2006 年　電子情報通信学会総合大会

BK-2-1

MUSES-C (HAYABUSA): The mission and results Jun’ichiro Kawaguchi, Akira Fujiwara, and Tatsuaki Hashimoto ISAS/JAXA 1. Introduction The engineering test spacecraft MUSES-C was launched on May 9th, 2003 and renamed in orbit as “Hayabusa”. The mission is to develop and demonstrate key technologies which are essential for the sample and return from small bodies, that is, 1. Interplanetary cruise via ion engines as primary propulsion 2. Autonomous navigation and guidance using optical measurement 3. Sample collection from asteroid surface under micro gravity 4. Direct ultra-high speed reentry for sample recovery from interplanetary orbit Hayabusa’s challenges are not limited to above four technologies. Bi-propellant small thrust (20N) reaction control system, X-band up/down communication, CCSDS compatible packet command and telemetry, duty guaranteed heater control electronics assuring heater power constraint, reaction wheel unloading via ion engines, PN-code ranging, lithium ion rechargeable battery, multi-junction solar cell, etc are also the first trial in Japan or in the world. The mission includes not only engineering experiments but also scientific observation of the target body, the asteroid Itokawa (1998SF36), of course. For this purpose, multi-band optical camera (AMICA), near infra-red spectrometer (NIRS), and X-ray spectrometer (XRS) are equipped. Navigation sensors such as a laser altimeter (LIDAR) and radio tracking data are also used to determine the gravity model of the asteroid.

20,000 hours on December 9th, 2004, though one of four engines is for redundant backup and has very few operational time. Fig.1 shows the history of ion engine operation. According to the generation power of the spacecraft, namely distance from Sun, the number of operational ion engines changed.

Fig.1 Operational history of ion engines

Fig.2 Moon from 340,000km

In the following sections, mission results by the end of 2005 are presented mainly focused on the asteroid vicinity operation, namely, approaching, rendezvousing, landing, and sampling. Fig.4 Japanese islands from 60,000km

2. Cruising to Itokawa Four ion engines on board the spacecraft worked as expected and total operational time of engines exceeded 1,000 hours on July 22nd, 2003, and

Fig.3 Earth from 295,000km

SS-22

2006 年　電子情報通信学会総合大会

BK-2-1

On May 19th, 2004, the spacecraft successfully changed its orbit by way of Earth gravity assist. Fig.2, Fig.3, and Fig.4 show Earth and Moon taken by AMICA on board Hayabusa. It became the first spacecraft in the world which performed the gravity assist, propelled by ion engines. The spacecraft was at the farthest from Sun (1.7 a.u.: astronomical unit) on February 18th, 2005 and the farthest from Earth (2.5 a.u.) on May 25th, 2005, respectively. They were also the world record of spacecraft using ion engines. Fig.5 Itokawa of different rotational phase Conventional radio-based tracking from the ground stations does not have enough navigation accuracy to approach to Itokawa, and radiooptical hybrid navigation is required. Itokawa was detected by an onboard star camera (STT) from July 29th to August 12th, 2005 and also by an optical navigation camera from August 23rd. The line-of-sight vectors to the target body provide good navigation information and the spacecraft was successfully guided to Itokawa. On September 12th, at 10:00 JST, Hayabusa made a final 7 cm/sec correction and became still with Itokawa at the distance of about 20 km from it. Rendezvous via electric propulsion was the world’s first achievement. Fig.6 Initial check-out data of NIRS

3. Global mapping of Itokawa To keep the relative position to the asteroid safely even when invisible time from ground stations, image-based autonomous navigation is needed. Hayabusa has a telescopic camera (ONC-T) and a couple of wide field of view cameras (ONC-W1 and W2), in order to detect the direction to the target body. LIDAR is used for range measurement against the asteroid surface. ONC-W1 and LIDAR are nominally used for autonomous relative position keeping. ONC-T is also used for scientific observation and called as “AMICA” for this purpose. In September, the spacecraft stayed at about 20 km altitude, and initial check-out of onboard autonomous functions was carried out. On September 30th, Hayabusa descended to the distance of about 7 km, Home Position, and detailed observation of Itokawa had started, using ONC-T/AMICA, NIRS, XRS, and LIDAR. Fig.5 is Itokawa images of different rotational phases taken by ONC-T. In September and October, about 1,500 images were obtained by ONC-T and ONC-W1. Fig.6 shows the near infra-red spectrum taken by NIRS.

Fig.7 Wire-frame shape model of Itokawa Three dimensional shape model of Itokawa was constructed on gourd from images and LIDAR data. (Fig. 7)

4. Landing and sampling In the end of October, Landing Site was selected at the Joint Science Team meeting, considering scientific interest and spacecraft safety. From the observations from Home Position, it was found that almost all surface of Itokawa were rocky or steep area, and “Muses-sea” was the only suitable landing point.

SS-23

2006 年　電子情報通信学会総合大会

BK-2-1

Fig.10 TM and Itokawa (2005/11/9)

Fig. 8 Artist view of Hayabusa landing The sampling method of Hayabusa is so-called touch and go way, that is, the spacecraft shoots a small bullet to the surface just after touch-down has detected, collects ejected fragments with sampler horn, and lifts off before one of solar cell panels might hit the surface. (Fig.8) To cancel the horizontal velocity, an artificial landmark named “Target Marker” (TM) is used. TM is a ball of 10cm diameter shown in Fig.9. Hayabusa equipped three TMs and one of them covered with shin film on which 880,000 names were written. TM is released at the altitude of about 30m. After TM is successfully captured and tracked with ONC-W1, the attitude of the spacecraft is aligned to the local horizon determined from four beams of Laser Range Finder (LRF) measurements. Then the spacecraft starts free-fall and touches down the asteroid surface. During the free-fall, some potential obstacles are checked with Fan Beam Sensors (FBS). If any obstacle is detected, sampling sequence is terminated and emergency assent is initiated.

Hayabusa tried the first rehearsal descent on November 4th. In this descent, the spacecraft was navigated and guided almost autonomously. But the effect of shadows on the rocky surface on images and orbital disturbance due to thruster firings for attitude control, the spacecraft did not descend as expected, and the descent was terminated by commands from the ground station at the altitude of 700m. The next practice descent was carried out on November 9th, for the test of LRF and TM tracking. And also, ground-based manual landmark tracking was introduced to assist onboard autonomous navigation. The practice was very fruitful and the performances of navigation sensors were validated. Fig.10 shows the image of released TM. ONC-W1 could successfully track it. During the descent, some close-up pictures were taken by ONC-T and Muses-sea area was found feasible for landing, while Woomera-desert which was the second candidate seemed inappropriate. (Fig.11 and 12)

Originally, one rehearsal descent and two actual landings for sampling were planed. Considering very asymmetric shape and rough terrain surface of Itokawa, however, total of three rehearsal or practice descents were carried out prior to actual landings.

Fig.11 Close-up image of Woomera-desert Fig.9 Target Marker

SS-24

2006 年　電子情報通信学会総合大会

BK-2-1

Fig.13 MINERVA Fig.12 Close-up image of Muses-sea On November 12th, the third descent was performed to deploy a small (about 500g) robot “MINERVA” which was expected to land, hop, and investigate on the surface. (Fig.13) Unfortunately, MINERVA was released at 200m altitude with ascent velocity and could not reach to Itokawa surface. However, it took the photo of Hayabusa’s solar cell panel shown in Fig.14. It was taken, selected, and transferred to Hayabusa autonomously. It demonstrated the miniaturization technologies for planetary robot.

The first landing for sampling was tried on November 20th. The guidance and navigation were all performed in order as planned. The guidance accuracy was within 30 meters in terms of the hovering point. Fig.15 shows the trajectory in semi-inertial coordinates. TM with 880,000 names was released at 40m altitude, and ONC-W1 tracked TM properly. Fig.16 is a lowaltitude image, in which the shadow of the spacecraft on the surface and the shinning released TM can be seen. The first touchingdown was unfortunately terminated by the obstacle detection of FBS, which has fan-shaped detection area beneath the solar cell panels shown in Fig.17.

Fig.14 Hayabusa’s solar cell panel taken by MINERVA

Fig.15 Descent trajectory on 20th

Fig.16 Navigation images (Left: taken at 30m altitude, Right: taken at 200m altitude) on 20th

SS-25

2006 年　電子情報通信学会総合大会

BK-2-1

deformation had detected and sampling sequence had completed. The guidance to the aimed landing point was perfect. In Fig.20, TM released on 20th can be seen in the same position. Fig.21 shows LRF data. Terrain alignment was successfully performed at about 7m altitude. Touching-down speed was estimated about 10 cm/second. When Hayabusa lifted-off, the +Z axis (High Gain Antenna axis) was 7 degrees off from the Sun direction as expected. Every instrument aboard functioned normally. Though it seemed perfect landing and sampling, after the spacecraft returned to Home Position, it lost attitude and can not communicate with highspeed link to the ground stations. Therefore detailed data can not be obtained so far.

1200mm

1160mm

Detection area

A

Cross section at line A 4400mm 1100mm Solar cell panel 250mm FBS-T 450mm FBS-R

Fig.17 Detection area of FBS

Fig.18 LRF data during bouncing and landing (The beam is canted with 30 deg from vertical) After the obstacle had detected, the spacecraft continued descending because attitude error was also large enough to prevent ascending thruster firing. As a result, the spacecraft did unexpected touch-down without sampling sequence, and stayed on the surface for about 30 minutes until forced ascent was commanded from the ground. Fig.18 shows the altitude measured by LRF. The second and final landing was performed on November 26th. The descent path taken was almost same as that at the 1st touching-down attempt, toward the west part of the Muses-sea. As already one TM was in the Muses-sea, to avoid the confusion, a new marker was not released, this time. That is, TM was not used for horizontal speed canceling, because we had confidence to control the spacecraft remotely from ground stations. And also the obstacle detection was set not to be referred to, since it seemed reporting too-sensitive signal on 20th. The touch-down sequence was set so that the lift off must be only after the sampler horn

SS-26

Fig.19 Descent trajectory on 26th

Fig.20 Navigation images on 26th

2006 年　電子情報通信学会総合大会

BK-2-1

.

Fig.18 LRF data before touch-down

5. Present status of Hayabusa On November 26th, there happened a leak from one of RCS thrusters (Bi-propellant chemical engines). Closing the latching valve for the thruster, acceleration due to the leak decreased. However, some troubles were followed after that, and the spacecraft finally lost its attitude. Communication link to the ground stations became impossible since December 8th. Present status of the spacecraft is estimated to be in a large coning motion due to the vapor gas that derived from the fuel leak-out and its command receiving antenna does not point to Earth. Though the analysis of the spacecraft dynamics says that the chance of recovery is about 70%, it seemed difficult to start ion-engine operation within 2005. Therefore, the project had determined that the return cruise should start from 2007 so that the spacecraft can return to the Earth in June of 2010, three years later than the original plan. As of the end of 2005, Hayabusa is still in the rescue operation. The project team is trying to send recovery commands with uplink frequency sweep, considering the temperature change of a command receiver.

6. Conclusions This report has presented the mission and results of Hayabusa. Though it is still uncertain whether the sample return of the surface material is possible or not, Hayabusa already provided a lot of engineering and scientific results.

SS-27