Study On The Application Of Refined CSMR Method In Rock Mass Classification Of Mine Slope

In order to overcome the problems of fuzzy value boundary and inconsistent standards encountered in the traditional CSMR method in the evaluation of rock mass quality of open-air slopes,the method itself and data collection are improved and refined,so that the rock mass can be improved and refined.The quality evaluation is more in line with the actual situation on site,and the evaluation results are more accurate.Data collection is based on standard drilling core cataloging and rock mass structural plane mapping.Basic data such as RQD,fissure ratio and structural plane occurrence,shape,roughness,filling and thickness are collected,and obtained through rock mechanics tests.The physical and mechanical parameters of the rock refine the data collection process.The improvement and correction of the CSMR method are mainly based on the problems of strong subjectivity in the value of rock mass conditions and poor applicability of complex rock mass in the process of field application.Combined with the research results of experts and scholars and the innovative methods in the evaluation process,the CSMR method is determined.The RMR value in the middle is the value without structural surface correction,and the uniaxial compressive strength,RQD,and joint spacing indicators in the RMR grading parameters are changed from the original fixed value to the piecewise function value according to the value interval,and the result is more accurate.;The value of F₃in CSMR method is changed from 0,6,25,50,60 to 0,6,10,25,30,which makes the classification result more reasonable;the values of F₁,F₂and F₃are selected with good fitting effect and high precision The Boltzmann model is fitted to obtain a continuous function of the value,which avoids the boundary and subjectivity of the value;the structural surface condition coefficient is refined according to the two standard dimensions of the development scale and the degree of integration of the structural surface,which makes the value process more accurate.Simple and clear;in the face of complex rock masses with multiple groups of structural planes,the weighted average method is used to correct the azimuth coefficient of the structural planes.Taking Kyisintaung Copper Mine as engineering examples,the rock mass in each zone of the two mines was evaluated according to the data collection and refined processing method of CSMR method,and refined CSMR values were obtained.The CSIR method and the traditional CSMR method were evaluated and compared,and the comparison results showed that the refined CSMR value was quite different from the RMR value,and the average error rate was-8.70%;the traditional CSMR value was greater than the refined CSMR value,and the average error was The rate of error is-2.42%,and the error is mainly reflected in the influence of the refined processing method on the value selection process and the result is more applicable.At the same time,based on the RMR value,the correlation coefficient between the RMR value and the refined CSMR value is0.9743,which is greater than the correlation coefficient between the RMR value and the traditional CSMR value,which is 0.9597,indicating that the refined CSMR value is more accurate than the traditional CSMR value.In addition,by collecting and refining the CSMR value of the rock mass data of the Kyisintaung Copper mine and other mines,innovatively refining the CSMR value,the cohesion and internal friction angle of the rock,and the cohesion and internal friction of the rock mass The BP neural network method is used to predict the cohesion and internal friction angle of the rock mass.In this paper,the 65 samples are divided into 55 training samples,and 10 are calibration samples.Through the prediction and comparison of the BP neural network method,the average error rate between the predicted value of the rock mass cohesion and the actual value is 2.84%.The average error rate between the predicted value of friction angle and the actual value is 1.88%,indicating that the prediction effect is good,and the refined CSMR value,rock cohesion and internal friction angle can be used as conditions for predicting rock cohesion and internal friction angle.There are few studies on the neural network identification of rock shear strength by refining CSMR values and multiple rock conditions,and it has good reference significance.