| As a large, open pit, origin of Bayan Obo mine is complex. There are many assumptions about the origin of dolomite and iron ore in Bayan Obo mine. Because of the rock layer combination character that killas cover dolomite, limitation of early blast technology and increase of slope exposing time, gradually, surface rock break increases, deep fracture and fault develops and stability of whole slope decreases since mining starts. Since 2000, instabilities and landslides appear in area C, instability tendency appears in area A by degrees. All above result in the fact that lower levels can’t close to the designed bound on time and a huge menace to safety of personnel and equipment was made. The directly and indirectly lost for that reason is nearly 100 million Yuan a year. Cost for survey and controlling of the slope is nearly tens of millions Yuan a year. In order to avoid landslide and big sliding stone in no instability and not strengthened areas, to ensure safety of open-pit mine production, evaluation of landslide hazard in these areas is necessary.In practice, the geological structure parameters of landslide occurred are important and instructive to forecast and evaluation of landslide hazard in the same area. The more similar on the geological structure parameters between no landslide area and landslide occurred area, the more likely to have a landslide. Based on this theory, These experiences from geological analysis and mechanical numerical inversion of landslides in area C are important to the landslide hazard evaluation in no instability areas.The radii and central locations of landslide bodies in area C were obtained by field measurement. According to original geological messages of area C, the numerical model of this area was established. Because of difficulty to get samples, the parameters of rock layers were obtained from inversion by software. The different levels of mechanics parameters of rock and structural plane were assumed with the design combination of orthogonal table. Inputting these parameters to MIDAS/GTS, the status of sliding and instability of slope were simulated, at the same time, radii of landslide bodies, central locations and relative replacements etc. were outputted. Constructing the neural network model in Matlab and training it with the mechanics parameters and characters of landslides, characters of landslides were inputted and mechanics parameters were outputted. When these outputs of trained neural network model converged to a stable level, the training finished. Inputting the field measurement parameters of landslide bodies, the mechanics parameters of rocks and structure layers were outputted from the model, so the parameters of structure layers which were difficult to obtain from field measurement and experiment were obtained.Changing parameter values of rocks and structure layers, the roles of stage slope, whole slope and residual rock strength in processing of landslide were identified by analyzing the parameter sensitivity of limit equilibrium.Comprehensive surveying 1-4th sliding bodies in area C, two different types of landslide were found. No.1 and No.2 landslide bodies have clear slippery beds and borders, conformed to the character of bedding rock landslide. No.3 and No.4 landslide bodies conformed to the character of breaking rock landslide. The extant borders and bed of No.1 landslide body have been found with exploration of electrical resistivity imaging(ERI) method in area C, message of No.2 landslide body couldn’t be got already.According to landslide analysis on area C and principle of landslide geology, when the stability evaluation of no landside occurred area is carried out, the parameters such as shallow breaking zones, breaking zones or faults combination patterns of middle and deep levels and contained water are important to occurrence of landslide. All these parameters can be detected by radar method, ERI method and seismic method.Models of seismic method are based on layered or approximate layered stratum. Researching the relations between artificial waves and received back waves, the message of underground structures and rocks can be obtained.Seismic methods include refraction method, reflection method etc. Rocks in this area are close to the layered, so the effective and regular reflection wave can be produced. With field survey, vibration source such as electric wheel dump trucks, electric shovels and electric drills, electric power noise source such as high-voltage power supply line were found. All these disturbance sources and background noises were investigated by MMS-1 which made in China University of Mining and Technology(Beijing)(CUMTB). Based on the experiments of types of seismic sources, dose of explosive, combination characters and geophone couplings, beating source was chosen. The field seismic parameters were fixed just as the following: track pitch was 2m, the minimum offset distance was 2 meters, 48 receiving, 12 times covered. With these parameters, seismic survey were carried out in 1544 cleaning platform, 1488 cleaning platform and top platform of area C. With Vista and other data processing and interpretation software, datum from field survey were processed by methods such as auto-adapted denoising, pre-stack time offset, and the outputted results showed the message in a range of about underground 100 meters. By comparing these results with geological datum and comprehensive explaining according to geological laws, the old geological structures such as faults were confirmed and the news were discovered. The seismic data showed that actual geological structure development degree were more complex than that showed by original drilling datum.Transmitting antenna of geological radar(GR) emits high frequency pulse electromagnetic waves to measured media, when the waves meet with the inhomogeneous bodies(interfaces), some of them will be reflected back and received by receiving antenna. Basis on the principle, geological radar can work to detect the abnormal underground geological bodies. In this thesis, GR made in CUMTB was used with 200 MHz antenna to detect the same lines of seismic survey. By locale experiment, measurement time window and collecting points of GR were fixed. With system of GR datum processing and explaining software produced by CUMTB, the datum obtained from field survey were processed with methods such as horizontal prediction filter and Hilbert transform. Explaining the results from GR data processing software, the shallow geological message from 5m under cover earth were obtained and the fractured and broken rock zones were identified. All these messages showed decrease of rock strength is caused by explosion and mechanical crushing etc. Surface rock decay degree can also be showed. The result showed that there were crevasse belts and crushed zones in 1544 and 1488 level which had reached dangerous proportion. The broken degree of 1544 level was higher than 1488 level, and broken rock landslide or collapse could occur more easily.The same lines of Seismic method and GR method were surveyed with ERI method. Because ERI method requires a straight measure line, so the straight parts of bend platform were surveyed with partition. Parameters in this survey were chosen as following: 60 electrodes, 118 meters a line. A 236 m survey line was arranged in the top platform of area C. For obtaining more messages about underground faults, on the basis of comparison between Wenner and differential method, differential method was chosen. The detection depth of area C is 76 m, others are 38 m. Concerning the segment height of slopes and distribution field of landslide bodies comprehensively, these parameters could satisfy the project requirement. Comparing message from inversion of ERI method, engineering geological datum and hydro geological datum, the middle and deep geological structures of stage slope were identified. Contained water of some areas were detected and identified by reviewing hydrogeology records and water controlling logs, the results were highly consistent. Some unrecorded contained water was discovered.There are two kinds of landslide risk evaluation methods, one kind is determination methods such as limit equilibrium method and 3-D finite element method, the other kind is random analysis methods such as comprehensive fussy evaluation and reliability analysis. Based on the following facts: 1 Getting inner parameters of rock mass by drilling is expensive. 2 gray-fussy comprehensive evaluation method can process low accuracy and fewer datum. 3 Other evaluation methods require high accuracy mechanical parameters and large number datum of rock mass structures. 4 Integrated geophysical exploration is cheap and fast, but the results about geological message have some uncertainty. According to all these factors, in this thesis, gray-fussy comprehensive evaluation method was chosen, and the message needed were collected by integrated geophysical exploration.According to analysis of all datum, factors such as middle-deep faults and broke zones combination mode, water conditions, shallow broke zones etc. are chosen as key recognition parameters during recognition of landslide hazard area in east pit. Because of the existing of sliding beds and sliding borders, bedding rock landslide shows the character of ‘two vertical and one horizontal line’. According to the hydro geological characters in east pit, the water existed in faults as beds and borders of landslide increase the possibility of landslide occurrence.Considering the views of relevant experts in the field and field work personnel, referring to relevant research reports, five parameters including stage slope were chosen as factors of gray-fussy comprehensive evaluation. Each factor had three levels which indicated risk levels of landslide occurrence. These factors were compared with each other in importance with AHP analysis, then weight matrix was worked out. Reviewing experience of similar projects and researching about fuzzy membership degrees, trapezoid and half trapezoidal function were chosen as fuzzy membership degree functions, and the values were worked out. Reviewing geophysical exploration laws and noise disturbances, the geological messages from geophysical exploration were evaluated in their unreliability. According to completeness of information or datum collection, the point gray degrees were assigned to fuzzy membership degrees. Then weight matrix were compounded and calculated, dangerous membership degrees of landslide occurrence were worked out. Consequently, the risk of landslide occurrence can be evaluated qualitative. Based on the results of evaluation and messages from geophysical exploration, the distribution map of sources of landslide disaster and its influence were drawn out. |
PDF Full Text Request