岩土力学研究的一些反思

传统三轴试验径变形的测量多采用链式环状传感器，假设试件为对称破坏，呈X型或沿中心轴对称破坏为合理测量方式，绘制偏应力与体积应变曲线斜率没有意义，如果在三向等量加卸载下斜率具有含义。弹性阶段比例极限和峰值应力恒定，外加应力大于比例极限应力时比例极限和峰值应力随损伤变化。破坏区荷载与位移关系曲线是材料力学与结构综合体现，应将应力和应变改为视应力与视应变；传统的迟滞回线面积是能量耗散表征是不正确的，应是能量存储、转换和耗散表征；沿中心轴破坏应为拉破坏，能很好地解释岩体中拉裂纹现象，因此提出循环加卸载试验以卸载曲线线性段作为材料参数确定依据。对于孔隙连通材料，施加应力大于比例极限应力，在不排水任意应力状态水压力卸载至0，以卸载曲线的线性段决定相应材料参数，建议传统三轴试验改为50 mm×50 mm×100 mm六面体试件。在比例极限应力内，损伤变量恒定，在外加应力大于比例极限应力时，沿应力主轴各方向损伤演化不一致，且随应力路径变化。

Reflections on rock and soil mechanics research

The achievements of rock and soil mechanics research have been obtained until now. A zipper type transducer is used to measure the radial deformation of a cylindrical sample of traditional triaxial test, in fact; it is assumed that the failure type of testing sample is symmetrical; it is reasonable for the failure to occur by “X” type or along the symmetrical axis of sample. The physical significance of the slope between the average stress (or deviatoric stress) and volume strain relationship doesn’t existed, but it is existed under the condition that an equivalent loading or unloading in the three directions is performed at the same time. Yield proportional limit stress and peak stress keep constant under elastic segment; when an applied stress is greater the yield proportional limited stress, the yield proportional limited stress and peak stress spaces are changed with the damage state; the load and displacement relationship is a comprehensive representation of mechanics and structure of material post-failure, the stress and strain must be changed to be defined as “quasi-stress and quasi-strain”. Traditionally, it isn’t reasonable for the areas of hysteresis to present the energy dissipation of rock and soil material under cyclic loading-unloading case; it is reasonable for the areas of hysteresis to behave the storage, conversion and dissipation of energy. Based on the mechanism analysis, a traction stress criterion must be suggested to describe the failure along the symmetrical axis of testing samples, that is rational to explain the fissure existence in the natural integrated rock mass. It is more reasonable for the material parameters to determine by the linear segment of unloading curve; a formula is suggested to classify the linear segment. When applied stress is greater than yield limit stress, a non-drainage test is performed for the porous interconnection material; the water pressure unloading is conducted to be near zero at a stress state; the material parameters can be obtained by means of linear segment of water pressure and strain curve. A hexahedron testing sample (50 mm×50 mm×100 mm) is suggested to take the place of cylinder. The damage variables keep constant within the yield proportional limited stress, when the applied stresses are greater the yield proportional limited stress, the damage evaluation is different in the different main stress directions, and is changed with stress paths.