A major challenge of infrastructure management is to predict the remaining capacity of degrading structures and safely prolong their lifetime. In reinforced concrete (RC) structures, concrete cracking has a significant effect on durability and stiffness properties. Structural integrity degradation is often assessed by estimating the global stiffness loss through vibration-based structural health monitoring. Yet, this is challenging as the modal characteristics might also be affected by environmental and support conditions. At the same time, the development of models that enable studying the modal characteristics of cracked concrete structures has received little attention so far. This paper proposes a novel, visual inspection-based method to predict the decrease in effective elastic moduli of existing concrete structures from observed longitudinal and transverse cracks which are typical for corrosion and load-induced damage in RC elements. Discrete and smeared finite element models are developed to establish a relation between the geometrical crack properties and the changes in the concrete's smeared dynamic stiffness parameters, as defined within an orthotropic material model. It is found that the crack pattern has a significant influence, with transverse cracks generally reducing the stiffness parameters more than longitudinal cracks. Experimental data support the proposed relations’ ability to tune the parameters of the orthotropic material model based on crack properties from corroded or mechanically loaded RC beams. The proposed relations enhance the assessment of serviceability limit states in RC beams and offer a valuable tool to evaluate dynamic test data obtained from on-site monitoring.

This paper presents a validation of a recently developed joint input-state estimation algorithm for force identification and response estimation in structural dynamics, using data obtained from in situ experiments on a footbridge. First, the algorithm is used to identify two impact forces applied to the bridge deck. Next, the algorithm is used to extrapolate measured accelerations due to wind loading to unmeasured locations in the structure. The dynamic model of the footbridge used in the system inversion is obtained from a detailed finite element model, that is calibrated using a set of experimental modal characteristics. The quality of the estimated forces and accelerations is assessed by comparison with the corresponding measured quantities. In both cases, a very good overall agreement is obtained.