Anti-collapse on the fragility and sensitivity analysis of composite frame with CFST columns based on the improved Latin hypercube sampling

The robustness of building structures is influenced by a range of uncertainties, including variations in materials and construction errors that occur during the design, construction, and utilization phases. However, the current uncertainty parameters fragility analyses primarily concentrate on the seismic assessment of reinforced concrete frames and steel frames. To evaluate the effect of uncertainty parameters on the robustness of concrete-filled steel tubular (CFST) composite structures, the improved Latin hypercube sampling (ILHS) method is used to extract 300 sets of uncertainty parameters samples, and the corresponding number of frames is modeled using the iFiberLUT fiber elements. The resistance of the structure for different failure conditions is obtained by the random pushdown method. This then leads to further analysis including structural fragility analysis and sensitivity analysis. The results show that the normal distribution could accurately describe the probability characteristics of CFST structures resisting progressive collapse. The coefficient of variation of the results increases to 13.2% and 18.3% with uncertain parameters. According to the standard, the designed structure has a significant safety margin during the beam mechanism stage. Under the same alpha max, the collapse probability of the edge column failure is increased by 45%66% compared to the interior column failure. The span, yield strength of steel, load and beam height have a greater influence on the structure, accounting for more than 80% of all parameters. Structures are more sensitive to variations in these parameters. Furthermore, under the same parameters, the area of the tornado diagrams for displacement response is increased by 18%37% than load response, which shows that the uncertainty parameters have more impact on the deformation.