Keywords: bridge, life-cycle management, deterioration, reliability.

More and more efforts are devoted to the development of better models for bridge life cycle cost analysis (LCCA). One major development in LCCA in recent years is to develop deterioration models for bridge components so that the times a component needs repair or replacement throughout its life span can be properly determined. The subjectivity in conventional LCCA can then be decreased and the accuracy of the results can be improved.

Reliability is the probability for a product to perform adequately under certain stress levels during its age or mission time. An example frequently used to illustrate the concept is strength vs. load. A product will fail if strength is less than or equal to the load. Research in introducing the reliability method for developing bridge deterioration models has been active in recent years. Many research studies can be found in literature [1,2,3]. Most of those studies require experimental data from major bridge inspections, and usually focus on one deterioration factor. In addition, many of them developed their models using a particular bridge component, such as a deck or pier, as the study object. Only very few of those models are developed for conducting bridge life cycle cost analysis.

This study employs the reliability method and develops a deterioration model for bridge components. A bridge component deteriorates yearly in its lifetime. It is a combined result of many possible factors, such as corrosion, chloride, cracks, and so on. Although arguable, this research takes a view that the results of a bridge visual inspection represent the combined acting effect of these factors. A condition index, based on the visual inspection data, is developed in this research to assess the condition of a bridge component.

For a bridge component, the condition in each year should be a random variable and has a distribution. Normally a maximum acceptable condition level can be set to determine if a bridge component will fail or not. Thus, the reliability of a component in a particular year is determined by assessing the probability that its condition is greater than the specified maximum acceptable condition level. As bridges age every year, the reliability of a component decreases every year. Model assumptions and the rationale are described, and the steps for applying the model developed are explained in detail in the paper.

The deck component in a new bridge in northern Taiwan is employed as a subject for demonstrating the use of the reliability-based model developed. This 10-span prestressed concrete bridge has a length of 500 m and a width of 20 m. Fifty-three similar bridges are identified using criteria such as climate, earthquake zone, distance to the shore, etc. Their ages range from 1 to 46 years. Between the data years of 1999-2003, a total of 1,651 records of span inspections of the fifty-three bridges are employed for conducting the regression analysis of the reliability index. The original is only 0.44. By further examining the data records and excluding the 'unreasonable' ones, it is attempted to increase the . Afterwards, the is improved to 0.67, which is more acceptable. However, there is still room for further improvement. The possible causes for the low are discussed in the paper.

In summary, the model developed features the following:

1. It is reliability based, which can simulate more closely the deterioration of bridge components in the real world and provides more accurate prediction of times for taking remedial actions in a component's lifetime. The accuracy of a bridge life cycle cost analysis can be improved as a result.
2. It employs the historic visual inspection data. Although arguable, the visual inspection data represents the combined acting effect of many factors and provides a more "complete" assessment of a bridge component's condition. In addition, for the purpose of conducting life cycle cost analysis, employment of the visual inspection data in the deterioration model should be more cost effective than that of the major inspection data.

1

Carlsson, F., Jeppsson, J., Thelandersson, S., "Reliability Analysis of Corroded Bridge Columns", Proceedings of the RILEM/CIB/ISO International Symposium, 271-275, edited by A. Sarja, Helsinki, Finland, 2000

2

Frangopol, D.M., Ghaaibeh, E.S., Kong, J.S., and Miyake, M., "Integration of Whole Life Costing with Reliability in Optimal Management of Bridge Networks", Proceedings of the RILEM/CIB/ISO International Symposium, 99-103, edited by A. Sarja, Helsinki, Finland, 2000.

3

Kawamura, K., "Method for Estimating Deterioration Factors of Concrete Bridges", Proceedings of the 2nd International Workshop in Life Cycle Analysis and Design of Civil Infrastructure System, 71-81, edited by A. Miyamoto and D.M. Frangopol, Ube, Yamaguchi, Japan, 2001.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £105 +P&P)