Open Access Open Access  Restricted Access Subscription or Fee Access

Implementation of a Structural Impairment Detection System on a 100 Year-Old Bascule Bridge



Critical instances of reliable bridge functionality arise in situations where waterway transportation conflicts with bridge traffic. Bascule bridges, or drawbridges, are one class of movable bridges that can be designed to accommodate both flows of traffic. While popular, heel trunnion bascule bridges have not been without problems and failures. Bascule bridges are subject to multiple cycles of opening and closing on a daily basis, and thus structural members undergo large stress ranges. This paper presents an implementation of a Structural Impairment Detection System (SIDS) that incorporates finite element modeling, instrumentation, and continuous evaluation of a testbed bascule bridge structure. Heel trunnion bascule bridges experience significant stress ranges in critical truss members. Finite element modeling of a testbed bridge provided (1) an estimate of nominal structural behavior, (2) an indication of types and locations of possible impairments, (3) a basis for the design of an instrumentation program appropriate for detecting possible impairments, and (4) data streams with which to train a neural SIDS. Analytical modeling was initially performed in SAP2000 and then refined with ABAQUS. Finite element models and experimental observations indicated maximum stress ranges of approximately 22 ksi on main chord members of the counterweight truss. Instrumentation used to observe structural behavior included electrical resistance strain gages, clinometers, and quadrature encoders. Ultimately, data streams from the testbed bridge are autonomously recorded and interrogated by competitive arrays of artificial neural networks for patterns indicative of specific structural impairments. A quasi static array of competitive neural networks was developed to provide an indication of the operating condition at specific intervals of the bridge’s operation. Based on neural algorithms trained on modeled impairments, the testbed operates in a manner most resembling one of two operating conditions: 1) unimpaired, or 2) an impaired member embedded at the southeast corner of the counterweight.

Full Text: