CFST Member Tensile Performance with GAN-Augmented Data for ML Prediction

Six experimental specimens and ten parametric models of square concrete-filled steel tubular members under axial tension form the basis of this dataset. The parametric analysis shows maximum load rising from 182 kN to 895 kN as the confinement coefficient increases from 0 to 0.99. Author Hongtao Zhang published the data on figshare in 2026, which includes results from finite element simulations, code comparisons, and machine learning predictions using a WGAN-GP data augmentation method.

Use Cases

• Training machine learning models to predict ultimate load and displacement based on parameters like confinement coefficient and slenderness ratio.
• Benchmarking design code predictions (e.g., Chinese code) against finite element simulation results.
• Evaluating the effectiveness of Generative Adversarial Network (GAN)-based data augmentation methods for small datasets in engineering.
• Conducting parametric studies on the tensile performance of concrete-filled steel tubular members.

Strengths

• Includes results from six physical specimens and ten parametric models, providing a basis for validation.
• Machine learning models trained on augmented data achieved high predictive accuracy, with R² scores of 0.997 for ultimate load and 0.9855 for displacement.
• Finite element simulation results showed good agreement with experiments, with simulation-to-test ratios all below 0.95.

Limitations

• Row count is unknown, which may limit suitability assessment for large-scale model training.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is very small at 5.5 KB, indicating a limited scope of data points.

Provenance

Source

figshare

Collection Method

Data was generated from experimental tests, finite element simulations, and parametric studies of concrete-filled steel tubular members.

Freshness

Last updated 2026-06-01 17:33:29; freshness should be verified.