Application of stochastic system identification to the study of the compliance of electroactive polymers.

Electroactive polymers have shown promising applications as transducers that can mimic biological muscle. The modulus or the compliance of many of these devices can change significantly as they are actuated making these materials attractive for applications that require tunable stiffness. We have developed a dynamic mechanical analyzer that is capable of making in situ measurements of the dynamic compliance transfer function of conducting polymers as a function of an electrochemical stimulus. We do this by simultaneously applying a stochastic stress waveform over a potential waveform and calculating the compliance as it changes over the course of electrochemical excitation. Using these signals we can calculate the compliance transfer function between 0.1 and 100 Hz and the impulse response function with up to 3% variation in its parameters. These functions are then computed as charge is injected into the polymer and it is shown that the low frequency gain of the transfer function can change by 30%-40% in the electrochemical system tested.

Thermo-mechanical characterization of polypyrrole compliance using stochastic system identification.

Conducting polymers such as polypyrrole are studied as novel biologically inspired actuators. Their capacity to generate stresses of up to 5 MPa, strains of up to 10% at low voltages (2 V) make them ideal candidates to be used as artificial muscle materials. It has been shown that the modulus of polypyrrole can change when the material is electrochemically excited. In this paper we develop a technique that uses a stochastic stress input that can be used to measure the compliance frequency response (between 10(-2) Hz and 100 Hz) of polypyrrole in-situ. We validate the compliance calculated from the stochastic stress input by comparing it with the compliance calculated from a single sinusoidal stress input. We also measure the compliance as a function of temperature using both techniques and show that the stochastic compliance follows the same trends as the compliance calculated from single sinusoidal stress input.