Interaction between Mass and Stiffness Parameters of Connections for Structural Parameter Estimation of a Steel Grid

Abstract

In any structural system, the mechanical properties of its connections impact the structure’s global performance. Particularly, the rigidity of joints plays a substantial role in the system’s vibration response. Since connections typically include several components with complex geometry, they are usually oversimplified in the modeling process. Therefore, the analytical model is typically a low-fidelity representation of the connections’ performance. Although simplifying joints for modeling purposes is a good practice for engineering applications, it may reduce the accuracy of the model predictions. Thus, it is essential to reconcile accuracy with efficiency, particularly for complex joint modeling. In addition to the rigidity, the mass of connections may be another influential structural parameter for the system’s global response. The joint mass may be underestimated in the modeling procedure because the elements of joints in the steel connection regions, including plates, stiffeners, and bolts, are not often modeled in simplified analyses. Hence, as the analytical model is updated, connections’ mass and stiffness need to be estimated and incorporated into the updated model. This paper studies the interaction of mass and stiffness estimation for bolted connections of an experimental steel grid. The structure used in this research is a well-known structure built, instrumented, and tested at the University of Central Florida. The laboratory setup simulates the behavior of deck-on-beam bridges. In its initial analytical model, all grid joints are assumed to be completely fixed, and no extra mass is considered in the joint regions. The model updating procedure is limited to estimating joints’ mass and stiffness values. The modal information of the structure is extracted from the measured frequency response functions of the grid excited in a vertical hammer impact test. Stiffness- and flexibility-based error functions are employed for the model updating. The results demonstrate how the selection of joints’ uncertain parameters can affect the procedure of finding suitable structural layouts for the semi-rigid connections of the grid. The model updating procedure may be generalized to condition assessment and structural health monitoring of bridges, especially when their connections are subject to damage and failure.