A comprehensive statistical analysis was performed on the condition data generated by a statewide underwater inspection of bridges in Louisiana. The research defined pertinent underwater structural decay characteristics and established a method for determining a bridge's propensity for underwater deterioration on the basis of those relevant factors. The inspection results were reviewed to ascertain the impact of human bias on consistency in the bridge condition ratings. Using a set of importance factors obtained through a questionnaire submitted to a group of inspectors, the fuzzy set theory was employed to assist in the removal of data incongruities. A combination of pure fuzzy set theory and classical binary weighting was found to produce optimal results. Bridge age, material type, and location were found to be significant in defining the rate of deterioration in Louisiana bridges. Linear least-squares, piecewise linear least-squares, and polynomial regression curves were matched to the overall underwater condition ratings. Comparisons of regression curves for neighboring states indicated reasonable consistency in results for differential inspection programs if similar inspection methodologies are employed. The correlation between above-water and subsurface inspection ratings was found to be poor for concrete and steel bent bridges but acceptable for timber bent bridges within a given age group. Additionally, a poor correlation was found between water quality data and underwater bridge deterioration rates. A methodology for determining the frequency and detail of future underwater inspection projects was developed on the basis of the deterioration trends and available bridge decay-defining characteristics discovered in this research.


  • English

Media Info

  • Features: Figures; References; Tables;
  • Pagination: p. 139-146
  • Monograph Title: Steel, concrete, and wood bridges
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 00711801
  • Record Type: Publication
  • Files: TRIS, TRB
  • Created Date: Sep 25 1995 12:00AM