Probabilistic Seismic Demand Modeling of RC Grid-Like Frames: Assessing Response Sensitivity Across Parametric Structural Configurations

Description

The research hypothesis of this study is that the seismic response sensitivity of RC grid-like frames is primarily governed by structural geometry and member stiffness ratios rather than material strength properties. Specifically, it is hypothesized that higher-order geometric configurations, such as increased story height and specific bay aspect ratios, lead to a higher "Vulnerability Index" (β), where the rate of increase in structural drift relative to seismic intensity approaches unity. The data, generated through approximately 4,974 non-linear time history analyses using the ETABS API across 90 structural archetypes, supports this hypothesis. The primary dataset (ETABS_D.csv) includes variables such as story height (2 to 10 stories), bay aspect ratios (0.33 to 3.0), concrete strength (20 to 35 MPa), and various member dimensions. By relating the Peak Ground Acceleration (PGA) to the Maximum Interstory Drift Ratio (MaxIDR) using logarithmic power-law fits (EDP=a(IM)^b ), the data shows a clear divergence in seismic behavior based on height. Notable findings indicate that taller structures (9–10 stories) exhibit high sensitivity, with β values ranging from 0.93 to 0.96. Conversely, shorter buildings (2–5 stories) display a "hardening" response, with lower sensitivity indices between 0.57 and 0.65. Interestingly, the data demonstrates that concrete compressive strength has a negligible impact on response sensitivity, with β remaining relatively constant between 0.51 and 0.63 despite strength variations. The most critical factor identified is geometric stiffness; certain combinations of column and beam sizes achieved the highest vulnerability index of 0.98, suggesting that the "Strong-Column/Weak-Beam" principle is the most vital design consideration for these frames. To interpret this data, researchers should focus on the slope (β) of the Probabilistic Seismic Demand Model (PSDM). A β value closer to 1.0 indicates a linear relationship where seismic demand increases proportionately with intensity, representing higher vulnerability. Values significantly lower than 1.0 indicate a structure that possesses greater inherent reserve capacity or "hardening" against increased seismic loads. This data can be used to develop simplified fragility curves and to inform performance-based design strategies. By understanding which parametric configurations lead to higher β values, engineers can optimize member sizing to ensure that grid-like frames remain resilient under varying seismic intensities without relying solely on higher-grade materials. The dataset serves as a benchmark for validating simplified analytical models against rigorous non-linear simulation results. This enables a more nuanced understanding of how global structural parameters—rather than localized material upgrades—dictate the safety of grid-like reinforced concrete systems in seismically active regions.

Steps to reproduce

The methodology for this research is based on a systematic parametric framework and an automated analytical workflow designed to quantify the seismic sensitivity of RC grid-like frames. To arrive at the data, a full factorial design was established comprising 90 distinct structural archetypes, representing typical urban construction. These configurations varied across six primary control variables: vertical topology (1 to 10 stories), plan geometry (bay aspect ratios from 0.2 to 3.0), member stiffness (systematic variation of column and beam cross-sections), slab thickness (150–200 mm), and concrete compressive strength (25–35 MPa). High-fidelity finite element models were then constructed in ETABS, utilizing rigid floor diaphragms and lumped seismic mass. Material nonlinearity was captured via a lumped plasticity approach with uni-axial moment hinges for beams and coupled axial-biaxial interaction hinges for columns, strictly following ASCE 41-17 protocols.To define the limit states for probabilistic assessment, nonlinear static pushover simulations were executed for all 90 archetypes. This process involved correlating physical damage progressions, such as plastic hinging and stiffness deterioration, with the Maximum Interstory Drift Ratio (MaxIDR). Rather than using empirical drift limits, five distinct Damage States (DS1–DS5) were discretized based on specific plastic hinge transitions (States A through E). A sequential search algorithm monitored the models during the pushover to record the precise base shear and roof drift at each transition, ensuring the multi-linear backbone curves accurately mapped structural degradation.The core data gathering was conducted through Incremental Dynamic Analysis (IDA), using an automated interface coupling ETABS with MATLAB via the ETABS C# API. Nine ground motion records from the PEER NGA-West2 database were selected to account for aleatory uncertainty. These records were iteratively scaled from a baseline of 0.05g until numerical non-convergence or collapse occurred. The tool extracted time step vectors and floor displacement histories to define the governing Engineering Demand Parameter (EDP) as the MaxIDR across all stories.The resulting dataset of approximately 4,974 simulations was synthesized into a Probabilistic Seismic Demand Model (PSDM). Using a power-law formulation ($EDP = a(IM)^b$) linearized in logarithmic space, a regression analysis was performed to extract the slope coefficient, $\beta$. This Response Sensitivity Index quantifies the rate at which structural demand increases relative to seismic intensity. To validate the findings, the interstory drift thresholds were benchmarked against FEMA 356 and ASCE 41 standards. The alignment at the extensive damage state (DS3) within code-recommended ranges confirmed that the models successfully replicated expected stiffness degradation, justifying the dataset as a robust basis for fragility assessment.

Institutions

• Universiti Malaysia Pahang

  Pahang, Kuantan

Categories

Licence