A nonlinear regression model for peak-failure strength prediction of rockfill materials is proposed. It is based on the relationship between the peak-failure stress ratio and the normalized confining pressure as well ...A nonlinear regression model for peak-failure strength prediction of rockfill materials is proposed. It is based on the relationship between the peak-failure stress ratio and the normalized confining pressure as well as the relationship between the normalized peak-failure stress ratio and the exponent function of the intermediate principal stress ratio. This model can well predict the variations of the peak-failure stress ratio with the initial confining pressure and the intermediate principal stress ratio for different rockfill materials under different general stress paths. Comparisons of the measured and predicted results show that the peak-failure strength under the constant-p' and constant-b path is larger than that under the constant-σ'_3 and constant-b path. The predictive capacity of the proposed model for the peakfailure stress ratio is better than that for the peak-failure friction angle.展开更多
基金financial support from the National Natural Science Foundation of China(Grant Nos.51509024 and 51678094)the Project funded by China Postdoctoral Science Foundation(Grant No.2016M590864)
文摘A nonlinear regression model for peak-failure strength prediction of rockfill materials is proposed. It is based on the relationship between the peak-failure stress ratio and the normalized confining pressure as well as the relationship between the normalized peak-failure stress ratio and the exponent function of the intermediate principal stress ratio. This model can well predict the variations of the peak-failure stress ratio with the initial confining pressure and the intermediate principal stress ratio for different rockfill materials under different general stress paths. Comparisons of the measured and predicted results show that the peak-failure strength under the constant-p' and constant-b path is larger than that under the constant-σ'_3 and constant-b path. The predictive capacity of the proposed model for the peakfailure stress ratio is better than that for the peak-failure friction angle.
