Distributed Control of Long Slender Robotic Manipulators
- Tuesday, 01 February 2011
Shape memory alloy wires modulate the end-effector contact load for sample capture applications.
Sample acquisition from small bodies is a key capability needed for proposed comet and asteroid sample return missions. This investigation determines how to utilize shape memory materials (SMMs) in sampling booms that result in improved sampling capability. By training an SMM wire to remember a given straight or curved shape when heated or cooled down to a given temperature, a long wire with low inherent bending stiffness may be ejected or unreeled from a spacecraft and then transformed into a long thin beam via a controlled material phase transition. Once the phase transition has been induced, the wire exhibits a bending stiffness that did not exist before, and the deployed appendage operates now as a stiff robotic arm.
Application." class="caption" align="left">The figure shows a conceptual view of how artificial manipulator tensioning would occur via an embedded SMM wire, and a proposed sequence for sample capture phase. Conversely, when the phase transition reverses, the original state of string behavior dominated by axial tension is recovered and the appendage can be reeled back inside the spacecraft.
The current state-of-the-art approach utilizes either an articulated arm whose limited length would require the spacecraft to approach the surface within about 3 meters, which would pose a significant risk to the spacecraft, or a Storable Tubular Extendable Member (STEM) boom, which would allow for sampling at slightly larger distances than 3 meters but would not allow for control of the interaction forces between the end-effector and surface, thus limiting the sampling to a few seconds due to the risk of buckling.
Utilizing SMM technology in sampling booms could enable innovative boom development that would allow for controlled small body sampling from a longer boom with lower risk to the spacecraft, and for longer sampling durations than possible with existing articulated arms and booms.
The proposed solution is to use distributed control of the long structural member via an active material (shape memory alloy wires in this application) to modulate the contact load that the end-effector would experience. The contact load would need to be modulated about a reference value, chosen to be 30-40 N for 2 to 3 seconds in this application, so that the end effector could collect sufficient regolith mass. Specifically, the problem was considered of how to control a sampling system that exhibits a contact phase, and that represents an attitude-stabilized spacecraft connected via an adaptive boom to a tip mass.
The tip mass is assumed to contact the surface of the body for a specific duration of time during which the specific mechanics of the sampler device would enable the material collection and transmit a horizontal and vertical force to the spacecraft-sampler system. The actuation inputs are the vehicle thruster forces, the contact force at the surface, and the boom distributed control inputs, for example Joule heat in the shape memory wire embedded in the boom. The distributed control algorithm is such that a load cell measures the contact load. This load translates into an equivalent amount of compressive stress in the beam, hence into an equivalent stress in the shape memory wires supported by the beam.
When the contact has been made, a command is issued to raise the amount of Joule (electrical) heat to the wires, which then transition from the martensitic to the austenitic state, becoming stiffer. The stress in the beam is then increased, and its global shape changes in a distributed manner, so that now buckling is precluded. The curvature changes and the effect of the back-reaction to the spacecraft during contact is delayed, hence long arms (100’s of meters, even kilometers, hence the potential application to tethered spacecraft as well) could be used within this actuation scheme to drive an active end effector to a surface remotely.
The simulation studies demonstrate that contact force modulation through distributed control of the boom elasticity causes weak dynamic coupling between the spacecraft and the end-effector and has an inappreciable effect on the stability of the system. Also, with the new concept the stiffness of the end-effector arm could be actively modulated so that the back-reaction on the spacecraft could be greatly reduced. A variety of asset deployment or sample capture scenarios would therefore be possible that could potentially minimize the dynamic interactions with the spacecraft during the maneuver.
Intelligent materials such as shape memory alloys can potentially provide a mechanically simple and affordable solution for sample capture and return across a wide range of mission types, including sample acquisition and return from small bodies such as asteroids and comets, but also for missions including aerobots (blimps, balloons) sampling the surface of Titan, Venus, or Mars.
Phase-transition based manipulation can also potentially impact other areas such as rendezvous/docking and remote target capture from a distance, thus reducing potential impact hazards and costly final approach maneuvering. This specific application would result in a unique and new capability by enlarging the range of applicability of autonomous robotic systems to proposed missions involving sample collection and return, on account of the adaptivity and robustness of this manipulator to a wide range of end conditions during sampling. Other applications include the use of long thin beams or tethers, which are actively controlled in a distributed fashion to modulate the end load. Sample capture, collection, and delivery of assets at long distances from the spacecraft appear possible with this innovation.