Structural Condition Assessment with Structural Health Monitoring Systems and Nonlinear Simplified Models
How can we tell if a tall building is damaged right after an earthquake? This study shows that tracking frequency changes alone is not enough, and combines them with inter-story drift ratios estimated through an optimized nonlinear simplified model — so even floors without sensors can be assessed. New threshold values are proposed for deciding a building's performance level, and the method is validated on a real earthquake-damaged building.
DOI: 10.1002/eqe.4132 ↗Implementation and Application of a SHM System for Tall Buildings in Turkey
SHM systems have been mandatory for tall buildings in Turkey since 2019. This paper shares real long-term monitoring experience from three instrumented high-rises in Istanbul — including modal identification before, during and after a Mw 5.7 earthquake. A Timoshenko-beam model updated with measurements estimates the motion of non-instrumented floors, enabling precise drift-based condition assessment, and the in-house real-time monitoring software platform is introduced.
DOI: 10.1007/s10518-021-01230-w ↗Modal Identification of the First Bosporus Bridge During Hanger Replacement
When the 1973 Bosporus Bridge had its inclined hangers replaced with vertical ones in 2015, our team measured the bridge's vibrations at every stage. Deck-dominated modal frequencies dropped by 6–16% while tower modes stayed nearly unchanged — findings confirmed by finite element models of both configurations. A rare, full-scale record of how a major retrofit changes the dynamics of a landmark suspension bridge.
DOI: 10.1080/15732479.2020.1717551 ↗System Identification and Seismic Performance Assessment of a Stone Arch Bridge
Historic masonry bridges carry big unknowns — material properties, boundary conditions, modeling assumptions. On a railway line crossing the North Anatolian Fault with 41 registered historic bridges, modal properties of 12 stone arch bridges were identified from ambient, impact and train-passage vibrations. Finite element models were then updated to match reality, giving a reliable basis for the seismic assessment of heritage structures.
DOI: 10.1080/13632469.2019.1692740 ↗Instantaneous Damage Detection of Bridge Structures and Experimental Verification
Can we watch damage happen in real time during an earthquake? Using an extended Kalman filter, elemental stiffness values of a bridge are identified instant by instant from vibration measurements — locating and quantifying damage as it develops. Verified on a large-scale shake table test of a three-bent concrete bridge, where identified stiffness losses matched the observed damage.
DOI: 10.1002/stc.229 ↗Large-Scale Shake Table Test Verification of Bridge Condition Assessment Methods
A fundamental question of monitoring: can low-amplitude vibrations measured before and after an event reveal how much stiffness a structure has lost? Two large-scale shake table experiments on reinforced concrete bridge specimens — progressively damaged by increasing ground motions — show that identified stiffness degradation is consistent with the experimental hysteresis and can be quantitatively linked to the remaining capacity of components.
DOI: 10.1061/(ASCE)0733-9445(2008)134:7(1235) ↗Long-Term Monitoring and Identification of Bridge Structural Parameters
What does five years of continuous monitoring actually tell us? A 111-m concrete bridge instrumented with 13 wireless accelerometers produced 1,707 traffic-induced vibration datasets. Using neural networks, stiffness, mass and soil-spring values were identified year by year: about 5% decrease in the first modal frequency and 2% in superstructure stiffness over five years — a template for evidence-based, long-term asset management.
DOI: 10.1111/j.1467-8667.2008.00572.x ↗Structural Reliability Estimation with Vibration-Based Identified Parameters
Monitoring data is most valuable when it changes the risk numbers. Here, stiffness and damping identified during damaging shake table tests are injected into reliability analysis: Monte Carlo–generated structural parameters are updated in a Bayesian sense with the measurement-based results. The residual reliability computed with updated parameters comes out lower — showing that ignoring monitoring data can overestimate the safety of a damaged structure.
DOI: 10.1061/(ASCE)EM.1943-7889.0000066 ↗Damage Detection Based on Damping Analysis of Ambient Vibration Data
Most damage detection needs a healthy "baseline" to compare against — this method doesn't. By extracting nonlinear damping from ambient vibrations with the random decrement technique, seismic damage in a large-scale concrete bridge model was identified without any pre-damage reference: nonlinear damping grows as damage grows, and correlates strongly with stiffness loss.
DOI: 10.1002/stc.296 ↗