Water presence causes a dramatic reduction of sandstone strength. Under compressive stress conditions, the strength of a rock sample is controlled by frictional parameters and the fracture toughness of the material. Here, we report fracture toughness, frictional and uniaxial compression tests performed on five sandstones under dry and water-saturated conditions, that provide new insight into the mechanical influence of water on sandstone strength. The mechanical data show that water saturation causes a reduction of i) the fracture toughness and fracture energy ranging from 6 to 35% and 21–52%, respectively; and ii) the static friction coefficient ranging from 0 to 19%. The results suggest that the water weakening in sandstones (with a reduction of the uniaxial compression strength ranging from 0 to 30%) is due to the reduction of the fracture toughness and of the static friction coefficient of the materials. The measured fracture toughness and frictional parameters are then introduced into two micro-mechanical models (a pore-emanating cracks model and a wing crack model) to predict the water weakening. It is shown that the models predict the water weakening relatively well with a general slight overestimation (10–20%). Finally, a parametric analysis on the wing crack model revealed that a sandstone's absolute strength can be estimated by means of combined physical and mechanical parameter measurements.

水的存在会导致砂岩强度的显着降低。在压应力条件下，岩石样品的强度由摩擦参数和材料的断裂韧性控制。在这里，我们报告了在干燥和水饱和条件下对五种砂岩进行的断裂韧性、摩擦和单轴压缩试验，这些试验为水对砂岩强度的机械影响提供了新的见解。力学数据表明，水饱和导致 i) 断裂韧性和断裂能分别降低 6% 到 35% 和 21% 到 52%；ii) 静摩擦系数范围为 0 至 19%。结果表明，砂岩中的水弱化（单轴抗压强度降低 0% 至 30%）是由于材料的断裂韧性和静摩擦系数降低所致。然后将测得的断裂韧度和摩擦参数引入两个微观力学模型（发孔裂缝模型和翼状裂缝模型）以预测水弱化。结果表明，这些模型相对较好地预测了水的减弱，但总体上略有高估（10-20%）。最后，对翼裂纹模型的参数分析表明，砂岩的绝对强度可以通过物理和机械参数测量相结合来估计。然后将测得的断裂韧度和摩擦参数引入两个微观力学模型（发孔裂纹模型和翼裂纹模型）以预测水弱化。结果表明，这些模型相对较好地预测了水的减弱，但总体上略有高估（10-20%）。最后，对翼裂纹模型的参数分析表明，砂岩的绝对强度可以通过物理和机械参数测量相结合来估计。然后将测得的断裂韧度和摩擦参数引入两个微观力学模型（发孔裂纹模型和翼裂纹模型）以预测水弱化。结果表明，这些模型相对较好地预测了水的减弱，但总体上略有高估（10-20%）。最后，对翼裂纹模型的参数分析表明，砂岩的绝对强度可以通过物理和机械参数测量相结合来估计。