Electroactive Polymers As Artificial Muscles and Sensors: A Systems Perspective

Sponsors: National Science Foundation, MSU Intramural Research Grants Program, US Civilian Research and Development Foundation, and International Society for Optical Engineering (SPIE)

Introduction

Control-oriented, Physics-based Modeling 

Integrated Sensory Feedback 

Micromanipulation and Biological Applications

Introduction

Electroactive polymers (EAPs), also known as Artificial Muscles,  are a class of smart materials that show strong coupling between the applied electric field and their mechanical strains.  The latter property enables us to use these materials as excellent actuators and sensors. We are particularly interested in two classes of EAPs, ionic polymer-metal composites (IPMCs) and conjugated polymers (e.g., polypyrrole),  since they produce large bending motions under low actuation voltages (about 1 V), are biocompatible and resilient, with tremendous potential in biomedical devices, micromanipulation, and biomimetic robotics.  Click the following two pictures for video clips  of  graceful movements of an IPMC actuator and a polypyrrole   actuator, respectively.

The goal of this project is to fully realize the potential of these emerging actuation and sensing materials, by taking a systems perspective to  

Control-oriented, Physics-based Modeling 

Our objective in modeling is to develop models for EAP actuators and sensors that capture dominant dynamics and yet are amenable to real-time control design. By taking a control systems perspective, we explore approaches to the derivation of analytical models from governing partial differential equations (PDE’s), with an attempt to bridge the gaps between complex physics-based models and oversimplified empirical models. For example, infinite-dimensional transfer function models have been developed for IPMC and polypyrrole sensors and actuators by exactly solving the (linear) dynamics-governing PDE with appropriately imposed boundary conditions. These models are expressed in terms of fundamental material properties and sample dimensions, and are thus geometrically scalable. Scalability  of the proposed models has been experimentally validated.

Work is underway to extend these models to nonlinear regimes, by drawing on tools in nonlinear elasticity, perturbation analysis, and dynamical systems. Model reduction techniques can be applied to the developed infinite-dimensional models, which leads to convenient model-based controller design. We have demonstrated experimentally the advantages of model-based designs in robust adaptive control and H¥ control of EAP actuators.

Click the picture below to see a video clip  of a custom-built mechanical excitation apparatus . The apparatus generates controlled linear displacement for characterizing the mechanical sensing behaviors of EAP materials. 

Integrated Sensory Feedback

Feedback is essential in ensuring precise control of EAP actuators in the presence of model uncertainties and noises. Using external, bulky sensors, however, is undesirable for EAP actuators in their biomedical, robotic, and micro/nanomanipulation applications because of size/weight constraints and safety concerns. Therefore, compact, accurate sensing schemes are of interest in the development of EAP-enabled systems. Several approaches are being investigated in our lab to detect the displacement/force generated by an EAP actuator in a way that takes no or very little extra space. One approach we have developed is to integrate  thin polyvinylidene fluoride (PVDF)  films with an IPMC to realize a sensory actuator, where PVDF provides feedback on both the displacement and force output of an IPMC actuator. Feedback control has been demonstrated with the integrated PVDF sensory feedback. The picture below shows an integrated IPMC/PVDF sensory actuator.

Micromanipulation and Biological Applications

One promising application area for EAPs is micromanipulation. We have built an IPMC-based micromanipulator and used it for micro-injection of fruitfly embryos.  This operation, necessary for genetic modification of Drosophila embryos, has traditionally be performed manually, which is time-consuming with low yield. With sensory feedback from the PVDF layer, we expect to automate the injection process with accurate force and position control. Shown in the pictures below:  an IPMC/PVDF equipped with a pipette (tip diameter 2 microns) ready for the injection experiment (left), and snapshots of the injection into a Drosophila embryo (right). Click the snapshots below (right) for a video clip of the injection process . We are also investigating the use of such micromanipulators in biomechanical study of single cells.