Earthquakes pose significant risks to buildings and infrastructure, prompting engineers and researchers to explore innovative methods for assessing structural resilience. A recent study introduces a novel importance analysis method—based on low deviation sequences and orthogonal polynomials—that aims to improve the computational efficiency of global sensitivity analysis (GSA) concerning structural seismic demands.
Conducted by researchers Xiuzhen Wang, Zhaoxia Xu, and Chuanzhi Sun, this study leverages advanced techniques to quantify how uncertainties in structural parameters affect seismic responses. Published on March 12, 2025, the research integrates time history analysis, using the OpenSEES simulation software, to evaluate the influence of various parameters on seismic demands. The team's findings indicate significant advancements over traditional analysis methods.
Significantly, the newly proposed method proves its capabilities by achieving reliable results with only 1,024 samples, whereas many established sampling methods often require thousands of samples, making the new approach more efficient. By employing low deviation sequences for sampling and orthogonal polynomial estimation techniques, Wang and colleagues demonstrated improved consistency and accuracy comparing their method against established alternatives, such as variance-based analysis and Tornado graphical sensitivity analysis.
The rationale for the study stems from the necessity of quantifying uncertainties when assessing structural performance under seismic loads. Traditional methods could fall short, particularly due to extensive sampling requirements dictated by Monte Carlo simulations. With the new method emphasizing information entropy, researchers were able to derive more accurate insight without the computational burden typically associated with larger sample sizes.
The study revealed interesting findings about the impact of specific parameters on seismic demand. For example, it was noted, "The influence of the representative value of gravity load (Ms) on seismic demands is substantial, whereas the influence of the modulus of elasticity of concrete (Ec) is minor." Such insights reveal the varying importance different parameters hold, which is beneficial for structural engineers when prioritizing design criteria.
A detailed case study illustrated the application of the proposed method on the analysis of three currencies of seismic demand: top displacement, base shear demand, and maximum inter-story drift angle. The researchers analyzed both reinforced concrete and steel-concrete frame structures under varying conditions, employing seismic waves such as the El Centro earthquake record to test the robustness of their findings.
The evaluation process included the generation of extensive parameter samples using the Sobol sequence, enabling researchers to process more efficient analyses. Their methodology not only minimized the required sample size for achieving favourable results but improved the methodology's overall execution time.
Adaptive results indicated distinct patterns of influence across random parameters. For the top displacement demand, the representative value of gravity load(Ms) was found to be particularly impactful, alongside the damping ratio (ζ). Simultaneously, for base shear demand and maximum story drift angle, differing weighting of parameters was observed, showcasing the complexity inherent within structural assessments.
The importance of thorough parameter analysis reveals itself not only within the scope of reinforcing structures against seismic damage but also highlights opportunities for future development within the field of earthquake engineering. By adopting methods such as those pioneered by Wang and colleagues, there lies great potential for enhanced structural safety practices.
Overall, this innovative approach fosters significant advancements in structural safety assessments under seismic loading conditions. The effectiveness of the proposed method intertwined with clear benefits reaffirms the need for continuous evolution within analytical frameworks to optimize engineering practices effectively. With the demonstrated reliability of fewer samples yielding comparable results to traditional techniques, building resilience to seismic threats can be analysed with greater efficiency and accuracy, laying the groundwork for safer and more sustainable engineering practices moving forward.