The invasion created by drilling fluids around borehole walls creates considerable variations on mechanical and physical properties of both rock materials and discontinuities. However, up to date there is no comprehensive study has been performed on the variation of fracture toughness values of rocks under drilling fluid saturation with different compositions. By considering this deficiency, the variation of fracture toughness values of five different rocks interacting with drilling fluids with different compositions was investigated in this study. The experiments showed that the negative effect of drilling fluids on fracture toughness values of rocks has a broad range and the addition of polymers to the drilling fluid prevents a dramatic decrease in the fracture toughness values of rocks. Statistical evaluations showed that the fracture toughness values of rocks can be predicted from tensile strength with moderate error by simple regression analyses. However, the multivariate regression analyses produced statistically significant prediction equations with stronger coefficients of determination. On the other hand, the prediction capability of Adaptive Neuro Fuzzy Inference System (ANFIS) model was higher than the prediction performance of simple and multivariate regression analyses.

钻井液在井壁周围造成的侵入使岩石材料和不连续性的机械和物理特性发生了相当大的变化。然而，目前还没有对不同成分钻井液饱和度下岩石断裂韧度值的变化进行综合研究。考虑到这一不足，本研究研究了五种不同岩石与不同成分的钻井液相互作用时断裂韧度值的变化。实验表明，钻井液对岩石断裂韧度值的负面影响范围很广，在钻井液中加入聚合物可以防止岩石断裂韧度值的急剧下降。统计评估表明，岩石的断裂韧度值可以通过简单的回归分析由抗拉强度预测，误差适中。然而，多元回归分析产生了具有更强决定系数的统计上显着的预测方程。另一方面，自适应神经模糊推理系统（ANFIS）模型的预测能力高于简单和多元回归分析的预测性能。