Finite Element analyses are increasingly utilized to evaluate the serviceability and stability of geotechnical structures under complex loading conditions. While advanced constitutive models provide a more accurate representation of the stress–strain behavior of soils, simpler models remain more widely used in practical applications. One major reason for this is the large effort required to calibrate the parameters of advanced models, demanding extensive laboratory testing and specialized expertise. To make hypoplastic models more accessible in practice, this study proposes a simplified calibration procedure. The objective is to estimate the hypoplastic parameters α and β based on grain size distribution characteristics (d50, Cu). 22 sand mixtures with linear grain size distribution curves (L-sands) were used, varying in both d50 and Cu. The hypoplastic parameters were determined using the “numgeo Automatic Calibration Tool”, based on existing laboratory data (e.g., from oedometer and triaxial tests). Correlations between α, β and d50, Cu were derived and then validated with available data for natural sands. Validation showed that the simplified calibration approach gave good estimates of α and β. However, slight deviations in the peak shear strength were observed for certain materials in dense states. Despite these variations, the simplified method offers a reasonable compromise between accuracy and effort required for extensive laboratory testing.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Simplified Calibration of Hypoplastic Model Parameters Based on Grain Size Distribution Characteristics

  • Joana Bombe,
  • Christoph Schmüdderich,
  • Jan Machaček,
  • Merita Tafili,
  • Torsten Wichtmann

摘要

Finite Element analyses are increasingly utilized to evaluate the serviceability and stability of geotechnical structures under complex loading conditions. While advanced constitutive models provide a more accurate representation of the stress–strain behavior of soils, simpler models remain more widely used in practical applications. One major reason for this is the large effort required to calibrate the parameters of advanced models, demanding extensive laboratory testing and specialized expertise. To make hypoplastic models more accessible in practice, this study proposes a simplified calibration procedure. The objective is to estimate the hypoplastic parameters α and β based on grain size distribution characteristics (d50, Cu). 22 sand mixtures with linear grain size distribution curves (L-sands) were used, varying in both d50 and Cu. The hypoplastic parameters were determined using the “numgeo Automatic Calibration Tool”, based on existing laboratory data (e.g., from oedometer and triaxial tests). Correlations between α, β and d50, Cu were derived and then validated with available data for natural sands. Validation showed that the simplified calibration approach gave good estimates of α and β. However, slight deviations in the peak shear strength were observed for certain materials in dense states. Despite these variations, the simplified method offers a reasonable compromise between accuracy and effort required for extensive laboratory testing.