PITA Fiscal Year 2010 Projects - Innovative Infrastructure Systems

Bridge Performance Evaluation Using Integrated Finite Element Models and In-Service Monitoring Data

Principal Investigators: Shamim N. Pakzad, John L.Wilson

The advances in micro-electro-mechanical-systems (MEMS) technology as well as wireless communication in the past decade have lead to a new generation of wireless sensor networks (WSNs) which are capable of measuring structure’s response, performing on-board filtering and data processing tasks, and relaying the information to base stations for further analyses. The low manufacturing, instrumentation and maintenance cost of these devices has made it possible to deploy dense sensor networks on important structures to monitor its condition on a permanent or semi-permanent basis. A significant area of related research is the integration of WSNs with advanced mathematical models of the structure to accurately diagnose its condition and provide prognosis for its future behavior and maintenance. A class of such mathematical methods is finite element models, which are widely used in the design and retrofit stages of structures, but are often not verified and validated by measured data. The focus of this project is the initial development of a methodology to merge the measured response of a monitored bridge to ambient or strong motion excitations, with a finite element model of the bridge and update the model to represent the current status of the structure. In this context, finite elements model updating will be studied in three general categories:

  1. Model updating for stochastic systems using features of the data (modal properties for example).
  2. Model updating for input/output systems using time and frequency-domain histories.
  3. Adaptive model filters for online real-time updating.

The following general tasks are required to achieve these objectives:

  • Creation of an adequately complex finite elements model for the bridge.
  • Proper parameterization of the FE model.
  • Feature extraction from the measured data.
  • Parametric optimization to update the FE model.
  • Validation of the updated model using reference data.

Development and implementation of such system will provide unprecedented insight into the condition and behavior of the structures. The bridge engineers and owners will be able to use this information to evaluate the performance of the structure under ordinary or extreme loads in real time and plan appropriately to remedy the possible structural deficiencies.