Abstract
Large-span single-layer reticulated shell structures are highly sensitive to load distribution. Variations in snow load distribution significantly impact the stable bearing capacity of these structures, where unfavorable patterns can lead to catastrophic failure. This study utilizes a single oblique rod type reticulated shell model to investigate sensitivity across longitudinal and span directions. By dividing the structure into eight stripes in each direction and applying permutations of snow loads, stability is assessed via the arc-length method. Results indicate that the structure is nearly five times more sensitive to span-wise distribution (coefficient 0.404) than longitudinal distribution (0.083), with the most critical loading zone identified between the 4th and 8th rows.
Introduction
Single-layer cylindrical reticulated shells are widely employed in modern architecture for large-span facilities such as gymnasiums and industrial warehouses due to their aesthetic appeal and structural efficiency. However, their high efficiency is coupled with a significant vulnerability to load distribution. While traditional designs often focus on uniform full-span snow loads, historical collapses—such as those in Bucharest (1963) and Hartford (1978)—demonstrate that uneven or localized snow accumulation is a primary driver of structural instability. Asymmetric loading can trigger buckling modes that a uniform load might not, making it essential to quantify the sensitivity of these structures to various partial loading patterns.
Methodology
The research employs a numerical sensitivity analysis on a monoclinic single-layer cylindrical shell model with dimensions of 20m (length $L$), 15m (width $B$), and 5m (height $f$).
Structural Modeling
The model was constructed using ANSYS software, utilizing Beam188 elements to represent member bars composed of $\Phi 83mm \times 4.5mm$ seamless Q345 steel pipes. The geometric configuration and primary dimensions are illustrated in Figure 1.
Zoning and Loading Strategy
To evaluate the impact of localized snow, the roof was divided into a grid of 8 columns along the longitudinal direction and 8 rows along the span direction. Snow loads were applied to different zones based on permutations and combinations to simulate various accumulation scenarios. The zoning strategy for the span direction is shown in Figure 2.
Nonlinear Stability Analysis
Because reticulated shells typically fail through sudden buckling or loss of stability, the arc-length method was employed for nonlinear analysis. This iterative technique allows for the tracking of the equilibrium path even beyond the limit point of stability. The sensitivity coefficient ($S_{F,i}$) was calculated to compare the bearing capacity under partial load ($F_i$) against the capacity under a full-span load ($F_0$).
Results and Discussion
The analysis revealed a stark contrast in sensitivity between the two primary directions of the shell.
Longitudinal vs. Span-wise Sensitivity
The maximum sensitivity coefficient for longitudinal distribution was found to be 0.083, occurring when snow is concentrated in the first column. In contrast, the sensitivity to span-wise distribution reached a maximum of 0.404. This indicates that the structure is approximately 4.8 times more sensitive to how snow is distributed across its arch (the span) than along its length.
Identification of the "Danger Zone"
The most unfavorable loading arrangement was identified as partial loading on the 4th to 8th rows. This asymmetric distribution across the span creates a "push-pull" effect that significantly reduces the stable bearing capacity compared to a heavier, uniform load. This finding highlights why "worst-case scenario" design must look beyond just the total weight of snow and consider localized accumulation patterns.
Conclusion
This study quantifies the sensitivity of single-layer cylindrical reticulated shells to uneven snow distribution. The results highlight that span-wise asymmetry is the primary risk factor for structural collapse, with the 4th to 8th rows representing the most critical zone for snow accumulation. Future research should integrate wind-snow coupling, as wind is the primary environmental factor driving the unfavorable distributions identified in this analysis.
Summary updated