Soft robotics are attractive to researchers and developers due to their potential for biomimicry applications across a myriad of fields, including biomedicine (e.g., surgery), the film industry (e.g., animatronics), ecology (e.g., physical ‘animats’), human–robot interactions (e.g., social robots), and others. In contrast to their rigid counterparts, soft robotics offer obvious actuation benefits, including their many degrees of freedom in motion and their potential to mimic living organisms. Many material systems have been proposed and used for soft robotic applications, involving soft actuators, sensors, and generators. This review focuses on dielectric elastomer (DE)-based actuators, which are more general electro-active polymer (EAP) smart materials. EAP-based soft robots are very attractive for various reasons: (a) energy can be efficiently (and readily) stored in electrical form; (b) both power and information can be transferred rapidly via electrical phenomena; (c) computations using electronic means are readily available. Due to their potential and benefits, DE-based actuators are attractive to researchers and developers from multiple fields. This review aims to (1) provide non-experts with an “easy-to-follow” survey of the most important aspects and challenges to consider when implementing DE-based soft actuators, and (2) emphasize current solutions and challenges related to the materials, controls, and portability of DE-based soft-actuator systems. First, we start with some fundamental functions, applications, and configurations; then, we review the material models and their selection. After, we outline material limitations and challenges along with some thermo-mechano-chemical treatments to overcome some of those limitations. Finally, we outline some of the control schemes, including modern techniques, and suggest using rewritable hardware for faster and more adaptive controls.

The Environmental Protection Agency (EPA) of the United States recently released the first-ever federal regulation on per- and polyfluoroalkyl substances (PFASs) for drinking water. While this represents an important landmark, it also brings about compliance challenges to the stakeholders in the drinking water industry as well as concerns to the general public. In this work, we address some of the most important challenges associated with measuring low concentrations of PFASs in drinking water in the field in real drinking water matrices. First, we review the “continuous monitoring for compliance” process laid out by the EPA and some of the associated hurdles. The process requires measuring, with some frequency, low concentrations (e.g., below 2 ppt or 2 ng/L) of targeted PFASs, in the presence of many other co-contaminants and in various conditions. Currently, this task can only (and it is expected to) be accomplished using specific protocols that rely on expensive, specialized, and laboratory-scale instrumentation, which adds time and increases cost. To potentially reduce the burden, portable, high-fidelity, low-cost, real-time PFAS sensors are desirable; however, the path to commercialization of some of the most promising technologies is confronted with many challenges, as well, and they are still at infant stages. Here, we provide insights related to those challenges based on results from ab initio and machine learning studies. These challenges are mainly due to the large amount and diversity of PFAS molecules and their multifunctional behaviors that depend strongly on the conditions of the media. The impetus of this work is to present relevant and timely insights to researchers and developers to accelerate the development of suitable PFAS monitoring systems. In addition, this work attempts to provide water system stakeholders, technicians, and even regulators guidelines to improve their strategies, which could ultimately translate in better services to the public.

Dielectric elastomers (DEs) exhibit remarkable properties that make them stand out among other electroactive polymers. Various types of actuators based on DEs have been used in applications that include artificial muscles, Braille displays, and robotic joints. In particular, conical dielectric elastomer actuators (CDEAs) are very attractive due to their multiple degrees of freedom (DOF) and easiness of construction. In this study, an energy method is used to derive an improved mathematical model for a double-cone dielectric elastomer actuator (DCDEA) capable of predicting horizontal and rotational displacements. To create the model, a new variable is introduced into the equations, the azimuth angle. In addition, a new pattern of electrodes is proposed as a method for achieving five DOF using only half of the electrode connections of traditional DCDEAs. Experimental tests are carried out and used to validate the proposed model. Results show very close agreement. A limiting aspect of the proposed model is that it relies on two experimental correction coefficients. Nonetheless, the model derived provides a means to more accurately implement automatic control to robotic systems that use DCDEAs (work in progress).

As the development of soft actuators progresses, new methods for more agile actuator control are required. While there are hardware technologies that can provide such agility (e.g. field-programmable-gate-array electronics), mathematical models tend to slow the controlling processes. For the case of double conical dielectric elastomer actuators (DCDEAs), complex material and physical models are often utilized; however, such models require extended computational time which limits real-time control and prediction, especially in small, portable applications. The foregoing problem is augmented, if agile applications require reprogramming on the fly . In this work, a parameter-based dimensionless model applicable to DCDEAs was developed. The proposed model: (a) relates a dimensionless displacement with a characteristic dimensionless voltage via a power-law equation; (b) is robust enough to accurately describe 5-Degree-of-Freedom motion; (c) is scalable (at least within the range of geometries and parameters used in this study); (d) due to its mathematical simplicity, it could be used for more agile controls. For the development of the model, multiple DCDEAs were manufactured by varying design parameters such as pre-stretch, thickness, dielectric constant, shear modulus, electrode area, and inner and outer radii. By applying the Levenberg-Marquardt best-fit algorithm to experimental data sets containing voltage and (translational and rotational) displacements, a general relationship was found between parameter-based dimensionless quantities and a general dimensionless displacement. The relationship was further simplified via a single dimensionless number related to the actuator design. The model was found to be accurately predictive for VHB4910 and VHB4905. The resulting model is simple and has potential to be used for a wide range of dielectric-elastomer materials.

Hyperelastics.jl is a Julia (Bezanson et al., 2017) implementation for the largest (70+) collection of hyperelastic material models in existence. The package provides a set of analytical and data-driven strain energy density functions (SEDF) and the tools required to calibrate the models to material tests. The package is designed to leverage multiple-dispatch to define a common set of functions for calculating the SEDF, Second Piola Kirchoff stress tensor, and the Cauchy stress tensor. The package provides: 1) a material model library that is AD compatible and 2) a set of extensible methods for easily defining and testing new material models. The package leverages the ContinuumMechanicsBase.jl pacakge for defining the continuum scale quantities and their corresponding relationships.