Seismic Risk Scenario in Grenoble (FRANCE) Using Experimental Dynamic Properties of Buildings

Assessing the vulnerability of a large set of buildings using sophisticated methods can be very time consuming and at a prohibitive cost, particularly for a moderate seismic hazard country like France. We propose here a low-cost analysis using an experimental approach to extract the elastic behaviour of existing buildings. An elastic modal model is proposed for the different types of building tested in Grenoble (France) thanks to their experimental modal parameters (resonance frequencies, modal shapes and damping), which are estimated using ambient vibrations surveys. Sixty buildings of various types were recorded. The building integrity is then calculated considering an accelerogram scenario provided by seismologists as input and considering an integrity threshold based on the FEMA inter-storey drift limits. Even if the level of damage remains unknown, we conclude that masonry buildings undergo more damage (70% of buildings damaged) than RC buildings. Finally, extracting modal parameters from ambient vibration recordings allows us to define, for each class of building, its ability to support seismic deformation in case of earthquake.

💡 Research Summary

The paper addresses the challenge of assessing seismic vulnerability for a large inventory of existing buildings in a region with moderate seismic hazard, where conventional detailed analyses are often prohibitively expensive and time‑consuming. The authors propose a low‑cost, experimentally driven methodology that relies on ambient vibration recordings to extract the elastic dynamic properties of each structure. Sixty buildings of various typologies (masonry, reinforced‑concrete, and mixed systems) located in Grenoble, France, were instrumented with a small number of sensors per building. Continuous ambient vibration data were collected for 15–30 minutes per site. Using stochastic subspace identification (SSI) and frequency‑domain decomposition (FDD), the authors identified the fundamental natural frequencies, mode shapes, and modal damping ratios for each building. These modal parameters were then used to construct a simple linear elastic modal model for each building class. Masses were estimated from floor area, height, and material densities; stiffness matrices were derived from the identified frequencies and masses; and the experimentally measured damping ratios were directly incorporated.

The elastic models served as the basis for a deterministic seismic scenario analysis. A synthetic accelerogram, supplied by seismologists and calibrated to represent a realistic Grenoble earthquake (including appropriate peak ground acceleration, duration, and spectral content), was applied as input to all models. The resulting floor‑by‑floor inter‑story drift responses were computed and compared against the FEMA 356 drift limits (typically 0.5 %–1.5 % of story height). Buildings whose maximum drift exceeded the prescribed limit were classified as “damaged.”

The results reveal a stark contrast between construction types. Approximately 70 % of the masonry buildings exceed the FEMA drift thresholds, indicating a high likelihood of significant damage under the considered scenario. In contrast, reinforced‑concrete (RC) structures show a much lower damage probability (below 15 %), reflecting their higher stiffness and inherent damping. Mixed‑type buildings display intermediate performance, suggesting that the interaction between RC frames and masonry infill walls critically influences seismic resilience.

Key insights emerging from the study include:

1. Feasibility of Ambient Vibration Testing – Ambient vibrations, which require only short recording periods and minimal equipment, can reliably capture the essential modal characteristics needed for a first‑order seismic assessment. This dramatically reduces the cost and logistical burden compared to forced‑vibration or detailed finite‑element modeling.

2. Utility of Linear Elastic Modal Models – Even though the models are linear, they provide a rapid comparative metric of vulnerability across a large building stock. By focusing on inter‑story drift, the approach aligns with widely accepted performance‑based criteria and can be integrated into municipal risk‑mapping workflows.

3. Typology‑Based Prioritization – The pronounced vulnerability of masonry structures suggests that city authorities should prioritize retrofitting or detailed evaluation of these buildings, especially in historic districts where such typologies dominate.

4. Scalability and Transferability – The methodology does not depend on detailed as‑built documentation, making it applicable to other regions with limited structural inventories. Moreover, the extracted modal databases could be linked to real‑time sensor networks for continuous health monitoring and rapid post‑event assessment.

The authors acknowledge limitations: the linear elastic assumption cannot capture post‑yield phenomena such as cracking, crushing, or pounding, which may lead to under‑estimation of actual damage. Ambient vibration data can be contaminated by cultural noise, requiring careful preprocessing. Future work is proposed to integrate nonlinear pushover analyses, develop hybrid models that combine experimental modal data with simplified plastic hinges, and explore automated data acquisition using wireless sensor networks.

In conclusion, the study demonstrates that ambient‑vibration‑derived modal parameters, when embedded in simple elastic modal models, constitute an effective, low‑cost tool for preliminary seismic risk screening of large building inventories. The approach yields actionable insights—particularly the heightened vulnerability of masonry buildings—supporting evidence‑based decision‑making for seismic retrofitting programs and emergency preparedness in moderate‑hazard regions like Grenoble.