135 research outputs found
Modelling and control of double-cone dielectric elastomer actuator
No full textAmong various dielectric elastomer devices, cone actuators are of large interest for their multidegree-
of-freedom design. These objects combine the common advantages of dielectric
elastomers (i.e. solid-state actuation, self-sensing capability, high conversion efficiency, light
weight and low cost) with the possibility to actuate more than one degree of freedom in a single
device. The potential applications of this feature in robotics are huge, making cone actuators very
attractive. This work focuses on rotational degrees of freedom to complete existing literature and
improve the understanding of such aspect. Simple tools are presented for the performance
prediction of the device: finite element method simulations and interpolating relations have been
used to assess the actuator steady-state behaviour in terms of torque and rotation as a function of
geometric parameters. Results are interpolated by fit relations accounting for all the relevant
parameters. The obtained data are validated through comparison with experimental results:
steady-state torque and rotation are determined at a given high voltage actuation. In addition, the
transient response to step input has been measured and, as a result, the voltage-to-torque and the
voltage-to-rotation transfer functions are obtained. Experimental data are collected and used to validate the prediction capability of the transfer function in terms of time response to step input and frequency response. The developed static and dynamic models have been employed to implement a feedback compensator that controls the device motion; the simulated behaviour is compared to experimental data, resulting in a maximum prediction error of 7.5%
A Simulation Tool for Robotic Active Debris Removal with minimum reaction space manipulator
No full textSimulation of robotic space operations with minimum base reaction manipulator
No full textAutonomous robotic capture has been identified as a key technology for On-Orbit Servicing (OOS) and Active Debris Removal (ADR) missions. However, manoeuvring spacecraft-mounted manipulators is a challenging task since it generates disturbance torques on the satellite. To mitigate this problem several minimum reaction control strategies have been developed to reach the desired End-Effector (EE) pose while minimizing the dynamic disturbances transferred to the spacecraft by the robotic arm.
This paper presents the development of a Simulation Tool in the MATLAB/Simulink environment capable of simulating the dynamics of a satellite equipped with a 7-DoF robotic arm during the target capture phase. The manipulator, as well as the spacecraft, is implemented by using Simscape Multibody and joint actuators are modelled as Brushless DC motors controlled by PID controllers. The spacecraft attitude is Nadir-Pointing, it is controlled by means of quaternion feedback and Linear-Quadratic-Regulator (LQR) and it is actuated by three Reaction Wheels (RW). Orbital perturbations such as non-spherical gravity potential (EGM2008 model) and atmospheric drag are considered. In addition, the minimum reaction control strategy called Kinetic Energy Minimization (MKE) is employed during the robotic arm manoeuvres.
The goal of this study is to compare the performances obtained with the MKE method with those achieved by using the classic Inverse Kinematics (IK) in the free-flying case. The numerical results confirmed that MKE method is to be preferred since it minimizes the control torque that the Attitude Control Subsystem (ACS) must provide and reduces the EE orientation and position errors
Compliant joint to reduce docking loads between CubeSats
No full textThis paper presents the design, modelling and experimental verification of a damping joint for application to miniature space docking mechanisms. The joint design is based on deformable elastomeric elements, thus avoiding any sliding or contact between moving components. Numerical FEM simulations have been conducted in order to quantify the joint mechanical characteristics (rigidity and damping coefficient) as a function of the main design parameters (geometry, material, assembly). The obtained parametric relations provide an estimate of the joint characteristics based on the selected design. An equivalent visco-elastic model is developed and implemented in dynamic simulations. The results of the experimental evaluation of the joint design provide a validation of the developed models and prove the advantage of adopting damping joints in docking applications between small satellites, like reduced contact loads and enhanced damping
Innovative technologies for the actuation of space manipulators
No full textIn this work, innovative technologies for the actuation of space robotic systems are investigated as possible alternative to traditional motors. The research activity focused on double-cone Dielectric Elastomer Actuators (DEAs). The most notable results achieved are predictive models for the static and dynamic performances estimation of the mentioned devices and experimental validation of both single actuators and a robotic arm prototype.
The general objective of the thesis is to evaluate innovative actuation technologies for space robotics; the main expected output of the research is the feasibility proof of a robotic space system based on low-TRL (Technology Readiness Level) devices. This objective is achieved by fulfilling two secondary goals:
- development of models to predict the actuator performances and validation of ready-to-use design tools;
- experimental evaluation of a multi-body manipulator prototype in laboratory environment.
The motivation on which this work is based, comes from the wide interest on robotics that recently grew among the space community. A large variety of space missions can benefit from the implementation of automated systems reducing risks, costs, delays and errors deriving from human interaction (i.e. astronauts or ground operators) with space vehicles and structures. On-Orbit Servicing (OOS) missions, in particular, are based on robotic servicing vehicles that perform complex tasks on client objects enabling unprecedented scenarios of improved accessibility to space. Future effective and efficient exploitation of space is strongly dependent on the development of key technologies to support existing and planned orbital assets, aiming to extend spacecraft operational life and to boost mission flexibility. Investigation on innovative actuation technologies is critical to improve space robotics performances and enable new applications. The TRL advancement of young technologies is at the basis of the development of new systems.
To date, a considerable number of relevant applications of robotics have been operated in space; main tasks include assembly of complex structures, manipulation of client vehicles and support to astronauts activities. Five human operated manipulators have equipped the Space Shuttle or the International Space Station (ISS), along with a variety of other experimental demonstrators; three examples of humanoid robotic astronauts have been tested and reached different levels of development; a wide range of autonomous demonstrative OOS missions have been conceived and designed, are currently under development or, in some cases, have been flown with success; several planetary probes and (partially) autonomous rovers have been operated on the surface of extraterrestrial bodies like the Moon or Mars. These missions and others constitute the solid background on which this work is based and consolidate the motivation behind the research. The past and present trend in the space sector is to seek improved capabilities, flexibility and autonomy of vehicles, assigning a prominent role to robotics as a key enabling technology.
By far the most common actuators in space systems are conventional DC drives like stepper motors and brushless motors: the first are used in robotic arms for control simplicity and positioning accuracy, the second are the standard option in reaction wheels. In some cases brushed DC motors (in sealed or planetary environment) and, less often, voice coil motors have been used. Innovative technologies, like smart materials, are rarely adopted mainly due to reliability and heritage reasons. In general, the space community is very conservative and new technologies have to be proven fail safe and robust, and, for this reason, well-known solutions are often preferred. Nevertheless, implementation examples of smart technologies in space exist and they performed particularly well in off-nominal conditions, where traditional solutions show limitations. It is worth mentioning the most notable: piezo-electric actuators and motors, used in micro-positioning and precision pointing; shape memory devices, employed in release mechanisms; bimetallic actuators, implemented in single-shot systems and thermal control; Electro-Active Polymers (EAPs). The latter have not been extensively employed in space systems yet, although interest is growing around them on the basis of the appealing capabilities proved in many laboratory tests. A wide choice of alternative EAP materials and configurations have been proposed, with ample performance ranges. Dielectric Elastomer Actuators are a promising branch of EAPs family, whose space TRL is currently 2-3. Dielectric Elastomers are arguably the best performing EAPs and, for this reason, very appealing. DEAs have been selected to be investigated in this work for three main reasons:
- good compromise performances in terms of stroke/deformation, force/torque and time response;
- interesting characteristics like low mass and low power consumption, possibility to improve performances through design flexibility and modularity, multi-DoF configurations, simple manufacturing process, low costs, solid state actuation (no friction), self-sensing capability;
- highly innovative technology with low TRL.
Double-cone actuators are selected for their flexibility and multi-DoF architecture.
An example mission scenario is conceived and simulated in order to determine preliminary requirements for the robotic system and the single actuator. An Active Debris Removal (ADR) mission is selected as a key OOS application of robotic systems. In the considered scenario a large piece of debris (1400 kg) is captured by a small spacecraft by means of a multi-DoF manipulator. The debris is spinning with respect to the servicing spacecraft which is equipped with a robotic arm composed by a variable number of joints (1-3). The capture interface is rigid and guarantees the mechanical connection between the manipulator and the client object. Several simulations are performed with different initial conditions and capture strategies, including the options of a rigidly controlled or free flying spacecraft. The requirements have been defined in terms of forces/torques and rotations at the robot joints. The maximum angular deflection required to the entire robotic arm is 90 deg; torque and forces are strongly dependent on the initial debris (relative) angular momentum, thus it is possible to relax the joint requirement imposing stricter constraints to the target selection or relative navigation system of the servicer.
The double-cone DE actuator is based on two circular, pre-stretched membranes of elastomer coated with compliant electrodes on both sides. By applying high voltage to the electrodes, electrostatic forces squeeze the membrane reducing its thickness and, consequently, expanding the material in the plane. Such material deformation is exploited to displace the actuator shaft. Multiple DoF are obtained by selecting a proper electrode layout; a 2-DoF (one rotational and one translational) configuration is selected in view of the proposed robotic application. On the basis of the results available in literature, the commercial polyacrylic elastomer called 3M VHB 49XX is chosen. Proper electromechanical models are identified for the mentioned polymer.
Once a set of geometrical and manufacturing parameters are defined, numerical simulations based on literature as well as newly developed FEM models are performed in order to collect a large number of performance data. Interpolating relations are obtained from the collected data and allow to estimate the steady-state performances of the actuator. Torque/force and rotation/stroke are proportional to the squared value of applied high-voltage. The mentioned relations allow to compute the gain to which squared voltage has to be multiplied to estimate the desired quantity. The mean error on estimations is 6.1% for angular rotation, 10.6% for torque, 22.5% for linear stroke and 11.8% for force.
A different approach is adopted to model the dynamic behavior of DEAs: transfer function (TF) based models are developed from time dependent data collected through long term tests. The elastomeric material adopted in the device manufacturing shows a relevant viscoelastic behavior that considerably affects the time response of actuators. The TF approach is chosen to simplify the estimation of the transient behavior of DEAs and to provide a practical design tool for robotic applications. The prediction capabilities of TF models are evaluated by comparison with experimental step response. The mean error on the 70% rise time is 15% for angular rotation, 9.5% for torque, 14% for linear stroke and 14% for force; the mean error on amplitude for t > t_r is 4% for angular rotation, 4% for torque, 9% for linear stroke and 11% for force.
The developed models, both static and dynamic, are suitable for the implementation of control algorithms and, consequently, for robotic applications.
The capability to control the actuator is experimentally proven by testing Single Input / Single Output compensators to actuate both DoF independently. Laboratory tests are conducted in order to evaluate the step response of double-cone actuators. Good accordance is obtained between the simulated and the experimentally measured time response with errors compatible with the prediction inaccuracies of the mentioned models.
Finally, a multi-body application of double-cone actuators is designed, manufactured and tested along with a proper control algorithm. The robotic arm is composed by two double-cone DEAs mounted in series. Each actuator has two DoFs and the manipulator moves in the horizontal plane. Two degrees of kinematic redundancy are achieved in the manipulator by controlling only the in-plane position of the end-effector. The arm prototype is suspended by an inextensible cable that reduces the effects of gravity on the motion. The experimental task is the tracking of simple linear and arc trajectories. A vision system monitors the position of the end-effector (optical marker) and feeds the position information to a control computer that commands the voltage actuation to the joints through a properly designed control algorithm. The kinematic redundancy is exploited by the controller to optimize the end-effector trajectory to achieve a given objective: several control schemes with alternative optimization functions are designed and simulated numerically in order to select the best performing option. The chosen control algorithm aims at the minimization of joint variables in order to reduce the risk of actuators saturation. The system performs well and the maximum position error norm is 6.4% of total path length for linear trajectory and 6.8% for arc trajectory
Design of an innovative Dielectric Elastomer actuator for space applications
No full textThe capability of Dielectric Elastomers to show large deformations under high voltage loads has been deeply
investigated to develop a number of actuators concepts. From a space systems point of view, the advantages
introduced by this class of smart materials are considerable and include high conversion eciency, distributed
actuation, self-sensing capability, light weight and low cost. This paper focuses on the design of a solid-state
actuator capable of high positioning resolution. The use of Electroactive Polymers makes this device interesting
for space mechanisms applications, such as antenna and sensor pointing, solar array orientation, attitude control,
adaptive structures and robotic manipulators. In particular, such actuation suers neither wear, nor fatigue
issues and shows highly damped vibrations, thus requiring no maintenance and transferring low disturbance
to the surrounding structures. The main weakness of this actuator is the relatively low force/torque values
available. The proposed geometry allows two rotational degrees of freedom, and simulations are performed to
measure the expected instant angular de
ection at zero load and the stall torque of the actuator under a given
high voltage load. Several geometric parameters are varied and their in
uence on the device behaviour is studied.
Simplied relations are extrapolated from the numerical results and represent useful predicting tools for design
purposes. Beside the expected static performances, the dynamic behaviour of the device is also assessed and the
input/output transfer function is estimated. Finally, a prototype design for laboratory tests is presented; the
experimental activity aims to validate the preliminary results obtained by numerical analysis
A miniature stabilized platform for lasercom terminals on-board nanosatellites
No full textIn the last decade miniature satellites have become attractive due to their inherent advantages: the reduced mass, production cost and time, as well as the low launch cost allow small companies, corporations and universities to access space easily. Moreover, constellations based on miniature satellites for observation, mapping or telecommunication purposes could represent an alternative to systems based on larger platforms, thanks to the further cost reduction due to mass production. However, pico- and nano- satellites still present severe technical limitations which prevent their exploitation for complex or high-performance missions. In particular, the reduced available power and volume restrict up/downlink data rates to a few hundred kbit/s.
To this day, optical links represent the unique viable solution to increase dramatically the communication capabilities of nanosatellites. In fact, only lasercom technology, thanks to the very narrow beam emission, permits to achieve data rates up to some Gbit/s with devices which can fit on a nanosatellite host bus in terms of volume, mass and power consumption. RF systems with comparable performance would inevitably exceed the resources available on such miniature platforms.
However, the extremely stringent pointing accuracy and stability required by optical link terminals are not compatible with the actual and perspective attitude control performance of nanosatellites.
In order to overcome these technical limitations, the authors are developing a miniature actively stabilized platform capable of rejecting the residual bus vibration and provide a small optical link device with a vibration-free base. Its exploitation will allow to relax the requirements on both the attitude control system of the host spacecraft and the pointing system of the optical communication terminal, making the miniaturization of the latter easier. The device consists of a parallel, three rotational degrees of freedom platform controlled by means of three identical actuators based on piezoelectric elements. The active control is required to manage low-frequency vibrations, while high frequency disturbances are rejected by high-stiffness elastic elements. The parallel configuration is chosen for its simplicity, symmetry and overall stiffness.
In this paper, the design of the actively stabilized platform is presented, along with the numerical simulations performed in order to evaluate the system performance
Ground Facility for Validation of Proximity Operations: a Hardware–In–the–Loop Experiment
No full text- …
