Experimental Study Of Important Influencing Factors For Strain Rockburst

Rockburst is a complicated failure process affected by various factors, which can be summarized into two categories, namely, the internal cause and the external cause. Internal factors mainly include high ground stress, rock lithology, rock structure and geological structure and the external factors include hydrogeological condition, construction factors and the stress wave, et.al. Researchers show that rockburst is actually the release process of elastic strain energy stored in rock and then influencing factors for rock mass high energy storage can largely influence rockburst occurrence. There are two necessary conditions for the formation of rock mass with high energy, namely, capability to store largy elastic strain energy and high stress concentration level. Rockburst occurs easily in rock mass with high energy storage and stress concentration. This paper introduces three important influencing factors including unloading rate, rock scale and rock combination forms. Using true-traxial unloading rockburst analogue test system, we performed these three series of parallel tests.This novel true-triaxial testing system is comprised of three main parts, i.e., the loading/unloading device, a high speed camera system, and a AE monitoring system. The main machine is consisted of three mutually independent loading systems. The loading/unloading device, which is comprised of the main stand, a hydraulic control apparatus and force- measuring transducers, can provide dynamic loading and unloading independently in three principal stress directions. During a test, one surface of the specimen can be unloaded suddenly under the true triaxial compression condition, simulating the stress condition for a rock mass at the free excavation boundary in underground excavations. Based on the rockburst mechanism, three loading paths corresponding to three types of rockburst are proposed. “Loading- Unloading-Keeping” is to simulate instantaneous rockburst, “Loading-Unloading-Axial Loading” to simulate delay rockburst and “Loading-Stepped unloading, stepted loading- Axial Loading” to simulate pillar burst. Thus, three series of rockburst tests considering different unloading rates, different rock scales and different rock combinations have been conducted:(1) Using the granite samples with good homogeneity cored from the same place and according to the different excavation methods, four tests with unloading rate of sudden unloading rockburst, 0.1MPa/s, 0.05MPa/s and 0.025MPa/s were tested. The tests results indicate that further decreases in unloading rate do appear to drive the specimen failure at a low stress state and the ratio between critical failure strength and uniaxial compression strength varies in the range of 1.25~1.77. With different unloading rates, the failure modes of specimens changes largely. Rock specimen with higher unloading rate has more obvious dynamic failure process such as debris ejection and has bigger left pit. However, rock specimen with lower unloading rate has more quiet failure process such as spalling damage and has smaller left pit. Observing the SEM photos of fragments from rock fracture surface, the micro tensile fracture features are more obvious for granite with higher unloading rate. Meanwhile, the screening experiment is performed on fragments genera ted from rockburst tests and the size of fragments are measured. It can be obtained that the total amounts of fragments and fragments in micro grain, fine grain and medium grain have an increase trend with the increase of unloading rate indicating more intensive dynamic failure process due to rapid excavation. Fragments generated from rockburst mainly exhibit platy structure and the proportion of fragments with blocky structure having larger ratio of length to thickness is bigger when the unloading rate is higher.Acoustic emission(AE) is an important indicator for rock damage level and it can be found that the AE cumulative energy release has an increase trend with the increase of unloading rate. According to the AE energy evolution, critical points can be determined and more than 90% energy is released at the final rockburst time. With the help of waveform analysis for AE signals, we can better understand the rupture source features. Using the method of Short Time Fouries Transform(STFT), we can obtain the 3D time-frequency spectrum map for each critical point and it has changed from single wave with lower amplitude and shorter duration to multiwave with higher amplitude and longer duration of 2ms, indicating more energy release. All of the AE signals for the whole test are processed using Fast Fouries Transform(FFT) and the frequency which has the biggest amplitude is defined as major frequency and extracted out. By plotting the major frequency distribution in the whole time domain, the frequency bands for these four tests with unloading rates from a high level to a low level can be determined which are 60~100kHz,60~100kHz,100~125 k Hz and 140~150 kHz. With the decrease of unloading rate, AE signals amount decreases and the frequency band up-shifting occurs. It can be seen that rock with higer unloading rate has more signals with higher frequency components indicating more significant tensile rupture. The b-value in Gutenberg-Richter formula which can reflect the relationship between wave amount and wave amplitude is proposed and the b-value of amplitude for major frequency in each critical point is calculated. At the first loading stage, b- value increases stably due to the gradual micro crack formation in rock and then b- value decreases rapidly as the load increases meaning the micro crack coalescence and the macro crack formation in rock. That is to say more and more AE signals with higher amplitude generate. Finally at the rockburst stage, rock specimen step into unstable failure state and a large amount of AE events with high amplitude produce so that b-value drops down to the lowest value. Note that rock with higher unloading rate has lower b-value in rockburst time indicating more large-scale damage occurrence and larger energy release.(2) Using the granite samples with good homogeneity cored from the same place, a series of rockburst tests with specimens’ heights of 150 mm, 120 mm, 90 mm and 60 mm are conducted to study sca le effect. It can be obtained that further decreases in specimen height do appear to drive the specimen failure at a high stress state and the ratio between critical failure strength and uniaxial compression strength varies in the range of 0.93~1.15. With different sample height, the failure modes of specimens changes largely. At the higher sample height, extensile and splitting failures are dominant and then the mixed failure mode including tensile and shear appear with decreasing height, until the single-plane shear becomes the main failure mode at the lowest height and the main fracture plane has a degree of 45 o from the vertical axial. Fragments in coarse grain have the biggest proportion and an increase trend with the increase of specimen height. However, fragments in fine and medium grains have an inverse decrease trend with the increase of specimen height. At the higher sample height, the ejected fragments are more and the dynamic failure process are more obvious which is consistent with the rockburst failure characteristics. With the decrease of specimen height, the average ratio of length to thickness also has decrease trend indicating pronounced blocky structure. Both the intergranular micro crack and transgranular micro crack are generated in rocks with different heights by observing the SEM images.According to the AE energy evolution, critical points can be determined and more than 60%~80% energy is released at the final rockburst time. With the help of waveform analysis for AE signals, we can better understand the rupture source features. Using the method of Short Time Fouries Transform(STFT), we can obtain the 3D time- frequency spectrum map for each critical point and it has changed from single wave with lower amplitude and shorter duration to multiwave with higher amplitude and longer duration of 2ms, indicating more energy release. All of the AE signals for the whole test are processed using Fast Fouries Transform(FFT) and the frequency which has the biggest amplitude is defined as major frequency and extracted out. By plotting the major frequency distribution in the whole time domain, we can find that the frequency bands for these four tests are much different. With specimen height from a high level to a low level, the frequency components in high frequency band significantly decrease indicating the weakened tensile fracture. The b- value of amplitude for major frequency in each critical point is calculated too. At the first loading stage, b-value increases stably due to the gradual micro crack formation in rock and then b-value decreases rapidly as the load increases meaning the micro crack coalescence and the macro crack formation in rock. That is to say more and more AE signals with higher amplitude generate. Finally at the rockburst stage, ro ck specimen step into unstable failure state and a large amount of AE events with high amplitude produce so that b-value drops down to the lowest value. Note that rock with lower height has lower b-value in rockburst time indicating more large-scale damage occurrence and larger energy release.(3) Using the granite and gneiss samples with good homogeneity cored from the same place, a series of rockburst tests are performed on intact granite specimen, intact gneiss specimen and combined granite- gneiss specimen to study the rock combination form effect on rockburst. It can be obtained that intact granite can store more elastic strain energy and has highest failure strength which is about 2.08 times of its uniaxial compression strength whereas intact gneiss has lower failure strength which is about 1.65 times of its uniaxial compression strength. For combined granite-gneiss specimen, it has similar failure strength with intact gneiss and it is 1.10 times of granite uniaxial compression strength and 1.63 times of gneiss uniaxial compression strength. That is to say intact rock has higher capacity to store elastic strain energy and combined layer existing in combined rock reduces its capacity for energy storage.Intact granite experiences violent burst process including crack growth and fragme nt ejection whereas intact gneiss has more silent failure process with less fragments ejec tion and more spalling. However, combined granite-gneiss specimen basically has no fr agments ejection and only has splitting and flaky debris dropping down. Fragments in coarse grain and medium grain have the biggest proportions of all of these specimens. Intact granite has the biggest proportion in micro grain, fine grain and medium grain indicating more generated fragments and violent dynamic failure process. Intact granite has the least average ratio of length to thickness and other two specimens have relatively higher average ratio meaning the intact granite generating more fragments with blocky structure and experiencing more obvious dynamic failure. Both the intergranular micro crack and transgranular micro crack are generated in rocks with different combination forms by observing the SEM images.According to the AE energy evolution, critical points can be determined. Comparing to other three tests, intact granite has the largest energy which is nearly ten times of others. With the help of waveform analysis for AE signals, we can better understand the rupture source features. Using the method of Short Time Fouries Transform(STFT), we can obtain the 3D time- frequency spectrum map for each critical point and it has changed from single wave with lower amplitude and shorter duration to multiwave with higher amplitude and longer duration of 2ms, indicating more energy release. All of the AE signals for the whole test are processed using Fast Fouries Transform(FFT) and the frequency which has the biggest amplitude is defined as major frequency and extracted out. By plotting the major frequency distribution in the whole time domain, the frequency bands for these three specimens GR,GN and GRN can be determined.Intact granite has larger major frequency both in medium- low frequency band and medium- high frequency band whereas intact gneiss has more singles in highe frequency band and then combined rock has its major frequency components in middle frequency band. That is to say intact granite ruptures in both tensile and shear modes and intact gneiss mainly ruptures in tensile mode whereas combined rock has its failure mode between these two. The b-value of amplitude for major frequency in each critical point is calculated too. At the first loading stage, b-value increases stably due to the gradual micro crack formation in rock and then b-value decreases rapidly as the load increases meaning the micro crack coalescence and the macro crack formation in rock. That is to say more and more AE signals with higher amplitude generate. Finally at the rockburst stage, rock specimen step into unstable failure state and a large amount of AE events with high amplitude produce so that b-value drops down to the lowest value. Note that intact granite has lower b-value in rockburst time indicating more large-scale damage occurrence and larger energy release.