Training data and results: Reliability assessment of FRP-RC columns using AK-MCS with a piecewise two-branch approach

Data behind the column examples of the article "Reliability Assessment of FRP-RC Columns Using AK-MCS with a Piecewise Two-Branch Approach" (A. Goldack, R. Koehler, A. U. Syed; submitted to ArXiv as preprint). The dataset contains, for five configurations of a slender FRP-reinforced concrete column (l60n07, l60n09, l80n04, l40n04, l40n07): - training_points/: all non-linear FE analyses (OpenSeesPy) used to train the two-branch Kriging surrogate, with basic random variables, surrogate inputs, FE resistance, governing failure mode and origin of each point; - input/: geometry, design load, distributions of the random variables and the surrogate and AK-MCS settings including random seeds; - results_summary.csv: Monte Carlo, FORM and importance-sampling results, leave-one-out residuals and the FE spot check at the design point, as reported in Tables 4 to 6 of the article. Together with the method described in the article, the training points are sufficient to rebuild the surrogates and to recompute the reliability indices. The Monte Carlo and importance-sampling pools are not included, because they can be regenerated from the distributions and random seeds. See README.md for a description of all columns.