Comparative Analysis of Simulation Models for Digital Twins and Damage Identification on Bridges: A Steel Bridge Case Study Using Strain Measurements

Abstract

Structural Health Monitoring (SHM) and Digital Twin (DT) technologies are transforming the management of civil infrastructure by enabling real-time condition assessment and predictive maintenance. A key component of these advancements is the development of accurate simulation models that represent physical assets in virtual environments. This paper explores different simulation approaches for DTs and damage identification on bridges, using a steel bridge with an orthotropic deck as a case study. Various state-of-the-art modeling techniques, including truss models and finite element models, are developed based on the bridge’s as-designed geometry. Their performance is evaluated in terms of computational efficiency, ease of parameterization, and accuracy in capturing structural behavior, with strain measurements from a load test serving as a reference. Statistical metrics assess the alignment between simulated and observed data. While all models involve certain simplifications, they demonstrate good agreement with experimental results. The findings highlight the importance of selecting an appropriate modeling approach based on specific SHM and DT objectives, such as real-time monitoring or detailed structural analysis. While truss models offer advantages in computational efficiency and straightforward parameterization, more detailed finite element models provide higher fidelity for local structural behavior. This study underscores the potential of even simplified models to achieve reliable predictions of global structural behavior when key geometric features are appropriately represented. Ongoing work further refines these approaches to enhance their applicability in damage identification, contributing to the advancement of DT-based predictive maintenance for bridges.