Session: SYMP 6-4: Smart Materials and Actuators

Paper Number: 139481

139481 - Design and Demonstration of a 3d Soft-Robotics Module Based on Rolled Dielectric Elastomer Actuators (Rdeas) 

With increasing numbers of robots working cooperatively with humans and navigating unstructured environments, the concept of soft robots has emerged as an alternative approach to traditional rigid robots. By either consisting partially (segmented soft robot) or fully (continuous soft robot) of compliant materials, high adaptability to its surroundings and better safety characteristics can be achieved. Common actuation principles of soft robots include pneumatics, hydraulics, shape memory alloy wires (SMAs), and, as in this case, dielectric elastomer actuators (DEAs).

In this work, we present the design and a demonstrator of a bending module for a segmented soft robotics arm, actuated by rolled dielectric elastomer actuators (RDEAs).

DEAs allow the unique combination of high-energy density and high bandwidth. Additionally, they are capable of self-sensing and composed of cheap, readily available materials with inherent compliance. All those benefits are possible without the need for an external pump.

Our rolled dielectric elastomer actuators (RDEAs) consist of two sheets of 50 µm-thick silicone film, coated with a carbon-black-based electrodes. The two sheets are stacked and wound in a spiral-like fashion, to create a flexible roll. This roll is terminated by a combination of commercially available wire-end ferrules and conductive silicone glue, as presented in our previous work. Our RDEAs were specifically developed for the use in soft robotics, with an elongated form-factor, robustness and the availability of a simple, analytical, physics-based model in mind.

The segment itself is composed of two circular platforms, made from PCBs. On each platform, a central rigid rod is attached. The endpoints of those rods are interconnected by a low-friction ball joint. The two platforms are connected by nine RDEAs, arranged around their circumference. By strategically applying high voltages to three groups of RDEAs, the segment bends in a specified direction in 3D space. Using an energy-based model, the geometric parameters of the system can be tuned: either to maximize the bending angle up to 30°, or to maximize the range of the bending angle between 0° and 20°.

To illustrate the characteristics of the system, a demonstrator will be presented. It allows playing back prerecorded trajectories of the segment, as well as interactive input via the control sticks of a game controller. The demonstrator consists of the segment itself, three compact high-voltage amplifiers, and a microcontroller that is connected to a game controller via Bluetooth.

Eventually, multiple of those systems should be stacked to create a segmented soft-robotics arm with a high number of degrees of freedom. Future research also includes the adaptation of existing self-sensing and control algorithms to our system.

Presenting Author: Vignesh Venkatachalam Saarland University

Presenting Author Biography: Mr. Venkatachalam is pursuing his PhD in the field of hybrid smart materials based actuators from the Saarland University. He completed his Masters in the field of Mechatronics from the Hamburg University of Technology where he worked on developing a constant curvature model for shape memory alloy wire actuated continuum robots. He has also worked on multiple soft robotic actuator projects during his Masters and Bachelors.

Authors:

Julian Kunze
Giovanni Soleti
Daniel Bruch
Gianluca Rizzello
Paul Motzki
Vignesh Venkatachalam