Mexico City’s First Concrete Skyscrapers Still Stand Strong—Until the Soil Changes

Two reinforced concrete towers completed in 1950 on the campus of Mexico’s national university have quietly outlasted some of the most damaging earthquakes of the past century, including the magnitude 7.8 Guerrero event of 1957, the magnitude 8.1 Michoacan earthquake of 1985, and the magnitude 7.1 Puebla earthquake of 2017. A new study published in Built Heritage examines why these seventy-five-year-old structures, designed under an incipient seismic code that introduced earthquake accelerations only in 1942 with a coefficient of just 0.05 g, continue to function with only light damage. The answer, the researchers find, lies less in the sophistication of the original design than in the ground beneath the buildings—and the same analysis reveals a sobering counterfactual about what would happen if these frames stood on the city’s soft lakebed clays.

The two case studies, designated T15 and T10, were the tallest reinforced concrete high-rises in Mexico City when built and are now part of a university campus declared a UNESCO World Cultural Heritage Site in 2007. The T15 building rises fifteen storeys to 53.5 metres on rocky soil, with a rectangular plan of three transverse spans and thirteen longitudinal spans. The T10 building reaches ten storeys and 36.40 metres, with a single 5.40-metre transverse span and twelve longitudinal spans of 3.60 metres; its proportions make it slender under the Mexican seismic code, with a height-to-minor-base-width ratio greater than four. Both buildings embody the Modern Movement ideals of their era, with exposed concrete frames, multifunctional spaces, and façade masonry walls that are not connected to the primary structural system.

To assess their seismic behaviour, the researchers built three-dimensional finite element models in SAP2000, drawing structural configurations, material properties, and reinforcement details from the original design drawings held in archives, supplemented by in situ verification using LiDAR and ground penetrating radar. That survey revealed only minor deviations, chiefly column dimensions increased by five to ten centimetres because of additional cover. The concrete was modelled with a mean compressive strength of 20 MPa using Mander’s constitutive model, with a confinement ratio of 1.05 reflecting badly confined sections, and the reinforcement steel, with a yield strength of 420 MPa, followed the Park–Paulay relationship. Nonlinear behaviour was captured with fibre hinges distributed along frame elements using the Gauss–Lobatto integration method, together with a hysteretic pivot model that simulates stiffness degradation, reinforcement slip, and the pinching effect caused by cracks opening and closing.

The seismic input consisted of 52 pairs of bidirectional synthetic accelerograms from the Mexico City Seismic Code’s design action system, already adjusted to the site-specific design spectrum. Of these records, 16 correspond to stiff or rocky soil, 16 to transitional soil of alternating silty clays and stiff material, and 20 to soft lakebed clay, all based on a return period of 275 years associated with the collapse prevention limit state. For each record the team ran nonlinear response history analyses, extracting maximum roof displacements, interstorey drift ratios, ductility demands, and hysteretic energy, which was then converted into a global damage index through the methodology of Rodriguez and Padilla. This index compares the hysteretic energy dissipated per unit mass with the energy the same system would absorb in a linear-elastic response, mapping values onto performance levels from no damage to severe, irreparable harm.

On the rocky soil where the towers actually stand, the models predict nearly elastic behaviour. The T15 building shows mean damage indices of 0.23 longitudinally and 0.17 transversely, both below the 0.3 threshold for slight damage, with a global ductility close to 1.0. The shorter, stiffer T10 building performs even better, with mean damage indices of 0.048 and 0.080 and drifts of 0.20 to 0.40 percent, placing it within the immediate occupancy performance level. Elastic analyses under the design spectrum confirm that both buildings keep their interstorey drift ratios below the 0.015 life-safety threshold prescribed by the current Mexican seismic design code for low-ductility reinforced concrete frames. Pushover analyses show that in both structures the lateral capacity exceeds the seismic demand at the calculated performance points, although local weaknesses emerge: the T15 building concentrates maximum drifts between the third and sixth floors, while the T10 building develops a soft-storey mechanism on its seventh floor, where a reduction in column cross-section triggers an abrupt stiffness loss and collapse prevention is reached at a low global drift in the longitudinal direction.

The picture changes dramatically when the same buildings are analysed as if relocated onto transitional or soft soils. For the T15 building, mean damage indices on soft and transitional soils climb to 1.48 and 1.35 longitudinally and 0.95 and 0.88 transversely, with much higher dispersion. The T10 building reaches mean values of 0.93 and 0.88 longitudinally and, critically, 1.06 transversely on soft soil, with a standard deviation of 1.07 indicating that a substantial number of scenarios could push the response well beyond the severe damage threshold. The researchers identify what they call a drift–damage paradox: damage indices exceeding 1.0, signalling irreparable or total loss, occur at drift levels between 0.75 and 1.5 percent—values that remain compliant with the code’s life-safety drift limits. In other words, a building can satisfy code drift checks while accumulating damage that no repair programme could realistically address.

The physical basis for this vulnerability lies in the original design standards. Both towers were designed under 1940 specifications that permitted stirrup spacing up to three-quarters of the effective depth, with a maximum of 457 millimetres, producing unconfined concrete sections with inherently low ductility. A review against the current Mexican concrete design code found that while the T15 columns remain adequate, certain beams show slight noncompliance under modern demands, and excessive stirrup spacing in both buildings leaves the sections unable to dissipate energy effectively once they enter the inelastic range. Under soft-soil conditions, the study estimates these legacy frames would need global ductility capacities of roughly 2.5 to 3.0 to keep damage reparable—far beyond what their original detailing can deliver.

The numerical predictions align closely with what happened in reality. After the September 19, 2017 Puebla earthquake, whose ground acceleration at the rocky campus site exceeded that of previous subduction events, construction authorities conducted rapid post-earthquake inspections under CENAPRED and FEMA P-154 protocols. Both buildings were classified as Grade I minor damage, with structural integrity not immediately compromised, and operations resumed just 24 hours after the event. This observed Grade I damage corresponds well with the modelled damage indices of roughly 0.05 to 0.20 and unit ductility for stiff soil, lending the analytical framework an empirical anchor that many assessments of mid-century building stocks lack.

The findings carry weight well beyond two heritage towers, because the T15 and T10 buildings are representative of the reinforced concrete stock erected across Mexico City between 1950 and 1960. The study concludes that structural damage to similar mid-twentieth-century functional buildings is substantially underestimated when they rest on transitional or soft soils, and it argues that building codes should be reviewed for these systems, since drift-based limits alone prove non-conservative for legacy frames lacking ductile detailing. Even on their actual rocky foundation, some structural elements do not meet current code requirements, and the authors note that an assessment of whether reinforcement is appropriate is needed. Any intervention, however, faces a conservation dilemma: retrofitting could improve safety but risks altering the authenticity of UNESCO-protected structures, while repair strategies alone may prove insufficient under soft-soil conditions. The authors acknowledge limitations, including the fixed-base assumption, the exclusion of masonry façade walls that would likely add stiffness, and the absence of explicit fragility curves, and they call for future work on performance-based retrofitting strategies such as seismic isolation or energy dissipation devices tailored to heritage frames.

Subject of Research: Seismic performance assessment of the first reinforced concrete high-rise buildings in Mexico City using nonlinear finite element modelling and 52 bidirectional ground motion records.

Article Title: Seismic behaviour of the first modern high-rise buildings in Mexico City: lessons for future design

Article References: Santa Ana Lozada, P. R., & Peña Mondragón, F. (2026). Seismic behaviour of the first modern high-rise buildings in Mexico City: lessons for future design. Built Heritage, 10(1), Article 58. https://doi.org/10.1186/s43238-026-00292-6

Image Credits: AI Generated

DOI: 10.1186/s43238-026-00292-6

Keywords: Mexico City, seismic engineering, reinforced concrete, heritage buildings, damage index, soil-structure interaction, UNESCO World Heritage, interstorey drift, nonlinear analysis, earthquake resilience, building codes, soft soil