Smart database design for concrete durability analysis - An application in the Hongkong-Zhuhai-Macau bridge

Abstract

A complete durability database that provides consistent, reliable information is of great significance for concrete durability analysis. The critical challenge is that most durability databases are quite site-specific. When analyzing a local site, a model developed from the generic database may not be fully applicable to a local site. On the other hand, a model purely developed from limited site-specific data may be very imprecise due to significant statistical uncertainty. This study proposes a novel data-driven method for constructing a smart database for concrete durability analysis. The proposed database design is based on a smart hybridization of site-specific and generic databases, in which site-specific data governs when site-specific data are plentiful, but generic data governs when site-specific data are scarce. This method can handle the missing data which are usually encountered in site investigation and quantify the uncertainties. To demonstrate the effectiveness of the proposed method, a real application that provides the durability data for Hongkong-Zhuhai-Macau (HZM) bridge is presented. The main issues regarding the database design include data preconditioning, missing data recovery, and site-specific data transformation. The results showed that the proposed approach could significantly capture correlation patterns in site-specific data and, moreover, can make meaningful predictions even when data is scarce.