unrecoverable. The Tape Recorder was the primary means for getting the Probe data; without it only the abbreviated data set to be stored in the computer memory would be available. And, the Orbiter mission without the High Gain Antenna was crucially dependent on the Recorder---there would be no images without the Recorder!

Now, two months before arrival, we urgently jumped into three new efforts: troubleshoot the Recorder, figure out how to somewhat recover the Orbiter mission without it, and extend the computer storage of Probe data. This was when we had hoped to better prepare for contingencies that might occur within a few days of Jupiter arrival.

Incredibly, on the very same day the spacecraft Recorder failed to rewind, an identical recorder in our ground testbed malfunctioned. These turned out to be totally separate problems, but the coincidence greatly compounded the troubleshooting. The testbed unit had broken---the tape was pulled off the reel due to a circuit failure. Within one week the troubleshooting tiger team had determined that on the spacecraft the tape may just be stuck and slipping on the capstans. On October 20th the spacecraft Recorder was commanded to move the tape forward for a few seconds---it worked! We were still a very long way from knowing what caused the problem. At this point I made the most painful decision of my career. Until the Probe data was returned to Earth, we would use the Recorder only to record and playback the Probe data because of the paramount importance of the Probe data, and Probe data record and playback required only the most benign use of the Recorder. Since recording the absolutely unique fields and particles data in the Io torus was equally benign, we added that recording at no significant risk. All imaging and other high-rate data had to be eliminated from the arrival sequence. Independent of the Probe considerations, it would have been foolish to risk damaging the Recorder before we could determine how to use it safely at the imaging rates (high tape speeds) for the ten satellite encounter Jupiter orbital tour. (The loss of the high-resolution Io images was the biggest disappointment, but we may go back in an extended mission to get them.)

We have conclusively determined that the tape is sticking to a "dummy" erase head that is used as a tape guide in our Recorder. This sticking results in loss of tape tension and consequent tape slipping on the capstans when trying to run in reverse. After months of vigorous effort, the exact phenomenon that causes the sticking is still unknown, although we have some good candidates. In mid-March we held a two-day Workshop at JPL with a broad spectrum of tape recorder industry experts. They all concurred with our findings. Since the cause of the sticking is not known, we do not know how to prevent it. Accordingly, we are implementing an operational strategy that will always pull the tape forward to break the stick before running in reverse. We have done some limited testing and conditioning of the Recorder on the spacecraft, and it was very revealing. More testing is planned to validate the operational strategy. The Recorder has worked perfectly for all the Probe playback operations. Exhaustive ground testing of flight-like recorders is continuing.

Our Phase 2 Flight Software has been augmented to directly control some Recorder functions previously done autonomously by the Recorder. This will prevent the catastrophic failure that occurred in the Testbed Recorder. The new software also provides for detecting a stuck tape and stopping the Recorder in that event.

A very notable accomplishment was achieved by the Phase 2 Development Team within just two days after the October 11th anomaly. They devised a scheme for obtaining images by buffering them through the central computer without using the Tape Recorder. While this method is not anywhere near as effective as using the Recorder, it is infinitely better than no images at all, and it was a priceless salvation when we thought the Recorder was broken. It was so remarkably good that we suspended the Phase 2 development for one month to complete the preliminary design even after we believed the Recorder could be recovered. We plan to resume this development after the Phase 2 software is installed on the spacecraft just in case the Recorder does fail during the orbital tour.

The decision to eliminate the imaging and other high-rate recording operations required a complete rework of the spacecraft approach and arrival concurrent sequences. The rigorously bulletproofed and tested Relay/JOI critical sequence required only one-for-one replacement of five commands to accommodate the Io torus recording, which in turn required several real-time ground commands for complete fault protection.

The greatly reduced concurrent arrival sequence enabled the expansion of the Probe data storage into the now unused sequencing memory so that the Probe symbols could be stored single-string for 73 minutes.

Elimination of the imaging also eliminated the optical navigation on Jupiter approach. The combination of superb DSN Doppler tracking and an ingenious Navigation Team strategy enabled the successive cancellation of the three approach Trajectory Correction Maneuvers, the Orbit Insertion delta-V update commanding, and then an essentially perfect JOI performance enabled canceling the two post-JOI Orbit Trim Maneuvers (OTMs). The strategy, which ultimately advanced the first in-orbit satellite encounter (Ganymede-1) by one week, was developed to minimize the size of the OTMs-it surely did. We went ballistic from JOI cutoff to Apojove.

The Orbiter performed the Io torus recording, the Relay Link, and the JOI flawlessly. The star scanner became radiation saturated as anticipated, but there was no other radiation-induced disturbance as Galileo passed through by far the most intense radiation it ever will. Unlike Voyager, there were no Power-On Resets (PORs). Clearly, Galileo's designers did an excellent job in making it radiation hard.

The Perijove Raise Maneuver (PJR) was performed on March 14th raising perijove---the lowest point of the orbit---to 715,000 km above Jupiter to withstand the radiation for the upcoming eleven perijove passes during the orbital tour (our first perijove at Relay/JOI was only 215,000 km). This was the third and final burn of the 400-N main engine. The 378-m/s maneuver---over half the size of the JOI---just about doubled Galileo's orbital speed at its then farthest point from Jupiter (apojove) and used two-thirds of the propellant that remained on board after JOI! Recall that since the Orbiter Deflection Maneuver last July we have been concerned that the helium pressurant check valve on the Oxidizer side of the propellant system may be leaking. We autonomously isolated the helium supply at the end of PJR to "trap" the Oxidizer pressure below the Fuel pressure to eliminate the threat of Oxidizer vapor migrating to the Fuel side. Unfortunately, the Fuel check valve apparently malfunctioned during PJR so now the Ox pressure is higher than the Fuel pressure and the threat of migration remains. We are working with our German colleagues (DARA/DASA) to try to prove that the Ox check valve is now, in fact, closed and holding the higher pressure. Until we prove it or prove that Ox vapor migration to the Fuel side cannot rupture the Fuel pressurization line, we must maintain near-constant electrical power margin because the excess power is hardwire-shunted to the propellant tank heaters. This further complicates our orbital operations.

The Project Team is now focused on completing the ground testing of the new orbital phase flight software, planning the in-flight loading process, and preparing for Galileo's first in-orbit encounter---Ganymede-1---on June 27th. Galileo is being programmed to return the first of our images from the Jupiter system beginning in mid-July. Let the Tour begin!!!