Experimental and numerical behavior of a deployable tool for cross-frame installation

The installation of cross-frames in highly skewed and curved steel girder bridges poses significant erection challenges. A common problem is that the girders deflect and twist under their own self-weight, creating differential vertical displacements and/or rotations that may prevent the fit-up of cross-frames. This paper experimentally and numerically investigates the behavior of a deployable tool designed to overcome this fit-up challenge and enable the installation of cross-frames without force fitting. The tool, made up of a double-acting hydraulic jack and two tension members (e.g., cables, threaded rods), was experimentally investigated using a one-quarter scale test setup representing a portion (i.e., two girder lines of limited length) of a built bridge during one stage of erection where cross-frames would be installed between two adjacent girders. A total of seven scenarios were investigated, with varied parameters including (1) type of geometric misalignment between girders (differential vertical displacement, differential rotation, or combination thereof) and (2) magnitude of geometric misalignment. The behavior of the girders and the tool was monitored as the tool is deployed (i.e., the hydraulic jack is extended in displacement control) to achieve the desired geometry adjustment between the girders. Measured behavior was compared with numerical predictions from a three-dimensional finite element model. Findings from this scaled study demonstrate the tool's potential to provide the necessary geometry adjustment to enable cross-frame installation. This work highlights the potential for the tool to enhance constructability and safety during the bridge erection phase, offering a practical solution to improve efficiency in the field. • Tool for cross-frame installation is numerically and experimentally evaluated. • Scaled experiment demonstrated capability of tool to correct misalignments. • A finite element model was developed that generally predicted behavior. • Recommendations for the deployment of the tool are provided.