FE Model-Based Bridge Health Monitoring
By
Alex Krimotat
Vice President
SC Solutions, Inc.
Sunnyvale, CA
Abstract:
Civil infrastructure such as bridges, roadways, and water pipelines, are critical to providing mobility, clean water, and other elements essential to maintaining an acceptable quality of life for the citizenry. A vast number of U.S. infrastructure facilities are in dire need of upgrade or replacement. A recent study reports the partial and total collapse of over 130 bridges in the United States from 1989 to 2001. Another report noted that in August 2007, fully a quarter of the nation’s highway bridges (approximately 600,000) were either structurally deficient or functionally obsolete. The potential cost of such an undertaking is very large. The total price tag to adequately address all of America’s infrastructure needs, ASCE concluded in 2005, would be $1.6 trillion over a five-year period. This figure excluded security needs, which have not yet been adequately assessed.
This presentation focuses on a research program that will transform the way in which safety and reliability of our nation’s bridges are ensured. The technology that will be developed will answer the following critical question that every bridge health monitoring system needs to address unequivocally:
“Is the bridge OK?”
The answer may be expressed in terms of simple actionable choices, e.g., using three levels (1) green (OK), (2) yellow (Warning: Inspect specific places on the bridge for corrective maintenance), or (3) red (shutdown the bridge immediately).
The high-level steps involved in the predictive health maintenance of a bridge, as presented here are:
Development of a high-fidelity Finite Element (FEM) model of the bridge.
Determination of the critical locations for sensor placement on the bridge.
Installation of the sensors on the bridge.
Calibration of the bridge model using sensed data.
Development of acceptable limits of excursions of sensed variables.
Real-time comparison of sensors with tabulated limits to provide real-time predictive maintenance decisions.
Many DOTs possess FEM models of their bridges. Many of them were developed using reliable software such as GTSTRUDL, and used as basis of the design and construction. Utilization of these models is one of the key elements of efficient and successful implementation of the model-based bridge health monitoring system approach described in this presentation.
