Search for: All records

Dielectric elastomers (DEs) are electro-active polymers that deform and change their shape when an electric field is applied across them. They are used as soft actuators since they are flexible, resilient, lightweight, and durable. Many models have been proposed to describe and capture the behavior of these actuators such as circuit representation, lumped parameter modeling, and physics-based modeling. In this paper, a hybrid between the physics and lumped parameter model is presented which is used to control the actuator. The focus of this paper is on a tubular dielectric elastomer actuator (DEA). The model proposed is validated with experimental data to evaluate its approximation to the physical actuator. The physics model offers the ability to describe how the material properties and actuator's geometry affect the dynamics and behavior of the actuator under different states. The lumped parameter model accounts for physical quantities that may not be fully expressed when formulating the physics-based equations. The discussed model performance is found to have an error less than 10% for the sinusoidal signals discussed. 

Dielectric elastomer (DE) materials, a category of electroactive polymers, can be used to design actuators that are flexible, resilient, lightweight, and durable. However, due to the uncertainties in its actuation dynamics, DE actuators always rely on feedback control to perform accurate and safe operations. In this paper, a tubular dielectric elastomer actuator (DEA) with self-sensing capability is developed. It does not require external devices to measure displacement for feedback control. The displacement of the actuator is controlled using a proportional-integral controller with the capacitance measured at high probing frequency as the self-sensing mechanism component of the actuator. By superimposing actuation and probing voltage and applying them to the DE tube, the actuation voltage activates the movement of the DE tube and the probing voltage is used for self-sensing. Fast Fourier Transform (FFT) is then used to filter a given frequency of the probing current and voltage and then calculate the capacitance from the probing current and voltage during each time window. With the relationship between capacitance and displacement of the DE tube, the displacement output is estimated online and self-sensing without an external sensor is achieved. The self-sensing signal is then used as a feedback signal in a closed-loop design to follow a reference signal for tracking. The experimental results validate the self-sensing of the DE actuator in feedback control.