Multifunctional Materials Group


                                 Participants
 

Faculty Leader	Ralph Smith
Postdoctoral Fellows	Michael Stuebner Jayasimha Atulasimha
Graduate Students	Lisa Downen Jon Ernstberger Xiang Fan Francesca Reale-Levis Ryan Siskind
Undergraduate Students	Kristopher Kleiner

Ralph Smith	Jayasimha Atulasimha	Lisa Downen	Jon Ernstberger	Francesca Reale-Levis	Ryan Siskind

                                      Research Highlights

The Multifunctional Materials Group focuses on model development, full and reduced-order simualtion techniques, parameter estimation, and control design for a range of materials including PZT, various ferromagnetic compounds, shape memory alloys, polymers and composites, and carbon nanotube-infused polymers.


1.  Model Development for Shape Memory Polymers

Shape memory polymers (SMP) are similar to shape memory alloys (SMA) in the sense that they exhibit the shape memory effect and capability to generate and recover from large strains.  However, they differ in the sense that these effects in SMP are due to complementary components at the molecular level including cross-links that determine permanent shape and switching segments that specify temporary configurations.

Shape memory polymers and composites provide the advantage of being lightweight, flexible, cost-effective, capable of generating larger deformations than SMA, and they can be tailored to meet specific control criteria.  Due to the molecular basis for the shape memory effect in SMP, they can be controlled by light or chemicals as well as by heating.  This provides the possibility for employing light as an input for deployment or shape control.  It also makes the compounds viable candidates for use as biological or chemical sensors.

Ryan Siskind's dissertation research focuses on the development and numerical simulation of constitutive and system models for shape memory polymers employed for high performance actuation. This involves experimental validation using data collected in collaboration with Chris Yakacki, MedShape Solutions, Inc.





2.  PZT-Based Actuators and Sensors

Actuators and sensors employing the ferroelectric material lead zirconate titanate (PZT) are being widely investigated for applications requiring high set point accuracy, large input forces, and broadband capabilities.  These properties make PZT transducers advantageous in applications ranging from laser positioning to vibration isolation platforms.  However, due to its ceramic nature, initial PZT devices proved brittle and were not easily molded to underlying surfaces.  This has recently been addressed through the development of actuators such as THUNDER (THin layer UNimorph Driver and sEnsoR), piezoceramic fiber composites (PFC) and Macro-Fiber Composites (MFC).

Michael Steubner is investigating the development of models and simulation packages for MFC for subsequent control design.  These models are being validated using experimental data collected by Dan Inman and his colleagues at VA Tech.






3. Model Development, Parameter Estimation and Control Design for Terfenol-D Transducers

Terfenol-D actuators and sensors are presently being considered for a range of applications including sonar transduction, high speed milling, torque sensing, and high fidelity loudspeakers.  For all of these applications, however, it is necessary to characterize the inherent material nonlinearities and hysteresis in a manner that facilitates design and real-time control implementation.

Jon Ernstberger's dissertation rsearch focuses on the development of highly efficient and robust parameter estimation algorithms to identify model parameters based on measured attributes of the data.  These algorithms are being tested using data collected by Marcelo Dapino and his group at OSU as well as colleagues at Etrema Products, Inc.

Lisa Downen is investigating the development of numerical techniques to reduce implementation times for algorithms.  She is presently focusing on algorithms for ferroelectric and ferrogmatic materials but due to the universal nature of the modeling framework, resulting techniques will also likely apply to shape memory alloys.

Michael Stuebner and Jayasimha Atulasimha are investigating the development and implementation of homogenized energy models (HEM) in colloboration with Billy Oates (FSU).  They are also investigating the development of control algorithms that are being tested at Etrema Products, Inc.  Xiang Fan's dissertation research focuses on the development and implementation of adaptive control algorithms for magnetic and ferroelectric actuators.


4. Development of Statistical Emulators and Metamodels

The homogenized energy model (HEM), used to characterize the nonlinear dynamics and hysteresis inherent to ferroic materials is based on energy analysis at the lattice level in combination with stochastic homogenization techniques that incorporate material and field nonhomogeneities.  Due to its energy basis, it incorporates a range of input behavior including stress, temperature and frequency dependencies.  However, its incorporation on a range of underlying physics gives it a level of complexity that presently precludes real-time implementation at for very high speed applications (e.g., 10 Khz or above).

Undergraduate Kristopher Kleiner and graduate student Francesca Reale-Levis are investigating the development of Bayesian and Gaussian emulators that incorporate fundamental physics, are highly efficient to implement, and statistically quantify the undercertainty associated with the resulting reduced-order models.

The development of emulators, including Kalman filters, is also being investigated by the NSF-supported REU team comprised of Joel Shor (Princeton University), John Wallace (VA Tech), Travious Glover (Albany State) and Lee Hallock (University of Wisconsin, LaCrosse).