- Phononic Crystals and Metamaterials with Microstructurally Controlled Wave Manipulation Capabilities
Phononic crystals are structural materials consisting of crystal-like periodic assembly of unit cells. We are interested in designing phononic crystals with enhanced and tunable wave control properties to be used as acoustic manipulators and adaptive vibration and sound isolation devices. Specifically we are interested in realizing crystals exhibiting substantial directionality and beam forming capabilities at low frequencies, a realm in which conventional phononic crystals operating in bandgap mode fail. Moreover, we envision a type of wave control that is operated by tuning the microstructural properties of crystals, i.e. by actively modifying the internal geometric and material properties of the unit cells without altering the global static characteristics of the primary cell network. We are exploring techniques for semi-active wave control based on piezoelectric microstructures as well as strategies involving nonlinear resonators for amplitude-dependent tunability. - Wave Manipulation via Granular Phononic Crystals
We explore the phonon manipulation capabilities of nonlinear and granular phononic crystals (GPCs) for application as smart metamaterials for vibration isolation and wave control. Granular phononic crystals consist of networks of particles arranged in space according to lattice topologies, featuring the periodicity and the phononic characteristics of crystal structures. As a result of the tight interplay between the dispersive mechanisms resulting from their topologies and the nonlinearities due to the inter-particle contact dynamics, GPCs feature a rich frequency- and amplitude-dependent dynamic behavior, which can be exploited to manipulate the elastic waves (phonons) propagating in the material. On a fundamental level, our investigation also aims at further unveiling and demistifying the complexity of wave propagation in nonlinear dispersive solids, by systematically mapping the distortion mechanisms to the topological and material properties of the media that generate them. - Anomaly Detection at the Intersection of Mechanics and Machine Learning
We are developing a body of strategies for data-driven, and model-agnostic detection of structural anomalies, defects and damage zones in highly heterogeneous solids. The conceptual keystone of our approach is the notion that the data structure of the response of a system to a mechanical stimulus provides a data map of the system and of its potential anomalies. This map contains the sparse code of the medium, where the adjective sparse implies that the anomalies manifest themselves as distinct and localized deviations from an otherwise shared (or typical) behavior. The task of detecting anomalies reduces then to the problem of learning the sparse code of the medium and finding the salient features (e.g., features that sufficiently deviate from he common behavior of the surrounding medium) in its mechanical response. To this end, we propose a body of methodologies that couple the spatial data richness and acquisition agility available via non-contact scanning laser sensing modalities with the data learning and classification capabilities of computer vision (CV) and machine learning (ML) algorithms.
Research Topics
Note: this page is a permanent work in progress. Research interests are broad and ever evolving and are only partially captured by the bullets below. The publication list provides a more comprehensive and updated overview of the group's activities.