A mechanistic approach to infer the load capacity of highway bridges with insufficient as-built data

Bridges are a fundamental component of any country's highway system. Therefore, the load carrying capacity of these structures should be assessed periodically in order to establish whether any measure should be taken to guarantee their safety such as restricting the load of passing vehicles, retrofitting or even structure replacement. The data contained in as-built documentation is one of the most important inputs required for a reliable bridge evaluation. However, the absence of this documentation has been identified as a common challenge that engineers must face. While the lack of complete data can be partly overcome with the help of field and laboratory testing, expert knowledge is usually relied on as a way to supplement missing information. In this paper, a methodology is proposed for the supplementation of missing as-build data, based on the review of past design and building practices, including common bridge design codes and loads. Knowing general geometry data, which can be obtained from available inventory data, as well as from simple on-site surveys, a method tool has been developed to automatically propose an expected bridge design, consistent with the year of construction. The structure detailing, resulting from this design, is then used to fill in for the missing data, allowing for an analytical evaluation to be conducted. The proposed method has been developed as part of a project to build a new Bridge Management System for the "Instituto Nacional de Vias", the Colombian government authority responsible for managing national public highways and is limited to conventional Reinforced and Prestressed Concrete Girder-slab bridges. The method has been verified by means of comparing predicted and actual data using bridges for which documentation is available.