Update to the Leps Strength Envelopes for Rockfill Based on 900 Large-Scale Triaxial Tests

This study presents a database of over 900 large-scale triaxial tests that were used to calibrate and refine the Leps strength envelopes for rockfill materials. A novel nonlinear shear strength criterion for rockfill is introduced that can replicate the Leps strength envelopes, is derived in terms of both shear/normal and principal stresses, employs a single fitting parameter with direct physical significance, and allows for smooth transitions between rockfill strength categories. An accompanying empirical model is presented to estimate rockfill strength based on void ratio and rock type, offering a practical tool for rapid application. The model simplifies to five strength envelopes and demonstrates strong predictive performance, while maintaining simplicity. The evaluation of prediction accuracy indicates that the strength model is applicable for soils with a median particle size greater than 2 mm and a fines content of less than 25%. Analysis by stress range reveals a limitation of power-law strength criteria: when calibrated using tests at very low confining stresses, power-law models can significantly over- or underestimate strength for normal stresses above 100 kPa. To address this, a 60-600 Guideline is proposed, recommending the exclusion of triaxial confining stresses below 60 kPa and inclusion of at least one test with a confining stress above 600 kPa to obtain reliable calibration of power-law strength criteria for rockfill from laboratory testing data. An improved methodology to convert triaxial test results into shear and normal stresses is also introduced, accounting for the nonlinearity of the failure envelope.