The motivation behind the research is both environmental and practical. The construction industry accounts for a substantial share of global resource depletion, while sustainably managed regrowing materials such as timber offer an abundant, self-regenerating resource. Engineered wood products like glued laminated timber have already enabled remarkable mass-timber buildings, including the Centre Pompidou-Metz and the Multihalle gridshell in Mannheim. But double-curved structural timber elements remain notoriously resource-intensive: they typically require extensive molding, gluing, and machining, or generate substantial waste through subtractive milling. Elastic bending offers an alternative form-giving strategy, in which initially straight planks are deformed into curved configurations while remaining strain-free along their length.

The geometric foundation of the approach lies in differential geometry, specifically in geodesic curves. A geodesic is the shortest and straightest path between two points on a surface, characterized by vanishing geodesic curvature. Unlike asymptotic curves, which only exist on surfaces with negative Gaussian curvature, geodesics can be generated on any surface, giving designers considerable flexibility. Crucially, planks aligned with geodesic curves remain tangent to the design surface, allowing them to be stacked efficiently in superposed layers. This property helps overcome the structural limitations that normally constrain bending-active structures, where material strength and bending curvature limit the cross-section of individual elements and make large-scale systems vulnerable to snap-through buckling.

Generating the geodesic curve network for the column was itself a computational challenge. Because the researchers wanted the lamellae to fully wrap around the column, with start and end points positioned above one another on the same generatrix, neither a boundary value nor an initial value formulation was suitable. Instead, the team used constrained relaxation: control points sampled along the generatrix were rotated around the central axis to form a polyline, which was then relaxed on the design surface using Kangaroo2, a live physics engine for form-finding. The process can be pictured as wrapping a rope around the surface and tightening it by pulling at each end. The resulting helix-like curves were verified by measuring their geodesic curvature, which remained well below ten percent of the normal curvature, an acceptable deviation from a true geodesic.

Materializing the geodesic network as physical planks required the strips to be developable, meaning they can be unrolled into flat sheets without stretching. Developable surfaces have zero Gaussian curvature, and the ruling directions of a developable strip along a spatial curve must be chosen as the vector conjugate to the curve tangent, not simply aligned with the Darboux frame as is common for planar curves. The conjugate vector, defined through the second fundamental form of the surface, corresponds to the intersection line of two neighboring tangent planes. Because the rulings are straight lines lying in the tangent plane, determining their lengths reduces to a straightforward geometric calculation that directly yields the plank width.

One subtle geometric singularity demanded special treatment. At a point in the upper region of the column, the curvature graph of the geodesic suddenly changes direction, and the curve normal rotates 180 degrees around the tangent vector. This can only occur where the normal curvature passes through zero, meaning the geodesic temporarily aligns with an asymptotic direction. At that point the strip undergoes pure twist rather than bending, a non-isometric deformation that does not yield a developable surface. The team tolerated a limited deviation from true developability, computing the resulting strain to be below one percent and assuming the material could accommodate the corresponding stresses. Notably, this non-developable region propagates further for wider planks, covering roughly 71 centimeters along the curve for the 8-centimeter-wide planks used, less than ten percent of the total area.

The structural innovation at the heart of the project is guided bending. An elastically bent element naturally deforms into a state of minimal elastic energy, known as the elastica shape. Forcing it to follow a different path introduces residual stresses and spring-back effects that cause geometric deviations from the intended design. To steer the deformation instead, the researchers introduced a series of incisions aligned with the conjugate vector, extending halfway through the lamella thickness and spaced 4 centimeters apart along the centerline. Acting as controlled hinges, these incisions direct the local bending at each point, guiding the overall deformation of the lamellae toward the target shape. Because the incisions were placed in the compression zone, simple experiments confirmed that lamellae achieved much smaller bending radii than when incisions were on the tension side.

The column itself was assembled from five continuous ash lamellae, 80 millimeters wide and 8 millimeters thick, arranged in six axes in both directions and bolted together at their intersections, following construction principles developed by Julius Natterer. Shorter infill lamellae filled the gaps created by the weave-like stacking, and the entire assembly was mechanically laminated with self-drilling timber screws, resulting in a compound depth of 72 millimeters comprising five continuous and four infill layers. The straight-development property of the geodesic planks simplified fabrication dramatically: bolt hole positions and incision angles could be mapped directly onto straight lines, allowing the entire production to be carried out with a standard three-axis CNC mill. During erection, a prefabricated reference scaffolding with predefined notches fixed each lamella endpoint precisely in space, and the column’s shape emerged as successive layers were installed and locked into their final geometry.

Finite element analysis, validated against the full-scale testing approach of Lara-Bocanegra and colleagues for elastically bent timber gridshells, revealed the column’s remarkable structural behavior. Owing to its proportions, with a height-to-width ratio of approximately five, the column acts structurally as a shell rather than a slender clamped column, so bending moments remain close to zero and the lamellae are verified essentially for normal forces. The asymmetric wind load case proved decisive, generating a rotational moment at the column head resisted by compression and tension forces within the column, with the inner layers experiencing the highest loads. The utilization of the inner lamella layer reached a maximum local value of 92 percent. Including the infill layers raised the buckling capacity from 113 to 135 percent of the design load in one configuration, confirming that their contribution is concentrated in compound stability rather than lamella verification, and was retained as a structural reserve rather than relied upon in design.

The project demonstrates that guided bending can significantly broaden the feasible design space of elastically bent compound beam assemblies, establishing a direct link between geometric design, fabrication logic, and structural performance. Unmodified lamellae of identical thickness could not have been bent around the column without structural failure, and although thinner lamellae could theoretically have achieved the target geometry, thicker lamellae proved structurally advantageous because fastener size scales with material thickness. The researchers acknowledge that only the orientation of the incisions was theoretically derived, while their width, depth, and spacing were kept constant, and that the precise effect of the incisions on load-bearing capacity remains to be quantified. Experimental validation of the built assembly also remains outstanding. Still, with nearly ten times more infill pieces than continuous lamellae produced, ongoing research aims to reduce or eliminate their necessity, improve fabrication efficiency, and calibrate incision patterns, paving the way for structurally efficient, adhesive-free curved timber architecture.

Subject of Research: Elastic bending and geodesic geometry for adhesive-free curved timber structures

Article Title: A guided bending strategy for the geometric and structural design of an elastically bent geodesic timber column

Article References: Andersson Largueche, D., & La Magna, R. (2026). A guided bending strategy for the geometric and structural design of an elastically bent geodesic timber column. Frontiers in Built Environment, 12, Article 1942204. https://doi.org/10.3389/fbuil.2026.1942204

Image Credits: AI Generated

DOI: 10.3389/fbuil.2026.1942204

Keywords: timber construction, guided bending, geodesic curves, elastic deformation, structural design, computational geometry, guided, bending, strategy, geometric, structural, design

The core problem the researchers identify is that current engineering practice generally relies on stochastic approaches for wind extremes and conventional extreme-value analyses for wave extremes. Because wind and wave assessments are carried out with different methodologies, the resulting design conditions can be physically inconsistent, and in some cases non-physical, meaning the paired wind and wave states used for structural design could not actually occur together in the real ocean and atmosphere. For a turbine tower that must simultaneously resist hurricane-force gusts and the slamming of enormous waves, such mismatches matter enormously.

A second challenge is observational. Tropical cyclones are rare events at any given location, and the relatively limited duration of observational records compared with the occurrence rate of these storms introduces substantial uncertainties when engineers try to estimate extremes with long return periods, such as the 50-year or 100-year conditions that anchor offshore design standards. A few decades of buoy measurements simply cannot capture the full statistical behavior of a hazard that may visit a site only a handful of times per century.

To overcome these limitations, the team turned to a Monte Carlo framework, a computational strategy that generates vast numbers of synthetic scenarios to build statistics that no observational record could provide. The methodology combines synthetic tropical cyclone tracks from the STORM database, developed by Bloemendaal and colleagues, with a parametric wind model from Vinour and colleagues and a tropical-cyclone-adapted configuration of the WAVEWATCH III wave model. Together these components generate large ensembles of tropical-cyclone-induced wind and wave conditions, effectively simulating thousands of storms that never happened but behave according to the physics of those that did.

The framework also incorporates two components that conventional approaches often neglect: the contribution of swells and of extratropical cyclones. Swells are long-period waves generated by distant storms that can arrive at a site independently of local weather, and they can coexist with a passing tropical cyclone, adding to the local wave field. The researchers introduce a statistical approach to account for these remotely generated swell systems, ensuring that the wave conditions paired with cyclone winds reflect what actually reaches the site.

Extratropical cyclones, the powerful mid-latitude storm systems that dominate weather in temperate zones, are handled through annual maxima statistics derived from the ERA5 reanalysis, a comprehensive global atmospheric and oceanic dataset produced by blending observations with numerical weather modeling. By folding these statistics into the same framework, the method produces joint wind and wave extremes that represent the full spectrum of severe storms a site may experience, not just the tropical variety.

The approach was applied at the NDBC 41002 buoy in the western North Atlantic, a location where long-term wind and wave observations are available for comparison. This validation step is critical: a Monte Carlo method is only as trustworthy as its agreement with reality. The results show that the coupled modeling framework agrees reasonably well with the historical tropical cyclone extreme statistics recorded at the buoy, giving the authors confidence that the synthetic storm ensembles reproduce the behavior of real events.

The payoff for the offshore wind industry is that the methodology provides physically consistent joint wind and wave conditions that can be directly used to construct environmental contours. Environmental contours are curves in wind-wave space that enclose combinations of conditions with a given probability of exceedance, and they are a standard tool for identifying offshore design points. Because the contours generated by this method pair winds with waves that genuinely co-occur, engineers can select design load cases that are both conservative and physically plausible, avoiding the non-physical combinations that can emerge when wind and wave extremes are estimated separately.

For regions eyeing offshore wind as a climate solution, from the Gulf of Mexico to the Caribbean and the western Pacific, the implications are significant. Turbines deployed in cyclone-prone waters must survive conditions far beyond those of the North Sea, and underestimating the joint severity of wind and waves could compromise structural integrity, while overestimating it inflates costs. A framework that generates thousands of physically consistent storm scenarios from synthetic tracks, adapted wave modeling, swell statistics, and reanalysis data offers a way to anchor those design decisions in robust statistics rather than sparse observations.

The study, authored by researchers at France Energies Marines and Ifremer in Plouzané, France, is distributed under a Creative Commons Attribution 4.0 license and is currently open for community discussion as part of the peer-review process at Wind Energy Science. As offshore wind farms push into hurricane territory, methods like this one, which treat wind and waves as the coupled physical system they truly are, may become essential tools for building clean energy infrastructure that can weather the storm.

Subject of Research: A Monte Carlo methodology for assessing joint extreme wind and wave statistics at offshore wind farms exposed to tropical and extratropical cyclones

Article Title: A method to assess wind and wave extreme statistics at offshore wind farms exposed to tropical and extratropical cyclones

Article References: Filipot, J.-F., Renaud, P., Prevosto, M., Leckler, F., Pouplin, C., & Raillard, N. (2026). A method to assess wind and wave extreme statistics at offshore wind farms exposed to tropical and extratropical cyclones. https://doi.org/10.5194/wes-2026-162

Image Credits: AI Generated

DOI: 10.5194/wes-2026-162

Keywords: offshore wind energy, tropical cyclones, extratropical cyclones, extreme value statistics, Monte Carlo simulation, WAVEWATCH III, environmental contours, wave modeling, swell, ERA5 reanalysis, structural design, Wind Energy Science

The stakes are enormous. In 2024, the United States experienced 27 separate billion-dollar weather and climate disasters, and tornadoes and thunderstorms accounted for more than half of them. Although violent tornadoes rated EF3 and above represent a small fraction of all occurrences, they dominate economic losses, infrastructure disruption and threats to life. Wind engineers, unlike meteorologists issuing real-time warnings, need long-term probabilistic estimates of peak wind speeds to inform design standards for buildings and infrastructure, and the broad EF categories obscure exactly the information they need most.

The core innovation of the study lies in what it predicts. Rather than assigning tornadoes to EF categories, the researchers model the probability that a tornado’s maximum damage wind speed (DWS) exceeds specific thresholds: 136, 140, 145, 150, 155, 160 and 165 miles per hour, all within the EF3 range of 136 to 165 mph. This range matters because it is a transition zone. Most structures built to current codes are expected to withstand EF2-and-below winds of 135 mph or lower, while EF4 and EF5 events above 166 mph often produce near-total destruction. The EF3 band, in which tornadoes cause a broad spectrum of severe but not necessarily complete structural damage, is where prediction offers the greatest engineering value, and EF3 tornadoes occur far more frequently than their stronger counterparts, providing more data for model development.

Building a trustworthy dataset was the study’s crucial first step. The team quality-controlled and cross-validated records from three federal sources: the Storm Events Database maintained by the National Centers for Environmental Information, the Damage Assessment Toolkit (DAT), and the Storm Prediction Center (SPC) tornado database. The DAT, which provides the most comprehensive post-event damage information, served as the primary source. Records with missing, inconsistent or duplicated values were flagged and investigated rather than simply deleted, and corrections were made only when supported by at least two independent sources, including National Weather Service and NOAA field survey reports. The result is a curated historical dataset of more than 630 authenticated EF3 through EF5 tornadoes in the United States from 2007, when the EF scale was introduced, through 2025.

With the data curated, the researchers turned to a Random Forest (RF) classifier, an ensemble method that constructs many decision trees and aggregates their predictions to improve accuracy and reduce overfitting. The model uses just four geometric features of the tornado path: track length, maximum width, and the sine and cosine components of the path bearing angle. The authors deliberately minimized the feature set because earlier studies showed that most variables in tornado databases contribute little predictive power, and because correlated variables such as duration, which tracks closely with path length, introduce redundancy. The bearing angle, a circular variable where 0 and 360 degrees mean the same direction, was decomposed into sine and cosine components to avoid discontinuities that would confuse the algorithm.

Model development centered on Iowa, home to the National Testing Facility for Enhancing Wind Resiliency of Infrastructure in Tornado-Downburst-Gust Front Events (NEWRITE) project and part of the midwestern tornado alley. Because violent tornadoes are rare, only 10 EF5 tornadoes occurred nationwide between 2007 and 2025, the team had to confront severe class imbalance. They enriched the training data with above-threshold tornadoes from across the country and applied the Synthetic Minority Over-sampling Technique (SMOTE) where necessary. Because this rebalancing inflates the apparent probability of extreme events, the researchers calibrated the raw RF outputs using logistic regression, following a method proposed by Dankowski and Ziegler, which produced closer agreement with observed Iowa exceedance rates than traditional Bayes prior correction.

The results are striking. Across all seven DWS thresholds, the Iowa RF models and their independent validation datasets consistently achieved accuracy and weighted F1 scores above 85 percent, with recall scores for both above-threshold and below-threshold tornadoes generally exceeding 85 percent. Permutation feature importance analysis revealed that tornado path width was by far the most influential predictor, followed by path length, with the directional variables ranking third and fourth. This makes physical sense: stronger tornadoes carve wider damage paths, and the relationship proved more direct than the link between intensity and path length, since some EF3 tornadoes leave shorter tracks than weaker storms. Comparisons with logistic regression and XGBoost baseline models showed broadly comparable overall performance, but the RF models maintained consistently higher recall for the above-threshold class, the metric that matters most for hazard assessment, where missing a dangerous tornado is far worse than a false alarm.

Perhaps the most consequential finding concerns geography. When the Iowa 136 mph model was applied to other states, performance varied substantially. Neighboring states such as Illinois, Missouri, Nebraska, Wisconsin and Minnesota performed comparably to Iowa, with Illinois reaching 89 percent accuracy, but Kansas, the Dakotas and Florida showed lower recall for above-threshold tornadoes, while Mississippi and Tennessee caught dangerous tornadoes well but misclassified weaker ones. In Kansas, path length even showed slightly negative feature importance, apparently because several EF3 tornadoes there had unusually short tracks. State-specific models developed for Missouri, Mississippi, Tennessee and Texas generally achieved accuracy between 75 and 95 percent, suggesting that tornado geometry-damage relationships are not uniform across the country and that state-specific or regional models may capture local variability better than any universal model.

The framework’s engineering implications extend well beyond classification. Structural engineers use fragility functions, conditional probabilities that a given building type exceeds a specific degree of damage at a given wind speed, to assess tornado risk. By combining the RF model’s DWS exceedance probabilities with state-specific fragility functions, the approach could estimate the overall probability that a residential building suffers severe or life-threatening damage in a tornado, moving the field toward performance-based design. The authors note that previous engineering-based damage assessment tools require extensive calculations for each location, whereas the machine learning framework embodies multiple variables at once and can be updated as new event data arrive each year.

The study also candidly acknowledges its limits. Only 10 EF5 tornadoes occurred in the study period, so the rarest and most destructive events remain poorly represented, and historical damage records carry uncertainties from survey practices, varying state building codes and evolving rating methodologies. The authors caution that model predictions should be interpreted as data-driven estimates rather than direct representations of tornado wind fields, especially since surface wind speeds during tornadoes remain nearly impossible to measure directly. Still, by demonstrating that a tornado’s path geometry, above all its width, carries quantifiable information about its destructive potential, the research opens a practical path toward probabilistic tornado hazard maps, fragility analysis and, ultimately, more resilient building design in the communities that need it most.

Subject of Research: Machine learning prediction of tornado damage wind speed exceedance from tornado path geometry

Article Title: Tornado damage prediction using machine learning

Article References: Li, Z., Cristea, N., Sarkar, P., & Reed, D. (2026). Tornado damage prediction using machine learning. Frontiers in Built Environment, 12, Article 1971791. https://doi.org/10.3389/fbuil.2026.1971791

Image Credits: AI Generated

DOI: 10.3389/fbuil.2026.1971791

Keywords: tornado, machine learning, random forest, Enhanced Fujita scale, damage wind speed, wind engineering, path width, Iowa, hazard assessment, fragility functions, Storm Prediction Center, structural design

The method at the center of the debate is the material intensity approach, the workhorse of bottom-up building material stock analysis. It works by multiplying a building’s gross floor area by an aggregated coefficient, expressed in kilograms or cubic meters of material per square meter of floor space. That simplicity is precisely why it dominates the field: it requires no component-level data, no floor plans, and no knowledge of internal layouts, making it feasible to estimate material stocks for entire cities, nations, or the global building stock. Those estimates matter enormously for the circular economy, because knowing what materials sit where in the built environment is the prerequisite for urban mining, the strategic recovery of steel, concrete, and timber from buildings at the end of their lives.

The problem, as Yingqi Jia and Chen Feng point out, is that the material intensity method treats all buildings of a given type as essentially interchangeable. Yet real buildings of the same class, built in the same era, can differ dramatically in footprint shape, corridor arrangement, number of units, room layouts, number of floors, and structural system configuration. Quantifying how much these design choices matter has long been blocked by a practical obstacle: detailed internal layout data for real buildings is scarce, and no city keeps a comprehensive database of floor plans suitable for this kind of controlled comparison. The researchers’ solution was to sidestep real-world data entirely and generate their own.

Their pipeline is a hybrid of two computational approaches, each compensating for the other’s weaknesses. Parametric modeling offers precise, systematic control over design parameters, which is exactly what a controlled experiment requires, but achieving architectural realism with parameters alone demands an impractically complex rule set. Deep learning generative models, by contrast, can produce realistic floorplans with minimal manual effort, but they do not allow explicit, systematic control of the variables under study. The team’s workflow combines both: parametric routines generate footprints, insert corridors, and subdivide space into units using a corridor-aware binary space partitioning algorithm, while a neural network called Graph2Plan, pre-trained on real-world floorplans, generates plausible room layouts within each unit. The pipeline then assembles three-dimensional structural components, resolves intersecting volumes with Boolean operations to avoid double counting, and calculates the material volume of every wall, column, beam, and slab.

Running this pipeline across the full parameter space produced 48,600 synthetic residential buildings, all with a constant gross floor area of 2,400 square meters, while systematically varying footprint shape among square, rectangle, L-shape, T-shape, and H-shape geometries; corridor type among single-loaded, double-loaded, and point-access systems; unit counts; room layouts; floor counts from one to six; and structural grid spacing and member sizing. The headline result is striking: total material stock ranged from 758 to 2,888 cubic meters across this design space, with a mean of 1,548 cubic meters and a standard deviation of 320. Under the traditional material intensity method, every one of those buildings would have been assigned the identical stock value, because floor area alone determines the estimate.

Disentangling which design parameters drive this variability required statistical analysis rather than inspection of individual cases. A multiple regression model explained more than 95 percent of the variance in material stock and revealed a clear hierarchy of influence. The number of floors was the single most powerful factor, alone accounting for roughly one-third of the variation, followed by the sizing of the structural frame, then the number of units and the footprint shape. Grid spacing and floor height came next, while corridor type and room layout had much smaller, though still statistically significant, effects. Every parameter in the model was significant at the p < 0.001 level.

The physical mechanisms behind these rankings are intuitive once spelled out. Holding floor area constant, the number of floors controls a building’s slenderness, and every additional floor must be enclosed by walls. In the synthetic dataset, wall volume grew from 600 cubic meters in single-story buildings to 1,186 cubic meters in six-story buildings, driving a 55.2 percent increase in total stock, while the structural frame grew far more modestly. Structural sizing mattered even more sharply within the frame itself: moving from slim to bulky member dimensions raised total material volume by 35.1 percent, and widening the structural grid from 4.0 to 8.0 meters cut it by 10.5 percent by reducing the number of columns and beams needed. Footprint shape operated through the surface-to-volume ratio, with the sprawling H-shape demanding 12.1 percent more material than the compact square, almost entirely in wall components, since frame volumes stayed nearly constant across shapes. Adding units raised stock sublinearly, because each new partition needs proportionally less wall, and going from one to three bedrooms per unit added only 5.4 percent.

Because the dataset is synthetic, the authors took plausibility seriously. They reproduced wall and slab volumes from an open residential building information model within 4 percent, checked that the 2,400 square meter floor area falls within real-world ranges observed in OpenStreetMap records across 17 cities on six continents, and compared converted mass-based intensities against the RASMI global benchmark database, producing a median of 1,276.77 kilograms per square meter against a benchmark of 1,285.52. They also examined structural calculation records for two real Hong Kong apartment towers, Tin Sam Villa and Millennium Court, finding derived column-and-beam volumes of 225.1 and 226.2 cubic meters, both close to the synthetic mean of 222.2 and well inside the synthetic range. Supplementary tests at 1,200 and 4,800 square meters confirmed the findings are robust to the choice of fixed floor area.

The practical implications cut in two directions. For analysts, the results suggest that material intensity databases should expand their building typologies beyond use type and construction period to include footprint shape, number of floors, and structural configuration, the parameters that dominate design-induced variability. Where such classification is not feasible, the pipeline can generate exploratory adjustment factors, nudging baseline intensities upward for slender, complex buildings and downward for compact, low-rise ones. For designers, the study offers a menu of material-efficiency levers: reduce slenderness, widen structural grids, slim down frame members, minimize unit divisions and floor height, favor square or rectangular footprints, and use point-access cores instead of corridors. The authors caution that these levers must be weighed against structural, regulatory, economic, and functional requirements they did not model, and that their analysis was limited to reinforced concrete frames, excluded foundations, roofs, and mechanical systems, and explored a theoretical design space rather than the distribution of real buildings, so the reported range represents potential variability rather than expected real-world spread.

What may prove most durable about the work is the methodological framework itself. By fusing parametric modeling, generative AI, and component-based material accounting into a fully automated pipeline, the researchers have created an extensible tool for running what-if experiments on building design and resource demand at a scale impossible with real-world data. The complete workflow, the database of more than 48,000 building configurations, and tutorials for reproducing the results are openly available on GitHub. As cities race to turn their building stocks into material banks for a circular economy, the study delivers a clear warning and a clear opportunity: floor area alone cannot tell you what a building is made of, but the design decisions that shape it can now be measured, ranked, and, ultimately, engineered for less.

Subject of Research: Sensitivity of building material stock estimates to architectural design parameters

Article Title: Assessing the sensitivity of material-intensity-based building stock estimates to design parameters

Article References: Jia, Y., & Feng, C. (2026). Assessing the sensitivity of material-intensity-based building stock estimates to design parameters. Journal of Industrial Ecology, 30(4), 2035-2050. https://doi.org/10.1007/s44498-026-00138-5

Image Credits: AI Generated

DOI: 10.1007/s44498-026-00138-5

Keywords: building material stock, material intensity, circular economy, urban mining, parametric modeling, deep learning, floorplan generation, structural design, reinforced concrete, industrial ecology, synthetic data, material efficiency

The research, conducted by Hessam Atghaee and Farhang Farahbod of Islamic Azad University’s West Tehran Branch together with Fayaz R. Rofooei of Sharif University of Technology, was published in the Bulletin of Earthquake Engineering. The team set out to answer a deceptively simple question: what happens to the in-plane seismic behavior of a concrete-filled steel deck diaphragm when it is simultaneously squeezed by out-of-plane gravity loading? Despite decades of use in commercial and industrial construction around the world, the coupled influence of sustained gravity loads and connection detailing on diaphragm response has remained insufficiently quantified, leaving a gap between how these systems are tested in laboratories and how they actually perform in buildings.

To close that gap, the researchers built and tested three full-scale diaphragm specimens, each measuring 2.5 by 5.0 meters, under reversed cyclic in-plane loading that mimics the back-and-forth shearing an earthquake imposes. The first specimen served as a control, carrying no gravity load and containing no mechanical connectors between the steel deck and the concrete. The second was subjected to a sustained gravity load of 600 kilograms per square meter, a realistic service-level loading, but likewise lacked mechanical connectors. The third combined the same gravity loading with self-drilling screws installed to enforce composite action between the steel deck and the concrete fill. This three-way comparison allowed the team to isolate the effect of gravity loading alone and then to evaluate whether mechanical fasteners could compensate for the degradation it causes.

The experimental program captured a rich set of response measurements, including hysteretic load-displacement loops, stiffness degradation from cycle to cycle, slip at the steel-concrete interface, and the progression of visible damage through to failure. The results were striking. Sustained gravity loading reduced the in-plane diaphragm capacity by approximately 20 percent compared with the unloaded control specimen. Just as importantly, gravity loading accelerated the development of damage under cyclic loading, meaning the diaphragm not only lost strength but also deteriorated faster as the simulated earthquake continued. In seismic terms, that combination is doubly concerning: a weaker diaphragm that degrades more quickly offers less protection precisely during the strongest and longest phases of ground shaking.

The third specimen told a more encouraging story. Adding self-drilling screws to enforce composite action between the deck and the concrete effectively suppressed interface slip, the relative sliding that undermines the ability of the two materials to work as a single structural unit. With slip controlled, the fastened diaphragm exhibited significantly reduced in-plane deformations and enhanced post-elastic stiffness, retaining its shape and load path integrity at displacement levels where the unfastened specimens had softened considerably. Interestingly, however, the mechanical fasteners did not raise the ultimate in-plane capacity. Their benefit lies in deformation control and stiffness preservation rather than in raw strength, a distinction that matters for performance-based seismic design, where limiting drift and damage is often as important as preventing collapse.

Complementing the laboratory work, the researchers developed a detailed three-dimensional finite element model of the diaphragms and validated it against the experimental measurements. With the model proven reliable, they used monotonic numerical analyses to reproduce the seismic backbone, or envelope, response of the specimens and to run parametric investigations that would have been impractical experimentally. One parameter stood out: reducing the concrete compressive strength decreased both diaphragm stiffness and strength, though it did not alter the underlying failure modes. In other words, weaker concrete produces a weaker, more flexible diaphragm, but the same mechanisms govern its failure. By contrast, the orientation of the applied in-plane loading relative to the deck corrugations had only a minor influence on seismic diaphragm response, a finding that simplifies design considerations for decks loaded in different directions.

The implications ripple through seismic design practice. Modern codes and design guides, including documents from the American Concrete Institute, the American Institute of Steel Construction, and the Steel Deck Institute, provide procedures for diaphragm design, but the new results highlight the importance of explicitly considering sustained gravity effects rather than treating in-plane and out-of-plane demands as independent. A diaphragm designed without accounting for gravity-induced degradation could enter an earthquake with substantially less reserve capacity than its calculations suggest. The 20 percent capacity reduction measured here is not a marginal correction; it is the kind of systematic bias that can erode safety margins across an entire building stock, particularly in regions where composite steel deck floors dominate mid-rise construction.

The study also carries a practical prescription. Because self-drilling screws suppressed interface slip and dramatically reduced in-plane deformations without adding ultimate capacity, they represent a relatively inexpensive retrofit and detailing measure that improves seismic performance where it often matters most: controlling damage and maintaining functional stiffness during shaking. For new construction, specifying mechanical fasteners to enforce composite action offers a straightforward way to build in that resilience. For existing buildings, the findings give engineers a quantitative basis for assessing whether unreinforced diaphragms under significant gravity load warrant strengthening, especially in seismic retrofit projects evaluated under standards such as ASCE 41.

Methodologically, the work demonstrates the value of pairing full-scale cyclic testing with validated finite element simulation. Physical tests of 2.5 by 5.0 meter diaphragms under simultaneous gravity and cyclic in-plane loading are expensive and logistically demanding, and only a handful of configurations can realistically be examined. The validated numerical model extends the reach of the three experiments, allowing the team to explore how material properties and loading orientations shape the backbone response across a much broader design space. The researchers note that the experimental measurements and simulation results are available from the corresponding author upon reasonable request, given the large volume of data and the project-specific nature of the models.

Ultimately, the study delivers new quantitative insight into gravity-influenced diaphragm behavior at a moment when earthquake engineering is increasingly focused on realistic, coupled loading conditions rather than idealized single-demand tests. Buildings do not experience earthquakes in isolation from the loads they carry every day, and the structures that connect those loads to the lateral system should not be designed as if they did. By showing exactly how much capacity sustained gravity loading can consume, and by demonstrating a simple fastening strategy that claw back much of the lost performance, the researchers have given the seismic engineering community both a warning and a workable solution. For the countless composite steel deck floors overhead in offices, hospitals, and warehouses worldwide, the message is clear: the weight they carry shapes how well they can protect the people below when the ground begins to move.

Subject of Research: Experimental and numerical investigation of gravity load effects on the in-plane seismic behavior of composite steel deck diaphragms.

Article Title: Investigation of gravity load effects on in-plane behavior of composite steel decks

Article References: Atghaee, H., Farahbod, F., & Rofooei, F. R. (2026). Investigation of gravity load effects on in-plane behavior of composite steel decks. Bulletin of Earthquake Engineering. https://doi.org/10.1007/s10518-026-02692-6

Image Credits: AI Generated

DOI: 10.1007/s10518-026-02692-6

Keywords: composite steel deck, seismic diaphragm behavior, gravity load effects, cyclic loading, self-drilling screws, interface slip, finite element modeling, stiffness degradation, in-plane shear, earthquake engineering, concrete-filled steel deck, structural design

The research, published in the International Journal of Intelligent Robotics and Applications, focuses on a cable-driven hyper-redundant robot, a class of machines defined by a very high number of degrees of freedom. Hyper-redundant manipulators are prized for their manoeuvrability in confined and challenging environments, from inspection ducts to maintenance scenarios where conventional rigid arms cannot reach. But that flexibility comes at a price. Because each joint can move independently, the space of feasible postures is vast, and each posture produces a different distribution of internal loads. The mapping from configuration to peak stress is highly nonlinear and can only be evaluated reliably through computationally expensive simulations, making an exhaustive search impractical even for a five-dimensional problem.

The core of the new workflow treats worst-case stress identification as a black-box maximisation problem over the robot’s configuration space. For any candidate pose, a recursive statics model first computes the forces and moments that the rest of the manipulator transmits to the component of interest, in this case the robot’s first link. Cable tensions, gravity and the geometry of the cable routing all enter this equilibrium calculation, which is solved sequentially from the distal end of the arm toward the base. The resulting interface reactions then serve as boundary conditions for a finite element analysis of the link, performed in ANSYS, which returns the peak Von Mises stress, the standard scalar measure engineers use to gauge how close a material is to failure.

Running that pipeline for every conceivable pose would be prohibitively slow, so the researchers built a surrogate model that learns the relationship between configuration and stress. The surrogate is a Gaussian process, a probabilistic model that not only predicts the stress at any pose but also quantifies how confident it is in that prediction. The team used an ARD Matérn 5/2 kernel, a mathematical structure whose per-dimension length scales reveal which joint variables matter most to the stress response, and trained it on an initial set of 200 poses generated by Latin hypercube sampling across the joint limits of minus 45 to plus 45 degrees. All inputs and outputs were z-score standardised before training to keep the numerics well conditioned.

Bayesian optimisation then took over. The algorithm repeatedly asks where the next finite element simulation should be spent, using an acquisition function called Expected Improvement that balances two competing urges: exploiting regions where the model already predicts high stress, and exploring regions where its uncertainty is large, since a hidden danger may lurk there. Rather than evaluating one pose at a time, the framework uses batch Expected Improvement, proposing ten configurations per iteration that are evaluated in parallel. Because the batch criterion accounts for correlations between the proposed points, it penalises redundant, near-identical samples and rewards diverse batches that jointly cover both promising and uncertain regions of the configuration space.

The optimisation converged after just six Bayesian iterations, corresponding to 250 evaluated configurations in total. The identified worst-case pose produced a peak stress of 14.23 megapascals, an improvement of 1.41 percent over the best configuration found in the initial random design. That modest gain is itself informative: the space-filling initial design had already sampled near the critical region, and the Bayesian loop refined and confirmed the most demanding neighbourhood rather than drifting elsewhere. The Expected Improvement values attached to proposed batches fell from 16.2 percent of the normalised output range at the first iteration to 0.21 percent at convergence, indicating genuine saturation rather than a premature stop.

Two further checks strengthened confidence that nothing dangerous had been missed. When the researchers optimised the final Gaussian process surrogate itself and validated its predicted optimum in ANSYS, the resulting configuration, although geometrically distinct from the observed worst case, produced approximately 14.2 megapascals, nearly identical to the maximum found during the loop. This suggests the stress-critical condition is not an isolated spike but a neighbourhood of similarly punishing poses, exactly the kind of insight engineers need when planning real trajectories. On an independent test set of 100 previously unseen configurations, the surrogate achieved a mean absolute error of 0.13 megapascals and a coefficient of determination of 0.992, and in the top decile of the stress distribution its uncertainty estimates proved slightly conservative, with no false negatives at either the 90th or 95th percentile danger thresholds.

The authors are careful to frame the case study as methodological rather than as a certification of the hardware. The first link they analysed is intentionally overdesigned, so the absolute stress values remain far below the 45 megapascal yield strength of the PA12 polymer from which the link is made. The load cases considered arise from self-weight, cable preload and configuration-dependent internal load transfer, with no external payload applied at the end effector, and the analysis is quasi-static, omitting inertial effects and distributed cable friction. Within those assumptions, however, the framework demonstrates something broadly useful: it can tell designers whether substantially higher stresses are likely to remain hidden in unexplored regions of the configuration space, and it can do so with a limited simulation budget.

The practical implications extend beyond design verification. Because the surrogate is trained directly on robot configurations, the same workflow can be used during operation to screen trajectories, flagging or excluding poses associated with elevated structural demand before the robot ever attempts them. This stress-aware trajectory screening could be particularly valuable for hyper-redundant and snake-like robots deployed in inspection and maintenance, where the machine’s posture changes continuously and a single unlucky configuration could compromise the mission. The authors also note that the approach could be extended to multiple links, alternative structural metrics such as fatigue, or constrained multi-objective formulations, and that nonlinear material regimes would require discontinuity-aware kernels or additional regularisation.

What makes the work resonate beyond robotics is its demonstration of a broader principle: when every evaluation is expensive, intelligence about where to look next is worth more than raw computational power. Bayesian optimisation with Gaussian process surrogates has already transformed topology optimisation, aerospace design and co-design of robot mechanics and behaviour, and this study extends that logic into the configuration space of the robots themselves. As machines grow more articulated and their design spaces explode in size, the ability to map where danger concentrates, with a few hundred simulations instead of millions, may become as important to robot safety as the strength of any individual part.

Subject of Research: Bayesian optimisation for identifying stress-critical robot poses in a cable-driven hyper-redundant manipulator

Article Title: Configuration-space Bayesian optimisation for stress-critical pose identification on a hyper-redundant robot

Article References: Poka, A., Ludovico, D., Verme, L. D. M. C. D., Manara, F., Canali, C., Berselli, G., & Caldwell, D. G. (2026). Configuration-space Bayesian optimisation for stress-critical pose identification on a hyper-redundant robot. International Journal of Intelligent Robotics and Applications. https://doi.org/10.1007/s41315-026-00586-z

Image Credits: AI Generated

DOI: 10.1007/s41315-026-00586-z

Keywords: Bayesian optimisation, hyper-redundant robots, Gaussian process, finite element analysis, Von Mises stress, surrogate modelling, cable-driven manipulator, robotics, structural design, worst-case stress, Expected Improvement, configuration space