砂土双屈服面的数值建模

通过研究中密砂在等p路径和增p路径下的三轴试验曲线,发现其应力-塑性应变关系曲线在常规应力范围内具有较好的量纲一化特性。选择合适的指标对砂土三轴试验数据进行量纲一化,以量纲一的试验数据为训练样本进行神经网络训练,得到了两种砂土比较理想的双屈服面神经网络模型。研究结果表明两种路径下屈服面是不同的,该方法能反映应力路径的相关性。

Numerical modeling for double yield surfaces of sands

Study shows that there is normalization characteristic about the curves of stress and plastic-strain of middle-dense sands in the range of routine stress under the stress path of constant p and increasing p.The triaxial test data are normalized by choosing proper normalization parameters.The neural networks are trained by regarding the normalized data as samples and then the double-yield-surfaces model of sands described by neural networks are obtained,which show that the double yield surfaces under the two stress paths are different.The modeling method suggested in this paper reflects the stress-path-dependency.