1. Introduction Even now, the Moon is a one of interesting celestial bodies for exploration or investigation in space development. The Moon has a lot of unclear questions as the origin, the chemical components, and the internal structure. Thus, according to these remarks, it is required to obtain the global information about the Moon for future lunar exploration missions. Therefore the authors propose a robotic system for burying a long-period seismometer under the lunar surface. In general, the lunar regolith has high adiabatic property, and the temperature is constantly -20 degree even if at night. This consequently enables the seismometer to measure for a long term. Also, the buried seismometer can realize better contact with surrounding regolith. This paper firstly describes the strategies of subsurface propulsion of the burrowing robotic system and required mechanisms. Then, this paper especially discusses an excavation mechanism which is very important for the system, and proposes a novel excavating mechanism. Fig. 1 Schematic of lunar subsurface exploration Table 1 Planetary excavation systems Mechanism Size Weight Robustness Purpose Bucket Wheel simple large heavy ~ sampling Penetrator simple middle long middle ~ investigation Drill simple narrow long light sampling Burrowing Robot complex compact light investigation Fig. 2 Overview of planetary excavation robots 2. Burrowing Robot System 2.1. Conventional Robots Planetary exploration with excavation of soil has received a lot of attention in the world. The comparison on planetary excavation systems is shown in Table 1, where the robustness means a capability to have some excavating points. So far, there have been proposed burrowing robots for planetary subsurface exploration [1]-[5]. However, those robots do not meet the scientific requirements. The overview of some proposed ideas which are classified in the length and the diameter is shown in Fig.2. ICEBREAKER and the Lunar-A penetrators of JAXA are not actually burrowing robot, but these are robotic systems for subsurface investigation or the similar objective. Therefore these two different methods are shown in Fig.2 for a comparison. 2.2. Strategies for subsurface propulsion Based on the past researches, the authors set the following assumptions: EThe burrowing robot is carried by the rover EPower is supplied by a cable from the ground station ETarget depth is several meters from the surface ETarget soil-layer is lunar regolith EThe robot diameter is about 0.1 m Next, the authors define the strategies for subsurface propulsion by the burrowing robot, and consider the following two phases. Several concrete methods are shown in Fig.3. (1) Make space ECompress fore-regolith ERemove and back transport fore-regolith (2) Advance forward EUtilize a contact with surrounding regolith EUtilize an excavated regolith ESelf generation unrelated to regolith Fig. 3 Strategies of subsurface propulsion 3. Propulsion Limit Depth Before considering the active generation of propulsion force as shown in Fig.3, this paper discusses the propulsion limit depth by only its weight. The main resistances for propulsion are lateral friction and excavating resistance by applying the simplified dynamics model as shown in Fig.4. For subsurface propulsion, the lateral friction is universal resistance. Firstly the soil pressure underground is calculated by the following equation. ( p gh)K 2c K 0 = + } i1j where, the notations used in this paper are as followings. : lateral soil pressure [kPa], p0 : external stress on the surface [kPa], : soil bulk density [kg/m3], g : gravity acceleration [m/s2], h : depth from surface [m] c : cohesion of soil [kPa], K : Rankinefs soil pressure coefficient, H : robot length [m], D : robot diameter [m], A : robot cross-section [m2] ( := (D/2) 2 ), : friction coefficient (soil ? robot ), L F : soil friction resistance [N], M : robot mass [kg], Also, the robot shape assumes to be cylindrical. When p0 and c are zero, and the total lateral friction resistance L F is calculated by the integrating through total lateral surface as following. ( ) 2 2 2 A K g hH H F A K gh dh h h H L ? = = ? ? @@ i2j Thus the propulsion limit depth is estimated for the burrowing robot which mass is M from the following equation (3). F ? Mg = 0 i3j As the above consideration, the propulsion limit depth respect to the robot mass is shown in Fig.5 where is 1600 [kg/m3], D is 0.1 [m] and H is 0.2 [m]. Thus, given the strict weight limits of space mission, the burrowing robot needs to generate the propulsive forces subsurface. Fig. 4 Simplified dynamics model of soil friction Fig. 5 Advancing limit depth by its weight 4. Compression Limit This paper considers the capability of compression fore-soil method as an excavation mechanism. The relationship at an extreme state between vertical compressing stress and shear stress by applying Mohr-Coulombfs failure criterion theory is shown in following equations (4) and (5). Ӂ@ sin 2 2 1 3 1 3 ? ? + = i4j = + c ? = cos tan 2 3 1 i5j Here the following new notations are defined. : shear stress [kPa], 0 : initial stress [kPa], 1 : maximum soil pressure [kPa], 3 : minimum soil pressure [kPa], : internal friction angle [rad], e : void ratio, e0 : initial void ratio, CC : compression index, B : generated space distance [m], x : effect compressing range [m], The compression index C C is introduced, which denotes the relationship between void ratio and compressing pressure as described in the following equation (6). 10 10 log d log e de CC = ?   = ? i6j Therefore, by integrating the above equation (6), the following equation is obtained. @0 0 10 e C log e C + ? ? ? ? ? ?? ? = ? i7j Then, the effect range is assumed by the following equation (8). This means that compression force affects to distance X from the bodyfs center line as the same force, and this can indicate the best condition about the compression to make space. ? ?? ? ? ?? ? + ? + = ? e e e e B xD 1 1 0 0 i8j The simulation result is shown in Fig.7. In the simulation, the parameters are 1 ? x ? 10 and e0=1.2, and D is 0.1 [m]. The simulation result shows that it can be too hard to make space. Thus it is concluded that because the lunar soil layer has high-filling state under the surface, a method to compress fore-soil is an unrealistic to make space and the robot has to remove and transport fore-soil to backward. Fig. 6 State model of compression method Fig. 7 Compression limit by Mohr-Coulombfs failure criterion with CC 5. Non-Reaction Mechanism 5.1 Proposal of Double Rotation System From the above considerations, the required properties for robotic excavation mechanism are following: EFore- soil removal and transportation backward EContribution of propulsion force EDust prevention mechanism Therefore, a screw drill is one of candidates, which has a series of spiral wing. However, there is a problem that a single spin drill has reaction to the body. This reaction can be friction as reducing the efficiency as well as leading to a wobbling of propulsion axis. So this paper proposes a new type spiral screw drill unit as a kind of novel excavation mechanisms to improve the problem. The double rotation screw is developed with N-RDM (Non-Reaction Drilling Mechanism). The double rotation methods are classified into three types as shown in Fig.8. The contra-rotor type (a) has one drilling unit which has the contra-rotation axis at the same of another rotation axis. The twin-rotor type (b) consists of two drilling units which have regular and contra rotation axis each other at binary positions. The dual-spin type (c) has one drilling unit which has dual different spin axis. Each type has two rotation axes, so it is possible to cancel the reaction. On the other hand, firstly the contra-rotor type is compact size, and it can be estimated that it has an equivalent excavating performance with the single screw drilling. Secondly the twin-rotor type has unexcavating space compared with the contra-rotor type because the shape of an excavated hole is anomalous. It is impossible to make one circular hole, and there is constraint for the body shape. Thus it can be estimated that it can be lower efficiency. Thirdly the dual-spin type has some external driving parts, and the efficiency is getting lower. According to these considerations, the authors firstly adopt contra-rotor type and will do some experiments in the near future. 5.2 SSD The schematic of the basic model, where it is call SSD (Single Screw Drilling Unit) unit, is shown in Fig.9(a), and also the prototype model is shown in Fig.9(b). The developed SSD is a conceptual model of the contra-rotor type screw. And it is a basic dynamics model for the theoretical analysis. The SSD unit consists of a body-part and an excavation- part. Furthermore, the excavation-part has an inner cone, called CONE, and a helical screw wing which wind around the CONE, called SCREW. From Fig.9(a), a parameter denotes the volubile angle from that central axis along the SCREW upper surface through from the top to end of the SSD. Then SC [deg] denotes the angle of inclination of the SCREWfs center position and p denotes the SCREW pitch, where these parameters are a function which is defined by . Fig.8 Schematic of N-RDM systems (a) Drawing (b) Picture Fig. 9 Model of SSD unit 6. Conclusion In this paper, the authors discuss a novel robotic burying seismometer system. And firstly this paper describes the limit of compression method for making space in subsurface and need of generation of a propulsive force for advancement. Then the authors propose a new excavating screw mechanism with N-RDM. Some experiments are under going for the validation of the proposed excavating method as future works. Reference [1] K. Watanabe, S. Shimoda, T. Kubota and I. Nakatani, hA Mole-Type Drilling Robot for Lunar Sub-surface Explorationh, Proc. of the 7th iSAIRAS, AS-7, 2003. [2] K. Yoshida, N. Mizuno, T. Yokoyama, Y. Kanamori, M. Sonoyama, T. Watabe, hDevelopment of Mole-type Robot for Lunar/Planetary Sub-Surface Exploration, and its Performance Evaluationh, Proc. of the 20th Annual Conf. of the RSJ (in Japanese), 1J35, 2002. [3] Y. Liu, B. Weinberg and C. Mavroidis, gDesign and Modeling of the NU Smart Space Drilling System (SSDS)h, Proc. of the 10th ASCE Aerospace Division International Conf. on Engineering, Construction and Operations in Challenging Environments (Earth and Space 2006), 2006. [4] R. Campaci, S. Debei, R. Finotello, G. Farziano, C. Bettanini, M. Giacometti, G. Rossi, G. Visentin and M. Zaccariotto, hDesign and Optimization of A Terrestrial Guided Mole for Deep Subsoil Exploration- Boring Performance Experimental Analysish, Proc. of the 8th iSAIRAS, 2005. [5] S. P. Gorevan, T. M. Myrick, C. Batting, S. Mukhrrjee, P. Bartlet and J. Wilson, hStrategies for Future Mars Exploration: An Infrastructure for The Near and Longer-Term Future Exploration of The Subsurface of Marsh, Proc. of the 6th Int. Conf. on Mars, 2003.