Abstract This paper presents an extensive shaking table campaign conducted on free-standing rocking blocks subjected to recorded earthquakes. The campaign investigates different sources and degrees of variability in the rocking response, considering material irregularities, differences in block aspect ratios, and variability in the input ground motions. The experimental programme consists of the development of experimental incremental dynamic curves, incorporating test repetitions, thereby addressing the repeatability limitation of most experimental campaigns, resulting in a dataset of over 300 shaking table tests. The analysis of the results is organised in a three-stage framework. First, the rocking response is examined through a deterministic approach based on time-history analyses. Subsequently, a statistical evaluation is conducted to explore the experimental relationships between engineering demand parameter and selected intensity measures. Finally, the pronounced influence of various sources of uncertainty on the rocking response highlights the necessity of a probabilistic interpretation of the data, culminating in the development of experimentally informed fragility curves. This paper introduces a novel methodology for the objective identification of experimental limit states for rocking structures, proposing ranges for three response thresholds corresponding to limited rocking, moderate rocking, and near-collapse conditions that can be adopted for vulnerability assessments.
Colombo et al. (2026) studied this question.