Accelerometry measurements using the Rosetta Lander's anchoring harpoon: experimental set-up, data reduction and signal analysis

In the years 2011-2013 the ESA mission Rosetta will explore the short perio d comet 46P/Wirtanen. The aims of the mission include investigation of the physical and chemical properties of the cometary nucleus and also the evolu tionary processes of comets. It is planned to land a small probe on the sur face of the comet, carrying a multitude of sensors devoted to in situ inves tigation of the material at the landing site. On touchdown at the nucleus, an anchoring harpoon will be fired into the surface to avoid a rebound of t he lander and to supply a reaction force against mechanical operations such as sample drilling or instrument platform motion. The anchor should also p revent an ejection of the lander due to gas drag from sublimating volatiles when the comet becomes more active closer to the Sun. In this paper, we re port on the development of one of the sensors of the MUPUS instrument aboar d the Rosetta Lander, the MUPUS ANC-Mi (mechanical properties) sensor. Its purpose is to measure the deceleration of the anchor harpoon during penetra tion into the cometary soil. First the test facilities at the Max-Planck-In stitute for Extraterrestrial Physics in Garching, Germany, are briefly desc ribed. Subsequently, we analyse several accelerometer signals obtained from test shots into various target materials. A procedure for signal reduction is described and possible errors that may be superimposed on the true acce leration or deceleration of the anchor are discussed in depth, with emphasi s on the occurrence of zero line offsets in the signals. Finally, the influ ence of high-frequency resonant oscillations of the anchor body on the sign als is discussed and difficulties faced when trying to derive grain sizes o f granular target materials are considered. It is concluded that with the s ampling rates used in this and several other space experiments currently un der way or under development a reasonable resolution of strength distributi on in soil layers can be achieved, but conclusions concerning grain size di stribution would probably demand much higher sampling rates. (C) 2001 Elsev ier Science Ltd. All rights reserved.