Estimating the static properties of rocks, especially low-strength rocks, is time-consuming, costly, and in some cases impossible. The current study was carried out to evaluate the petrographic (XRD, thin section, and calcimetry), physical (porosity, absorption, density), mechanical [uniaxial compressive strength (UCS), Young’s modulus (E_s), Poisson ratio] and dynamic [compressional wave velocity (V_p), shear wave velocity (V_s), dynamic modulus (E_d)] properties of the Godarkhosh dam site, in western Iran. Then, some relationships were proposed to estimate the mechanical properties using simple regression (SR), multiple linear regression, and artificial neural networks (ANN). The XRD analysis showed that the main clay minerals observed in rocks are Illite, Kaolinite, and Chlorite. Therefore, these clay rocks’ swelling potential is low. In addition, due to the high percentage of carbonate minerals in the marl samples, the mechanical and dynamic properties of the marls samples were higher than shale samples. Statistical analysis showed that both UCS and E_s have a significant correlation with physical properties and V_p. The relationship between UCS with these parameters is more than with the Es. Besides, the UCS and Es’s relationship with V_p were higher than the physical properties. Presented relationships were compared with previous suggested equations. The UCS and E_s relationship, based on universal average data, showed that there is a moderate correlation (RMSE = 0.30, R = 0.74) between these two variables. The ANN exhibits a higher accuracy than the MLR and SR methods in estimating the E_s and UCS. The neural network is also conservative in estimating the modulus of elasticity of the clay-bearing rocks; however, it is not conservative in predicting the UCS of these rocks.

估计岩石（尤其是低强度岩石）的静态特性既费时又费钱，而且在某些情况下是不可能的。目前的研究是为了评估岩相学（XRD、薄片和钙量）、物理（孔隙度、吸收、密度）、机械[单轴抗压强度 (UCS)、杨氏模量 ( E _s )、泊松比] 和动态[压缩波速度（V _p），横波速度（V _s），动态模量（E _d)] 伊朗西部 Godarkhosh 坝址的特性。然后，提出了一些关系来使用简单回归 (SR)、多元线性回归和人工神经网络 (ANN) 来估计机械性能。XRD分析表明，在岩石中观察到的主要粘土矿物为伊利石、高岭石和绿泥石。因此，这些粘土岩的膨胀潜力很低。此外，由于泥灰岩样品中碳酸盐矿物的比例较高，泥灰岩样品的力学和动力特性高于页岩样品。统计分析表明，UCS和两个Ë_小号具有的物理性质和显著相关V _p. UCS 与这些参数之间的关系比与 Es 的关系更大。此外，UCS 和 Es 与V _p的关系高于物理性质。提出的关系与先前建议的方程进行了比较。基于通用平均数据的 UCS 和E _s关系表明， 这两个变量之间存在中等相关性（RMSE = 0.30，R = 0.74）。ANN 在估计E _s和 UCS 时表现出比 MLR 和 SR 方法更高的准确度。神经网络在估计含粘土岩石的弹性模量时也是保守的；然而，预测这些岩石的 UCS 并不保守。