Relationships Between Petrologic And Physicomechanical Properties Of Granitic

Based on the borehole No.1 and No.2 granite of high-level radioactive waste disposal at Bayannuoergong in Alax,Inner Mongolia,we carry out the systematic experimental study.According to the depth and lithology,we selected 22 groups samples(10 of borehole No.1,12 of borehole No.2)to carry out the thin section identification,rebound test,wave velocity test and mechanical test.Through the thin sections identification,we get the petrological characteristics of the samples,including particle size,mineral composition,micro-cracks and so on,on this basis,we quantify the characteristics indicators that selected reasonably,in order to establish corresponding relationship with the physical and mechanical properties.Using the homemade rebound assist device,and adopting the effective testing methods,to experiment rebound test with the core samples to get the rebound values.Through compared with quartz particle size,found they have obvious correlation.Through wave velocity test,we obtain the compressional wave velocity and shear wave velocity respectively of granite.Compared with micro-crack density,we found that micro-cracks exists a threshold on the influence of wave velocity,that is to say,when the micro fracture density is greater than the threshold,they have a good correlation,but when it less than the threshold,their correlation is not obvious.Conducting uniaxial compression test to get the corresponding mechanical parameters,and establish the relationship between article size,mineral composition,micro-cracks above-mentioned and rock mechanics properties(including strength and deformation parameters obtained by uniaxial compression test),found that the mineral content and particle size have a greater effect on the mechanical behavior of the rock,and micro-crack density exists a threshold on the influence of the mechanical properties.Based on comprehensive consideration of every quantitative index and their connection with uniaxial compressive strength,methods of BP neural network are adopted to predict uniaxial compressive strength of granite with basic index parameters of mineral composition,density and longitudinal wave velocity,then grey correlation analysis is conducted to verify the rationality of chosen index parameters.