[This is a voluntary translation of Taro's Page February 5, 2009 on Hayabusa's Returning Home. ]
Hayabusa is returning home. Her ion engines have been successfully reignited. Hooray! The press release states it very briefly:
Since then, the HAYABUSA has been in a coast flight, but today [February 4] we reignited the ion engine for a powered flight at 11:35 a.m. (JST) after activating its reaction wheel and establishing the three axis attitude control.
Simpler it seems, but the reality is, two out of three reaction wheels are not functioning. Chemical thrusters are also not usable due to fuel leaks. It really is beyond anyone's imagination that the team has come to reestablished the three-axis control. So here's the unsung story:
Think of the orientation of Hayabusa as a human body. The antenna is at the top of the head, solar panels are on both sides like extended arms. Ion engines are on the back like a backpack. She has been in spinning mode to keep the attitude for the past sixteen months, without firing any engines nor thrusters. The Sun has always been kept to the spinning axis, which means the antenna and the solar panels are always facing to it.
So, to initiate the returning home procedure, the team should first stop the spin. The only remaining reaction wheel is, luckily, the yaw control, the same axis of the spin motion. That is, rotating the wheel would stop the spin. To achieve the full three axis control, two more axis to go. The ion engines on the back can sway five degrees each to up, down, right and left (gimbal-mounted). Originally designed for fine-control the direction of ion engine thrust, this gimbal structure can also be exploited to control the bowing down motion along the pitch axis.
One more axis remains, namely the roll axis, or, turning like cartwheel sidewise. Swaying the ion engines left or right gives the yawing effect, same as the remaining reaction wheel, so it's no use. Instead the team has decide to exploit the photon pressure coming from the Sun! The mass of Hayabusa is concentrated at the center of the main structure, and the solar panels are extending to both sides at the top of the main structure. So a small amount of torque arises when the panels receives the photons at a tilted angle. Just like a solar sailing.
Amazing, isn't it? The team has come up with this "Plan B" to regain the three axis control technique, almost like cheating an exam with an ultra-E moonsault! And that was in April 2007. They confirmed this technique to be usable, overwrote the control program with this alternative, put Hayabusa into the spinning mode, then hibernated her until now. A year and four months later, they reactivated her with already practiced maneuver, controlling her attitude, and firing the ion engines to start the still long process of coming home. I've never heard of using photon pressure for the attitude control. And yet they never hint us about their efforts. Quite cool, aren't they?
[Addendum by 5thstar]
This article has been linked by The Planetary Society Blog, NASAspaceflight.com and ForumAstronautico.it (English Translation) as of this addendum. Thank you Emily, for bringing this to the world's attention! Hayabusa team truly deserves it.
In response to Spaceater's question:
Ho letto l'articolo e non mi è molto chiaro come abbiano fatto a sfruttare il "solar sailing" per controllare il rollio dell'astronave: ruotavano i pannelli solari in modo che ricevessero più o meno fotoni in modo che la differenza di pressione dovuta alle particelle creasse un momento intorno all'asse di rollio? Perchè dalle immagini della sonda che ho visto i pannelli non sembravano mobili...
I read this article I am not very clear how they did it to exploit the "solar sailing" to control the roll dell'astronave: revolved solar panels to receive more or less photons so that the pressure difference due to particles to create a moment about the roll? Because of the probe from the images that I saw the panels did not appear to be moving ...
Good question! I am not clear either. So I googled a bit, and find out a Japanese article written by one of the Hayabusa team member Professor Hitoshi Kuninaka on ISAS Mailing List, essentially explaining the same thing back in August 2007.
He states that the photon pressure to a spacecraft generates a disturbance to the attitude control, depending on the angle of the Sun light and the cross section of the spacecraft. It generates a small torque as the center of the photon pressure versus the center of mass is slightly shifted.
So my guess is that they are carefully exploiting the combination of yaw, pitch and roll controll of Hayabusa at the same time by calculating the exact positions and vectors of the Sun light versus the center of the photon pressure versus the center of the mass. It ought to be a state of art software. I could be wrong.
Then I googled a bit more, and found his paper.
Deep space flight of Hayabusa asteroid explorer
Proc. SPIE, Vol. 6960, 696002 (2008); DOI:10.1117/12.782163
Online Publication Date: 15 April 2008
Since I don't have an access to this database, could someone who has an access download and check it?