The study, published in the Bulletin of Earthquake Engineering, focused on the Guangyue Tower, a traditional hybrid pavilion-style timber structure located in Liaocheng, Shandong Province, in northern China. Led by Xian-Cai Ren of Xi’an University of Architecture and Technology, together with colleagues from Liaocheng University and Qingdao University of Technology, the research team fabricated a 1/4.5-scale model of the tower and mounted it on a shaking table, a laboratory platform capable of reproducing the ground motions of real earthquakes. Three ground motion records of increasing intensity were applied to the model, allowing the researchers to track how the structure’s behavior evolved from mild tremors to rare, severe shaking.

Shaking table tests are among the most demanding experiments in earthquake engineering. The platform must replicate not only the peak acceleration of a seismic event but also its frequency content and duration, all while the instrumented model records accelerations, displacements, and dynamic properties at every level. For heritage timber structures, the challenge is compounded by the need to faithfully reproduce traditional joinery. The Guangyue Tower model incorporated the characteristic features of Chinese pavilion-style construction, including mortise-tenon joints, dou-gong bracket sets, and column feet that rest on stone plinths without rigid anchorage. These connections are deliberately loose by modern standards, and that looseness turned out to be central to the structure’s survival strategy.

Under strong seismic excitation, the model displayed exactly the kind of damage that conservators might fear: significant loosening at the joints and visible cracking in members. Yet the structural consequences were remarkably contained. The fundamental frequency of the model, a measure of its stiffness, dropped by 22.8 percent as the joints worked loose and energy was absorbed through friction and slippage. At the same time, the corresponding damping ratio increased by 59.3 percent, meaning the structure became far more effective at dissipating vibrational energy. In effect, the damage itself acted as a protective mechanism, softening the building and damping its response before deformations could grow dangerous.

Perhaps the most striking finding concerned acceleration. At every story of the model, the acceleration amplification factor, the ratio of the acceleration at that level to the acceleration at the base, remained below 1.0, and the factors decreased as seismic intensity increased. In conventional stiff buildings, accelerations typically amplify toward the roof, imposing large inertial forces on upper levels. The Guangyue Tower model did the opposite: the ground shaking was largely filtered out before it reached the upper stories. This behavior stems from the sliding and rocking of column feet and the energy dissipation at semi-rigid joints, which decouple the superstructure from the most violent components of ground motion.

Deformation capacity proved equally impressive. Under the rarely met earthquake, the most severe shaking level considered, the maximum interstory drift, the relative horizontal displacement between adjacent floors normalized by story height, reached 1/29. For many modern structural systems, drift ratios of this magnitude would raise concerns about collapse or permanent damage. Yet the timber model showed no obvious residual inclination after the shaking stopped. The structure simply returned to its original plumb position, its joints re-seating themselves as the motion subsided. This self-centering behavior, driven by gravity acting on rocking columns and the elastic recovery of timber, is a hallmark of traditional Chinese timber construction and a property that modern engineers often struggle to replicate with rigid connections.

To extend the experimental findings beyond the laboratory, the team developed a simplified numerical model of the structure and validated it against the shaking table measurements. The finite element model reproduced the dynamic responses of the test specimen with good agreement, capturing the frequency shifts, damping changes, and story-level accelerations observed experimentally. With a trustworthy computational surrogate in hand, the researchers could then ask a question that no ethical test program could answer directly: what would happen if the column feet were rigidly fixed to their foundations instead of resting freely on stone plinths?

The numerical comparison delivered a clear verdict. A model with rigid ground-story column foot connections did reduce the maximum interstory drift, making the structure nominally stiffer and less deformable. But this apparent benefit came at a steep price: the roof acceleration response increased significantly, exposing the upper levels of the building to much larger inertial forces. In other words, fixing the columns would trade controlled flexibility for amplified shaking at the top of the tower, potentially overloading roofs, brackets, and upper-story joints. The flexible, sliding column foot, so often dismissed as primitive, is in fact a deliberate seismic fuse that protects the entire building above it.

The implications reach well beyond a single tower in Shandong. Thousands of multi-story traditional timber structures across China, Japan, and Korea face seismic risk, and many have already been damaged in recent events, including the Ms 6.0 Luxian earthquake of 2022, which harmed numerous cultural heritage buildings. Rehabilitation programs guided by modern strengthening instincts, such as anchoring columns or stiffening joints, could inadvertently destroy the very mechanisms that have preserved these buildings for centuries. The study’s authors emphasize that their findings can serve as a reference for the rehabilitation of multi-story traditional hybrid pavilion-style timber structures, and the message for conservation engineers is unambiguous: preserve the flexible behavior of column foot connections if seismic resilience is the goal.

The work also adds to a growing body of experimental evidence on traditional timber engineering, from shaking table studies of Forbidden City palaces and Japanese pagodas to laboratory investigations of dou-gong brackets and mortise-tenon joints. Together, these studies sketch a coherent picture of an ancient building system that achieves earthquake resistance through energy dissipation, joint loosening, rocking, and self-centering rather than through the strength and stiffness that dominate modern codes. As China continues to invest in the protection of its timber heritage, validated numerical models like the one developed in this study offer a practical tool: they allow engineers to test rehabilitation strategies computationally, predict how a real tower will respond to future earthquakes, and ensure that interventions enhance rather than undermine the ingenious flexibility that has kept these wooden giants upright through the centuries.

Subject of Research: Seismic performance of multi-story traditional Chinese pavilion-style timber structures evaluated through shaking table tests and numerical modeling

Article Title: Seismic performance of multi-story traditional hybrid pavilion-style timber structures: shaking table tests and numerical analysis

Article References: Ren, X.-C., Meng, Z.-B., Wang, S.-W., & Cao, Y. (2026). Seismic performance of multi-story traditional hybrid pavilion-style timber structures: shaking table tests and numerical analysis. Bulletin of Earthquake Engineering. https://doi.org/10.1007/s10518-026-02643-1

Image Credits: AI Generated

DOI: 10.1007/s10518-026-02643-1

Keywords: timber structures, seismic performance, shaking table test, Guangyue Tower, mortise-tenon joints, dou-gong brackets, column foot connections, interstory drift, damping ratio, finite element model, heritage conservation, earthquake engineering

The research team focused on reinforced concrete hospital buildings resting on friction pendulum isolation systems, devices that sit between the building and its foundation and allow the superstructure to slide gently during strong shaking rather than deform violently. Two technologies were compared: the single friction pendulum (SFP) system, a proven design with one concave sliding surface, and the triple friction pendulum (TFP) system, a more sophisticated bearing whose four spherical surfaces engage sequentially as shaking intensifies. The TFP’s adaptive behavior means it is stiff under small motions, softer under moderate earthquakes, and stiffens again under extreme events, effectively retuning itself to each stage of the ground motion.

What makes the study distinctive is its life-cycle, probabilistic framing. Rather than testing a handful of design-level earthquakes, the authors computed annualized performance rates derived from site-specific seismic hazard curves, integrating the effects of every earthquake expected over the building’s lifetime, from frequent moderate shocks to rare catastrophic ones. Collapse safety was quantified through both the probability of collapse and the mean annual frequency of collapse. Structural performance was tracked through interstory drift, the relative displacement between floors that governs damage to both the frame and its partitions, cladding and mechanical systems. Functional continuity was assessed through peak floor accelerations, used as an acceleration-based proxy for the survival of sensitive equipment such as imaging machines, laboratory analyzers and intensive-care devices.

The modeling effort was substantial. Nonlinear response-history analyses were carried out in ETABS, with detailed hysteretic models capturing the behavior of the isolation bearings and the reinforced concrete superstructure. Concrete material properties were calibrated to Mexican production, and the hysteretic models of the frames drew on well-established formulations for confined concrete and degrading reinforced concrete response. The isolation systems themselves were designed under two regulatory approaches: a performance-based procedure aligned with the seismic isolator standard associated with ASCE 7, and the Mexican seismic provisions of Mexico City’s complementary technical norms, reflecting the study’s setting in one of the world’s most seismically exposed capitals.

The choice of Mexico City as the testbed is significant. The metropolis sits on a deep soft-lakebed basin that dramatically amplifies long-period ground motions, exactly the kind of shaking that resonates with flexible isolated buildings. The 1985 and 2017 earthquakes exposed the vulnerability of medical infrastructure, with hospitals suffering structural and nonstructural damage and, in some cases, losing the ability to treat patients precisely when demand for care peaked. The 2020 Mw 7.5 Oaxaca earthquake, arriving during the COVID-19 pandemic, further underscored how fragile hospital continuity can be. The new framework speaks directly to that experience by treating functionality, not just survival, as a design objective.

The results deliver a clear verdict. Across the configurations investigated, the TFP-based hospitals consistently achieved lower collapse probabilities, reduced interstory drift demands, and significantly lower peak floor accelerations than their SFP counterparts. The triple pendulum’s multi-stage sliding mechanism limited bearing displacements and kept the superstructure response within tighter bounds across the full spectrum of hazard, translating into enhanced structural safety and stronger functional continuity over the life cycle. The SFP configuration, by contrast, showed lower reliability, a higher likelihood of damage, and less effective protection of acceleration-sensitive equipment, meaning that even a hospital whose frame survives an earthquake could lose imaging suites, operating rooms or laboratory capacity to shaking-induced equipment failure.

The acceleration findings deserve particular attention from hospital planners. In conventional fixed-base buildings, floor accelerations can reach levels many times the ground acceleration at upper stories, tearing equipment from its anchors and rupturing piping and ductwork. Isolation dramatically cuts these demands, but the study shows the degree of protection depends strongly on the bearing technology. Because TFP bearings maintain lower transmitted accelerations across a wider range of earthquake intensities, they offer a more robust guarantee that medical equipment remains operational without costly post-event inspection and repair. In resilience terms, that difference can determine whether an emergency department admits patients within hours of a disaster or remains shuttered for weeks.

The framework’s annualized approach also changes how engineers can communicate risk. Instead of reporting performance against a single design earthquake with a fixed return period, the method expresses collapse risk, drift damage and equipment vulnerability as mean annual frequencies, quantities that can be summed over decades of service and compared directly across design alternatives. This gives hospital owners, insurers and regulators a common currency for weighing the higher upfront cost of advanced isolation against the expected lifetime losses avoided, including downtime, repair and the immeasurable social cost of a hospital that fails its community. The authors note that the approach unifies collapse safety, structural performance and functional continuity within a single probabilistic assessment, something conventional code-based design does not attempt.

Broader context supports the urgency of this work. Post-earthquake investigations after the 2010 Chile earthquake and the 2016 Kumamoto earthquake in Japan documented how nonstructural damage and equipment failures, far more than structural collapse, degraded hospital function. Researchers have repeatedly found that hospitals are among the most functionally fragile critical facilities, dependent on a chain of power, water, medical gases and functioning devices that shaking can sever at multiple links. By explicitly modeling acceleration-sensitive contents as part of the performance metric, the Mexican study moves hospital design closer to the resilience objectives that agencies such as PEER have long advocated, in which buildings are expected to remain operational, not merely standing, after design-level events.

The study’s implications extend beyond Mexico. As more countries adopt seismic isolation for critical facilities, and as new regulations for isolation in Mexico City come into force, the choice between simpler single-pendulum systems and adaptive multi-pendulum systems will recur in design offices worldwide. The evidence assembled here suggests that for buildings whose mission demands uninterrupted operation, the extra complexity of the triple friction pendulum buys measurable, quantifiable resilience: fewer collapses, less damage, and a far better chance that the lights, monitors and ventilators stay on when the city outside is in crisis. The datasets, synthetic ground-motion records and numerical routines behind the analysis are available from the corresponding authors on reasonable request, offering a foundation for engineers seeking to apply the same life-cycle logic to their own critical facilities.

Subject of Research: Life-cycle seismic performance and functional continuity of base-isolated hospital buildings using friction pendulum systems

Article Title: Life-cycle performance assessment of resilient base-isolated hospitals: collapse safety, structural performance, and functional continuity

Article References: Equihua, E., Flores, E., & Alamilla, J. L. (2026). Life-cycle performance assessment of resilient base-isolated hospitals: collapse safety, structural performance, and functional continuity. Bulletin of Earthquake Engineering. https://doi.org/10.1007/s10518-026-02694-4

Image Credits: AI Generated

DOI: 10.1007/s10518-026-02694-4

Keywords: base isolation, friction pendulum systems, hospitals, seismic reliability, functional continuity, collapse safety, interstory drift, floor acceleration, life-cycle performance, triple friction pendulum, Mexico City, earthquake engineering