The research, led by Azlan Adnan of Universiti Teknologi Malaysia together with Yuyun Tajunnisa and colleagues at Institut Teknologi Sepuluh Nopember in Surabaya, Indonesia, combines full-scale laboratory testing with advanced computer simulation to trace the entire chain of seismic damage, from the first hairline crack in a joint to the collapse of an entire frame. The work was funded by the Indonesian Endowment Fund for Education under the EQUITY Program and supported by Malaysia’s Public Works Department, which supplied technical data for the connection details studied.

At the heart of the study is a deceptively simple structural element. In many precast industrial buildings, particularly in Southeast Asia, beams are not cast monolithically with columns. Instead, they are supported on nibs, short cantilevered projections that carry the beam’s reaction through a combination of bearing pressure, dowel action of any projecting reinforcement, and friction. This arrangement speeds up construction enormously, but it also creates a discontinuity in the load path precisely where seismic forces concentrate. When a frame sways during an earthquake, the beam-to-column interface must transfer alternating moments and shears through a joint that was never designed as a continuous monolithic element.

To understand how such a joint actually deforms, the team built a full-scale specimen of a nib connection and subjected it to cyclic loading, the laboratory equivalent of the back-and-forth shaking an earthquake delivers. The test followed internationally recognized protocols for cyclic seismic testing of components, applying progressively larger displacement cycles while measuring the force resisted by the joint. The resulting load-deformation loops, known as hysteresis curves, revealed how stiffness degrades and energy is dissipated as cracking, bond slip, and concrete crushing accumulate at the nib. These experimental curves are the gold standard for characterizing connection behavior, but they describe only a single joint, not a whole building.

That is where the hybrid modeling strategy of the study becomes significant. The researchers first constructed a detailed three-dimensional nonlinear finite element model of the tested joint in ABAQUS, using plastic-damage constitutive models for concrete that capture the material’s degradation under cyclic loading. By calibrating this micro-scale model against the full-scale test data, they ensured the simulation reproduced the real joint’s stiffness, strength, and hysteretic shape. From the validated model, they extracted a moment-rotation relationship for the connection, a curve describing how much rotational resistance the joint provides as it is progressively damaged. Because this curve was synthesized from both experiment and simulation, the authors describe it as a hybrid moment-rotation curve, and it forms the critical bridge between component-level physics and building-level analysis.

The hybrid curve was then imported into SAP2000, a structural analysis platform widely used in engineering practice, as a custom nonlinear hinge placed at each beam-to-column interface of a complete precast frame. This allowed the global model to reproduce the complex, pinching hysteretic behavior observed in the laboratory rather than assuming an idealized elastic or perfectly plastic joint. With the frame model in place, the team ran two complementary forms of nonlinear analysis. The first, nonlinear static pushover analysis, pushes the structure laterally with a monotonically increasing force pattern to identify the sequence of damage states and the governing collapse mechanism. The second, incremental dynamic analysis, subjects the frame to a suite of earthquake ground motion records scaled to ever-higher intensities, tracking how drift demands and damage accumulate as shaking intensifies.

The output of these analyses is a set of fragility curves, probabilistic functions that express the likelihood of reaching a defined damage state at a given level of ground shaking, here quantified by peak ground acceleration. Fragility curves are the workhorses of modern seismic risk assessment: they allow engineers, insurers, and code writers to translate a hazard forecast into an expected probability of damage. For the precast frame with nib connections resting on rock sites, the results are encouraging. The probability of first cracking, corresponding to the Immediate Occupancy performance level at which a building remains safe to use without significant repair, begins to rise at a peak ground acceleration of about 0.20 g. More strikingly, the probability of reaching the Collapse Prevention threshold, the point beyond which the structure is on the verge of partial or total collapse, remains a negligible one percent even at 0.30 g, a level of shaking that would be destructive in many regions.

Those favorable numbers, however, come with an important caveat that the authors emphasize through parametric variation. When the same frame is assumed to stand on softer soils, local site amplification magnifies the ground motion transmitted to the structure, and the damage probabilities climb substantially. The finding underscores that soil-structure interaction is not a secondary refinement but a first-order control on seismic risk for precast buildings. Two identical frames on the same street can face materially different fragilities depending on what lies beneath their foundations, a fact with direct consequences for site selection, foundation design, and retrofit prioritization in moderate-seismicity zones.

The study’s broader significance lies in its methodology as much as its results. In regions of low to moderate seismicity, such as much of Malaysia and Indonesia outside the great subduction zones, large destructive earthquakes are rare enough that extensive experimental databases do not exist, yet frequent enough that vulnerability assessment matters for schools, hospitals, and industrial facilities built with precast technology. By anchoring a numerical model to a full-scale test and then propagating that validated connection behavior through pushover and incremental dynamic analyses to fragility curves, the researchers have demonstrated a transferable framework. Other connection types, from dapped-end beams to grouted duct and post-tensioned joints, could be assessed with the same pipeline, filling a persistent gap between idealized code assumptions and the idiosyncratic behavior of real precast interfaces.

For the precast industry, the message is cautiously optimistic. Nib connections, when properly detailed and founded on competent ground, appear capable of delivering adequate seismic safety in moderate seismic zones, with cracking initiating only at shaking levels that would already alarm occupants and collapse probabilities remaining vanishingly small at 0.30 g on rock. But the same analysis shows that the margin evaporates as soil softens, and that the connection, not the member, is the component that governs the whole frame’s fate. As cities across Southeast Asia and beyond continue to expand with factory-built construction, studies of this kind provide the quantitative evidence base that codes and practitioners need to ensure that speed of assembly never comes at the cost of resilience when the ground finally moves.

Subject of Research: Seismic fragility assessment of precast concrete frames with nib beam-to-column connections

Article Title: Seismic performance and fragility analysis of precast concrete frames with nib connections

Article References: Adnan, A., Tajunnisa, Y., Ramli, M. Z., Falahi Abdul Halim, N. H., Ab Kadir, M. A., Nurhadi, H., Yudoprasetyo, K., & Hariyanto, I. R. (2026). Seismic performance and fragility analysis of precast concrete frames with nib connections. Bulletin of Earthquake Engineering. https://doi.org/10.1007/s10518-026-02700-9

Image Credits: AI Generated

DOI: 10.1007/s10518-026-02700-9

Keywords: precast concrete, nib connections, seismic fragility, cyclic testing, incremental dynamic analysis, pushover analysis, soil-structure interaction, finite element modeling, earthquake engineering, beam-to-column joints, moment-rotation curves, ABAQUS