Differential MPM for System Identification and Structural Monitoring
Keywords: MPM, Structural Health Monitoring, Bridge Monitoring
TL;DR: We propose using differential MPM to approximate the material properties and loading conditions of structures for long-term monitoring.
Abstract: Nondestructive monitoring of existing structures is a challenge for structural engineers. There are many types of readily-available sensing technologies that can provide information regarding surface conditions of the structure. As such, there is a demand for computational techniques that can leverage this surface information to inversely solve for internal conditions throughout the structure. Differential material point method (MPM) is appealing for this inverse problem, as it can provide the approximation of material properties and load conditions from surface information via gradient-based optimization.
This presentation covers current work of how differential MPM can be used to approximate structural conditions in 2D examples. In addition, we discuss how differential MPM will be applied to a real structure: CFM-5 bridge. CFM-5 bridge in Barkston Ash, UK presents an asset monitoring and maintenance challenge. Since its construction in 1868, the masonry arch rail bridge has suffered damage and undergone a variety of repairs. The uncertainty of loading conditions, construction material properties, and the repairs' effectiveness motivates the need for data-driven system identification for automated monitoring. Fiber optic dynamic strain sensors have been installed along the underside of the bridge, and differential MPM will be applied to a model of the bridge and the dynamic strain data to approximate the Young's moduli of discretized portions of the bridge.
Submission Number: 18
Loading