Sensitivity Analysis in Structural Health Assessment Using Generalized Polynomial Chaos Expansion: a Full-scale Bridge Test Case

Kurzfassung

The contribution addresses the growing need to understand and protect bridge structural integrity amid crises, aging infrastructure, rising traffic loads, and climate-driven extreme weather events. It applies generalized Polynomial Chaos Expansion (gPCE) to build a surrogate model for a benchmark bridge at the University of the Bundeswehr Munich, enabling extensive monitoring studies. A sensitivity analysis is then carried out across various mechanical, loading, and damage conditions - particularly foundation settlement, a leading cause of failure in river and coastal structures - to reveal how different damages manifest in measured outputs and to assess their statistical relevance for structural reliability. The long-term goal is to feed these results into a model-driven decision-making framework that helps practitioners prioritize maintenance and optimize sensor placement, thereby improving safety and resource allocation in infrastructure management.