Rock parameters inversion for estimating the maximum heights of two failure zones in overburden stra

Mining Science and Technology( China)21(2011)41-47罐Contents lists available at Science DirectMining Science and Technology( China)ELSEVIERjournalhomepagewww.elsevier.com/locate/mstcRock parameters inversion for estimating the maximum heights of two failurezones in overburden strata of a coal seamLu Haifeng a, Yuan Baoyuan Wang Lin bb Chizhou Land Resources institute of planning 8 Survey, Chizhou 247100, ChinaARTICLE OA BSTRACTIn order to enter effective parameters of rock mass in a numerical model, the relationships betweenceived 24 May 2010mechanical parameters of rock and rock mass were obtained by an inversion method and an orthogonalL, given our measurements of the maximum heights of two failure zones in the Longdong coal mine.Accepted 2 July 2010Using the maximum heights of the caving zone and the water-conducting fractured zone as test indicesthe modulus of elasticity, the poisson ratio, cohesion and tension strength as test factors and differentlues of reduction enhancement factors as test levels, an orthogonal test was designed to obtain areters inversionoptimum simulation scheme From the analysis of different values of reduction enhancement factorshich affect the test indices, an optimum factor combination for modification of parameters could beMaximum heights of two failure zonesferred. By using modified parameters in our numerical simulation, the maximum heights of the cavingSimilar materal simulationzone and the water-conducting fractured zone in the extensive xiyi area were determined as 15.06 m and6.92 m. These values were almost the same as those obtained by similar material simulation (8.5 m and37.0 m)and empirical prediction(8.4 m and 34. 4 m). These results indicate that the modification ofparameters is a rational method.Copyright e 2011. China University of Mining Technology. All rights reserved1. Introduction2. Geological and hrydro-gical characteristics in theIf the mechanical parameters of rock were employed innumerical simulation of underground engineering, the accuracy of The extensive Xiyi area is located in the west of the longdongthe calculations would be difficult to guarantee. Therefore, they coal mine, northern Jiangsu province, eastern China( Fig. 1).have to be corrected. At present, rock classification systems and The maximum length of the area is 2220 m along the strike. Thisdisplacement inversion are usually used to modify the mechanical trend is 100-480 m long with a dip angle of 14. The primary coalproperties of rock [1-3). However, a considerable amountseam, the No. 7, is approximately 4 m thick. The first productionxpensive and time consuming exploration and observation needs level is at a depth of-285 m. It is a concealed coalfield, entirelyto be carried out on site when these two methods are used. Using covered with about 180 m thick unconsolidated Cenozoic alluvium.numerical simulation for obtaining the maximum heights of two the lowest aquifer with an average thickness of 10. 2 m in thefailure zones in the overburden strata of the No. 7 seam in the alluvium, covers the coal seam outcrop and is clearly a major hazaextensive Xiyi area of the Longdong coal mine as an example, the to safe mining practices, particularly in shallow depth mining.relationships between mechanical properties of rock and rock massAccording to the results of pumping tests at the bottom aquiferwereby inversion and orthogonal tests to solve these (location of pumping boreholes w w2, W3. wa and ws. as shown insimulation provides a new method for parameters Fig. 1). q values are more than 0. 1 L/(s m). Basing on the standards ofdetermination of rock mass.the State Coal Industry, the type of water abundance and perme-ability in this aquifershould be classified as medium (4 According tothe Bureau of the State Coal Industry, when the No. 7 seam was usedin shallow depth mining, water-proof pillars had to be designed toater from the bottom aquifer to flow into mineonding author, Tel. +86 13404145083中国煤化工 s were needed to preventE-mailaddress:luhaifeng7571@126.com(LHaifeng).1674-5264/5-see front matter Copyright o 2011, China UnCNMHGdoi:101016/mstc20102008L Haifeng et al Mining Science and Technology(China )21(2011)41-47tour line of seam foor Crop hme of scamFig 1. Contour map of No. 7 coal seam floor.3. Analysis of numerical simulation in height determination where p is density, E Young,s modulus and v the poissons ratio.of two failure zonesThese values increase with time t(year) and eventually approach3.2. Determination of mechanical properties of rock massused FLAC D code in our investigation. FLACD is a 3D fdifference program which uses an explicit, Lagrangian calculation 3.2.1. Choice of methodheme and a time-step solution to ensure that plastic collapse and In this coal mine the structure of the roof and floor of the coalflow can be modeled accurately. Since FLAc was developed seam in the #7126 face was almost the same as that of ourrticularly for geomechanical analysis and geotechnical applica- numerical model Fig 2). The location of the #7126 face is shown inons, it incorporates some special numerical considerations to Fig. 1. The#7126 work face of the No. 7 seam is, on average, 4.3 mimulate highly nonlinear behavior of geologic material, such as thick, dipping 7-12, with an average of 9. the strike length isstrain harden/soften, irreversible shear and compaction(61430-449 m and the dip width 90-164 m Comprehensive mech-In our study, the 3D model was 100 m(height)x 250 m(strike anized long-wall miwith full caving was used in the exploingth)x 223 m(dip width). The mode was divided into 35, 332 3D tation of the seam. In order to measure the maximum heights of theelements and 42, 739 nodes. Displacement of the model was caving zone and the water-conducting fractured zone, severalrestricted on the boundaries, horizontal displacement by the four boreholes were drilled in December 2002-January 2003.Based ondes and the vertical displacement by the bottom, with a unified these measurements, we calculated that the maximum heights ofrtical load exerted from above, simulating the weight of overlying two zones were 16.70 m and 35.00 m, respectivelyle model included 45 m thick roof strata. 20 m of uncon-ated layers and 31 m thick floor strata. The No. 7 seam is 228mand 4.0 m thick. a generalized stratigraphic column of the ThcknLathNoyy Formationextensive Xiyi area is shown in Fig. 2.rock are presented in Table 1. A Mohr-Coulomb soft strain model 830mUnconsolidatedwas used in association with the mohr-Coulomb failure criterionThe Mohr-Coulomb soft strain model requires paradescribing the rate of cohesion and friction, to drop as a function ofplastic strain. The determination of the parameters is a difficulttask Based on the results of Gao, Xie and wang. we have used theirH:Madstonearameters. shown in Fig 3(7, 8 The slope of the process of softstrain for cohesion and the friction were kept constant for differenttypes of rocks in the modelThe caved roof rock was simulated by a time dependent elasticmaterial with the following properties(9-1Medium sMeduum sundomep=1600+800(1-e-125)kg/m3E=15+175(1-e-125)MPa中国煤化工剧=005+02(1-e-12CNMHG”6hFig. 2 Generalized stratigraphic column of #7126 face in the extensive Xiyi area.L Haifeng et al/ Miming Science and Technology(China)21(2011)41-47Tabe 1Mechanical rock properties,Density pkg/cm) E(GPa)or(Mp strength Friction angle Residual friction Residual cohesiongeq(°)0300010005ym山2611202.10067000.700Fine sandstone02836.01900cording to the data about the maximum height of the two Poisson ratio wasgh a comprehensive consideration offailure zones in the #7126 face the relationships between the effect of theces, the optimal reduction factors ofmechanical properties of rock and rock mass were obtained by Youngs modulus,d tensile strength were 1/20. 1/20inversion and an orthogonal test. Specific methods were as followsind 3/20. The optimal enhancement factor of the poisson ratio wasUsing the maximum heights of the caving zone and the water- 1.25conducting fractured zone, as test indices, the modulus of elasticityhe poisson ratio, cohesion and tension strength as test factors anddifferent values of reduction enhancement at various test levels weused an orthogonal test to obtain an optimal simulation scheme.0.28From the analysis of the effect of various values of reductionenhancement factors on the test indices we inferred the best factorcombination for modification of the parameters. Thus, modification0.24of parameters of the rock mass could be obtained.0.203.2.2. Analysis of orthogonal testIn general. the model with the-Coulomb strquires the following eight paramehen the flacg0.16ode is used: density, Youngs moduratio, cohesionensile strength, friction angle, residual friction angle and residualohesion. based on the results of our investigation into the modi-fication of these parameters, Young's modulus, cohesion and tensilestrength are, in general, reduced to 1/ 3-1/5 of their original valuesor possibly even to 1/10-1/20 in some cases. The Poisson ratioere increased to as much as 1. 2-1 4 times their original valuesPlastic strain/103]. In general, the small differences in density and friction anglesetween rock and rock mass did not require modification of thestwo parameters and was therefore not considered by us. The resultsof modification of cohesion were used to modify the residualohesion. The results of factors and levels from the orthogonal test43are shown in Table 242By employing an orthogonal testing technique for 3 levels and 4factors, an orthogonal La3")table was used to arrange the exper-Nets iwhen no interaction between the factors was considered41alculation schemes were proposed as presented in Table 3.ed on the generalized stratigraphic column of the #7126face, we constructed a numerical model according to the method39escribed in Section 3. 1. The maximum heights of the two failurezones, obtained by numerical simulation, are presented in Table 2(methods of judgment will be introduced in Section 3.3). The38ased on the analysis suggested in Table 2. InTable 3, K represents the maximum height of the caving zone orwater-conducting fractured zone, given the condition of identicalPlastic strain/101factors and levels, where j represents the factor and i the level. R(R= Kimax-Kimin)represents the maximum difference of height ofig 3. Cohesion and fnction angles as functions of plastic strain.the caving zone or water-conducting fractured zone given theAs shown in Table 4. the optimal reduction factors of Young's Table2Factors and levels of the orthogonal test.modulus, cohesion and tensile strength were 1/20. 1/20 and 1respectively and the optimal enhancement factor of Poissons ratio Factor1.3 when the maximum height of the caving zone was used as theSecond level Third levelindex While the maximum height of the water-conducting frac-YH史中国煤化工mtured zone was used as the index, the optimal reduction factors ofYoungs modulus, cohesion and tensile strength were 1/20. 1/20NMHGIAOand 1/10, respectively. The optimal enhancement factor of theL Haifeng et al/Mining Science and Technology( China)21(2011)41-47Table 3Factors for le(3")orthogonal test and results of numerical simulation.Enhancement factor Reduction factor Reduction factorPoissons ratioof tensile strength of caving zone( m) fractured zone(m423423By using the optimal reduction factors( 1/20. 1/20 and 3/20)and stress. Zone ll, adjacent to the caving zone is defined as the water-he enhancement factor(1.25), our numerical simulation estimated conducting fractured zone. above the model, tensile stress exceedshe fractured zones to be 15.06 m and 36.92 m thick.tensile strength of the rock mass on the edge of the subsidenceThese results indicate that the values simulated with the basin, due to the sinking of the overburden strata Tensile cracksreduction factors(1/20. 1/20 and 3/20)and enhancement factor appear in this rock mass and hence this zone is also referred to as(1.25)were close to the measured results. For the purpose of the tensile and compressive stress zone. A bidirectional compres-reducing the error even more, an orthogonal test was used, again sive stress zone is located in the other zone and rock in this zone isased on these reduction and enhancement factors Using the same not destroyed (still in its plastic state)[14], as shown in Fig 4.Basedmethod, the reduction factors of Youngs modulus, cohesion and on the upper height of zones l and ll which is adjacent to zone l, thetensile strength were 3/40. 2/37 and 3/35. The optimal enhance- shape of the deformation curve and failure of the overburden stratanent factor of the Poisson ratio was 1. 22. The maximum heights can be determined, which allows for the maximum heights of thesimulated with these factors were very close to our measurements. caving zone and the water-conducting fractured zone to beTherefore, we conclude that the best factor combination for calculatedodification of the parameters was 3/40, 2/37, 3/35 and 1.22.Given this analysis, the maximum heights of the cavingand the water-conducting fractured zone in the overburden strata3.3. Analysis of the numerical resultsof the No. 7 seam were estimated at about 11.30 m and 36.20(Fig. 4).The best factor combination was used to modify the rockparameters. These were then used in the numerical simulation of 4. Comparison of resultsthe extensive Xiyi area. It is well known that the tensile strength ofrock mass is far less than its compressive strength due to theIn order to verify the foregoing numerical calculations, similapresence of distributed joints and fractures. the characteristics of material simulation and empirical prediction were applied tos w tensile strength obviously redistributed the stress in the analyze the heights of the two failure zones in the overburdenurrounding rock during mining. Hence, the maximum heightstrata of the No. 7 seamhe two failure zones can be determined based on the regularity ofstress redistribution of overburden strata after mining.4.1. Similar matenial simuFig. 4 shows the graph of principal stress distribution aftermining. It can be seen that vertical zoning of stress is very obviousOur similar material simulation was based on the geological andAccording to the magnitude and property of the principal stress, technological conditions of the extensive Xiyi area. Based on thehe vertical zoning of stress can be divided into three zones, ie (I) practical in situ situation and our experimental model anda low tensile stress zone. ll)a tensile and compressive stress zone considering the size of the model as well as the convenience ofand(lll)a compressive stress zone. If the rock stratum is in the measurements, we opted for a geometric proportion of 1: 100. iecaving zone, it will lose its resistance function of stress due to 10 mm of the model was equivalent to 1 m on the ground.serious bed separation and fragmentation of rock in general. According to the similarity criterion [15]. the bulk density proporHowever, the rock is still in a low tensile stress zone due to the tion was 1: 1. 6 compared with the real bulk density. Consequentlypresence of a stress arch after caving Zone l is defined as the caving the related simulation constants were as follows:zone. Although the rock strata of the fractured zone are in a plasticand destructive state, they basically maintain their original conti- for bulk density: C= 1/1.6=0.625nuity. They still maintain a certain ability to bear compressive for strength and stress: Ca=CC=0.00625Table 4Average value Reduction factor of Youngs modulus Enhancement factor of Poisson's Reduction factor of cohesionReduction factor of tensile strengthdexCaving zone(m) Water-conducting Caving zone(m) Water-conducting Caving zone(m) Water-conducting Caving zone(m)Water-conductingfractured zonefractured zone(mfractured zone(m)3723中国煤化工3750120114.11CNMHGL Haifeng et aL/ Mining Science and Technology(china)21(2011)41-47IIICompressiMax Compression=-1 197e+007PaLinestyleMax Tension =2. 186c+ 006PaFig. 4 Principal stress zones of mined rock mass.for time: C=C5=0.increase with the mining distance. due to the constraint of thefor the poisson ratio: Cu= 1.unconsolidated layers. As can be seen from Fig. 6. the maximumeights of the caving zone and the water-conducting fractured zoneThe fundamental components of our similar material siinduced by mining were 8.5 m and 37 m. The ratios of thewere fine sand, gypsum and lime. Suitable ratios of themaximum heights of the caving zone and the water-conductingntent and mechanical properties of the material werefractured zone to the mining height were 2.13 and 9.25by carrying out compression tests of a large number ofrespectivelymatching test samples, shown in Table 5The experiment was carried out on a plane stress model. In thisphysical model, the simulation height was 108.6 cm, including 4.2. Empirical prediction of maximum heights of two failure zones5 cm of unconsolidated alluvium and 26.6-71.6 cm of roof strata,he underburden was 8-53 cm thick and the extraction seam 4 cm.he choice of empirical formula was provided by the Bureau ofDuring the extraction 70 cm of pillar was left on the left side and the State Coal Industry (5). Based on our experimental results the60 cm on the right side to avoid boundary effects. The mining rock in the weathering zone with a strength generally lower thanstrike-ward length was 120 cm and the mining depth 200 c. TheO MPa, belonged to very weak to weak strata, while rock under themodel is shown in Fig. 5. The mining procedure was simulated by weathering zone with a strength of 20-40 MPa, belonged tocutting the seam from left to right in intervals of 10 cm.medium strata 5]. Hence we can conclude that the overburdenhe maximum height of the water-conducting fractured zone strata of the No. 7 seam in the study area belong to weak to mediumincreased gradually with the advance of the mining face. When thstrata Equations for medium to strong strata and for weak stratamining distance exceeded 90 cm, the top of the water-conducting were selected to predict the maximum heights of the two failureactured zone had already reached the unconsolidated layers zones. In order to reduce the difference in using the two equations,(Fig. 6). At this time, the maximum height of the water-conducting the mean square deviation was opted to be negative in the equationfractured zone was 37 cm. However, it will not monotonically for the medium strong strata. The mean square deviation wasComposition of simulation materialockThickness(m) Compressiveuk density Thickness Compressive Bulk density Sand: ime: gypsum Waterstrength(MPa)(10 N/m)(cm)rength(MPa)(10-N/mUnconsolidated layers 25.00m sandstone479.759⑦邱中国煤化工Proportions of the seam were: sand: lime: gypsum: water coal ash= 5: 0. 36: 0.36: 0. 79: 0.56.CNMHGb Proportions of the unconsolidated layer were: sand: lime: gypsum: water: sawdust =5: 0.36: 0. 36: 0. 8Haifeng et al/ Mining Science and Technology(China)21(2011)41-47Results by the formula forweak strata- conductingzone(m)fractured zone(m) zone(m) fractured zone(material simulation. They were almost the same as the maximumheight obtained by empirical prediction, when calculated by theequation for medium strong strata.5. Conclusions1)According to our measurements of the maximum height of twoones in the #7126 face, the relationships between mechanicalFig 5. Simulation model of similar material.properties of rock and rock mass were obtained by inversionand an orthogonal test. this provides a new method for thedetermination of parameters of rock mass by numericalselected positive in the equation for the weak strata. Stsimulation.expressions were as followsEquation for medium strong strata2)Numerical simulation, similar material simulation and empir-ical prediction were applied to predict the maximum height ofH=47M+19-22所100Mtwo failure zones in overburden strata of the no 7 seam16M+36-56Different values were obtained by these methods. The results ofthe comparison indicate that results from numerical modelingEquation for weak stratawere more accurate and are further explained by the rationalityof the correction factors of the parameters.H=62M+32+15,H=31M+5+4(5) 3)The maximum heights of the caving zone and the water-conwhere H and H are the caving zone and water-conducting fracof given parameter values made for limitations in our results. Intured zone, respectivelyorder to obtain better estimates, more test indices should behe average thickness of the No.7 seam in the extensive Xiyiconsidered. It should be noted that the parameters, determinedea was about 4.0 m. two failure zones were predicted accordingby the orthogonal test, were not the real parameters of rockto our use of comprehensive mechanized long-wall mining withass. But, more suitable results of simulation could be obtaineda one time full caving and mining overall height. The results arewith these parameters in order to reflect the actual situation.presented in Table 6.Results indicate that the maximum heights of the water-con Acknowledgmentsducting fractured zone and the caving zone, given our numericalmodeling, were almost the same as those obtained by similarThe authors gratefully aledge the support from theLongdong coal mine. We also thank Mr Lu of Nanjing University ofScience and Technology for his meaningful comments on thisCurvemanuscript and support.[11 Bieniawski ZT. Rock mass classifications in rock engineering John Wiley[2]Zhou]w. Xu WY, Tong FG. Zhu w) Liu XN Inversion forthe diversion tunnel No. 2urnal of Geotechnical Engineenng 2007: 29(10): 3393-400 in Ch[3 Q JX Wang L. Lu FC Wang w, Dong C. Inversion and prediction forCurve ol[51 State(tidings, rivers and railway lines. Beling: Coalaround roadway-insight from numerical modeling. Minang Science and18] Wang XB. Effect of softening modulus on entire deformational charac中国煤化工CN MHGOlongwall mining with top-coalcaving. intermational Journal of Rock MechanicsFig. 6. Curve of two failure zones as the mining face moved 90 cm.2005:2:219-35L Haifeng et aL Mining Science and Technology(China)21(2011)41-47[11] xie GX. Chang jC. Yang K. Investigations into stress shell characteristics of [13 Cai MF. Rock and soil mechanics and engineering. Beijing: Science Press: 2002[14 Gui HR, Zhou QF. Liao DS, Kang QG. Fan ZY, Zou H Prediction of maximum[12] Xie HP, Zhou Hw, Wang JA Li Lz, Kwasniewski MA. Application of FLAC toredict ground surface displacements due to coal extraction and itsomparative analysis. Chinese Journal of Rock Mechanics and Enginee15deling methods. Xuzhou; China University of Mining and996 In C中国煤化工CNMHG