Non-linear Dynamics group | Department of Engineering Science | University of Oxford

The Non-Linear Dynamics Group's Research

The group's work is focused in the study of Dynamic Systems that exhibit Non-Linearities. Within that framework, the directions of modeling, testing, identifying and mitigating the response of such systems are of interest. Current areas of research include:

• System Identification, Structural Health Monitoring and Observability of nonlinear structural systems
• Modeling Non-Linear Dynamic Systems
• Experimental Testing

The PI

Manolis Chatzis graduated from the National Technical University of Athens in 2007 with a Diploma in Civil Engineering and obtained an MSc from NTUA in Structural Engineering in 2008. He then joined Columbia University in the City of New York where he completed his PhD in 2012 and continued in Columbia as a Post Doctoral Research Scientist. He joined Oxford in 2013 as an Associate Professor. Manolis and his group work in improving the methods for modeling, identification and testing of non-linear dynamic systems.

System Identification, Structural Health Monitoring and Observability of nonlinear structural systems.

The group focuses on extending the methods that allow us to talk about the observability of non-linear systems for a sensor setup which measures some of the outputs. We have recently extended the observability framework to non-smooth systems and systems where some of the inputs are not being measured, and are working towards efficient implementations that allow for studying larger and more complex systems. We further work on modifying system identification algorithms to increase their robustness when dealing with more complex systems. We have recently suggested modifications of Non-Linear Kalman Filters for non-smooth dynamic systems. The improved algorithms are used for Structural Health Monitoring of infrastructure elements, i.e., using dynamic sensors such as accelerometers on those elements to detect damage and quantify it. Applications of interest include bridges, buildings and wind turbines. Another direction of focus is vibration monitoring in museums and museum exhibits.

Modeling Non-Linear Dynamic Systems

A significant focus within this direction is to accurately model and quantify the risk for rocking components. Those are components not tied or bolted to their support medium, unlike what is typically done in Civil Engineering structures, and thus are allowed to slide, rock and twist when subjected to ground motions. This applies to both structural components as well as non-structural, such as seismically isolated buildings, museum artifacts, hospital equipment, electrical transformers, supercomputers, intermodal containers or even household equipment. This direction further deals with estimating the risk of such systems and proposing suggestions to mitigate it. The group also focuses on studying behaviors related to damage and modeling the response of infrastructure elements as a complementary part to structural health monitoring. We have also interest in novel types of base isolation and magnetic levitating platforms.

Experimental Testing

Aside of the deployment of field sensors for the purposes of Structural Health and Vibration Monitoring, the group has an interest in design experiments that involve the hydraulic actuators and electric biaxial shake table of the Structural Dynamics Lab. The fusion of heterogeneous data is a direction of interest in various experiments. A new direction of interest is that of Real-Time Hybrid Testing of components. We are interested in designing the experiments to test structural and non-structural components and improving the algorithms for achieving reliable tests.