This paper presents a new proposed methodology for the translation of bridge CoRe elements inspection data into the FHWA NBI ratings. The existing FHWAs BMSNBI translator is first discussed, as well as its application to Florida bridge inventory. Using a set of inspection data where the NBI ratings were obtained simultaneously with the Pontis® element inspection data, the accuracy of the translator was assessed. A new proposed translator (NewTranslator) was developed in the Microsoft Visual Basic programming language, based on the computation of an element condition index and the use of these element indexes to compute a weighted index as NBI ratings for bridge major components (deck, superstructure, or substructure). For NBI ratings higher than 6, the NewTranslator program showed better accuracy than the BMSNBI Translator on bridge decks and superstructures and about the same accuracy for substructures. The deterioration patterns of the data from the NewTranslator were observed to be closer to those portrayed by the actual inspected ratings, except that the NewTranslator tended to show higher NBI ratings, even for older bridges. The overall network condition of bridges, both in terms of specified NBI condition ratings and the Bridge Health Index, was also better represented by the NewTranslator program. Statistical analyses using Kolmogorov-Smirnov and t-Tests indicated that the NewTranslator results are not precisely the same as the inspected ratings. Finally, discussions are presented on the elements smart flags, the different scales in element inspection versus NBI inspection, and optimal relative weights for each element within a bridge component.
Abstract