Comparing the available methods to determine the ultimate bearing capacity of rock foundations and verification based on the measured bearing capacity

Summary
In this research, the capability of the available methods to estimate the ultimate bearing capacity of rock mass is evaluated based on the statistical performance evaluation indices and measured collected data from the reliable literatures. Accordingly, the most appropriate methods are determined based on the minimum error and maximum conformity with the measured values and proposed to practically estimation of the ultimate bearing capacity of rock mass in engineering projects. Research findings can be successfully utilized by designers according to the governing conditions on the problem that help to save the cost and time in the understudied projects.
 
Introduction
Although design of foundations resting on rock masses is usually controlled by the settlement criterion, the bearing capacity of rock mass must be estimated to evaluate the stability. Therefore, in order to provide an efficient design of a foundation, it is crucial to estimate the bearing capacity of rock mass beneath it. There are four most usually used methods to estimate the bearing capacity of rock mass including codes, analytical, empirical and in-situ methods. Each of these approaches has some shortages in determining the ultimate bearing capacity. Thus, determination of the precise method to determine the bearing capacity is required in order to successfully implementation of the related projects.
 
MethodologyandApproaches
Statistical performance evaluation indices were used to compare and verify the available methods for determining the ultimate bearing capacity of rock masses. These indices are correlation coefficient (R), determination coefficient (R²), root mean square error (RMSE) and mean absolute error (MAE). For this purpose, the sufficient in-situ datasets were firstly collected from the literatures. Then, the above indices are computed for nine available methods. Finally, prioritization of the available methods were performed according to the calculated indices and obtained related graphs.
 
ResultsandConclusions
The main results and conclusions of this research can be summarized as follows:
1-       According to the R and R² indices, the most precise methods are LGP1, Goodman2 and LGP2.
2-         Based on the RMSE index, the most precise methods are LGP1, LGP2 and Bowles.
3-         On the basis of the MAE index, the most precise methods are LGP1, LGP2 and Bowles.
4-       The research results can be utilized for engineering applications and also for future researches to overcome the existing shortcoming of the available methods.