Evaluation of mine scale longwall top Coal caving parameters using continuum analysis

Mining Science and Technology( China)21(2011)787-796Contents lists available at SciVerse Science DirectMining Science and Technology( China)ELSEVIERjournalhomepagewww.elsevier.com/locate/mstcEvaluation of mine scale longwall top coal caving parameters usingcontinuum analysIsManoj Khanal ", Deepak Adhikary, Rao BalusuEarth Science and Resource Engineering Commonwealth Scientific and Industrial Research Organization, Queensland Center for Advanced Technology, 1 Technology CourtPullenvale, QLD 4069, australiaARTICLE INFOABSTRACTArticle history:A mine-scale analysis of Longwall Top Coal Caving(LTCC) is performed using a continuum mechanicsReceived 20 April 2011eived in revised form 18 May 2011finite element solver called COSFLOW. The uniqueness of COSFLOW is that it incorporates Cosseratcepted 16 June 2011continuum theory in its formulation for describing the load deformation of bedded rocks. it is shown thatAvailable online 19 December 2011uch a continuum based code is valuable for assessing the feasibility of introducing LTCC in any mineVarious LTCC parameters, for example chock convergences, top coal failure behavior, strata cavingmechanism, abutment stresses and vertical stresses, were evaluated for a mine using COSFLOWe 2011 Published by Elsevier B Vehalf of China University of Mining Technology.CosseratSimulation1 Introductionin China report that the LTcc is providing high productivity andefficiency in application It is also cost effective because the shearerLongwall Top Coal Caving(LTCC)is an underground mining slices only the bottom part of the seam and the top coal fracturesmethod developed for thick seam extraction, Depending on the due to gravity [2-6 the only additional cost will be added tocoal seam condition various layouts can be considered for the Ltcc the rear conveyor and slightly modify the chocks on the existing[1. The full seam height LTCC mining layout has been chosen for normal longwall equipment. But there are numerous intrinsicthis work as it takes full advantage of support pressure in fractur- and non- intrinsic parameters which govern the feasibility of tcoing and breaking top coal which increases caveability and draw- in any mine and have to be evaluated properly the intrinsicability of top coal, less gateroad development and maintenance, parameters are thickness of the coal seam, coal strength and defor-simple mining system and less equipment 111-mation properties, inclination of the coal seam, roof sandstoneIn full height LTCC layout, the thick coal seam is divided into strength and deformation properties and coal geology the non-two parts bottom and top parts, Fig. 1. the shearer slices the bot- intrinsic parameters are existing equipment support for the normaltom of the seam and the top coal is allowed to break under gravity. longwall extraction, life of the mine financial health of the mineNormally, the bottom coal has a uniform thickness. The bottom and a detailed geological study of the minepart is extracted using a normal longwall extraction techniqueThe LTCC method can be used to successfully extract up to 12 mwith 2.0-3.5 m face height. The extracted coal obtained from the in thick seams. a successful application of LTcc depends on theface is transferred via a conveyor belt, also known as Armoured caving mechanics layout and gas and dust control options overFace Conveyor(AFC), installed in front of the hydraulic support the rear AFC [7, 8]. Under favorable conditions, LTCC is an econom-near the cutting face an additional conveyor is attached at the rear ical underground mining technique. the top coal recovery dependsof the support to collect the fractured coal falling from upper sec- on the top coal fragmentation mechanism and fragment size distri-tion of the seambution generated during normal longwall extraction of the bottomThe LtCC technique is used extensively in China with over 100 coal. The classical continuum mechanics is well suited for thefaces producing over 200 MPa in conditions ranging from soft stress and deformation study of normal longwall and LTCC.(<10 MPa UCS coal)to hard (>50 MPa UCS coal)[2]. A lot of mines The in-situ stresses should be considered properly to assess therisk and efficiency of a modern day longwall mine. In addition, thew Corresponding author. Tel. +07 33274199stresses released and breakage mechanism of top coal should alsE-mail address: Manoj Khanal@CSIRO.AU(M. Khanal).be considered for a中国煤化工: The better1674-5264/ssee front matter o 2011 Published by Elsevier B.V. on behalf of China University of Mining Technoldoi:10.1016/msc011.06027CNMHG788M. Khanal et aL /Minming Science and Technology( China)21(2011)787-796432.3ITIBSS50Bottom coalFig 1. Top coal caving mechanism used in COSFLOW simulationimplementation of the LTCc may be achieved through past experi-SS40ence of mining in identical geological and excavation situations orfrom a detailed numerical modeling of the LTCC using comprehen-sive and accurate mine parameters. The past experience from amine can be a challenging tasks for a new mine as none of the383.7mines are same. Similarly a detailed modeling of a mine is alsoIll_Top_coalchallenging work as mine as a whole can be visualized as a heter376.2372.7I Bottom coalogeneous structure with built-in imperfections. Compared to ana-lytical modeling, numerical modeling provides relatively accurateresults In the past, different two dimensional and three dimen-sional numerical simulations were being used to analyze the feasi-Fg. 3. A typical core log of the model.bility study of LTCC in different mines(9-11]. This paper presents adetailed three dimensional analysis of various parameters whichmay impact successful implementation of LTCC Support(chock) Table 1convergences, top coal failure mechanism and recovery estimation Different cases of the modeL.for top coal, roof caving mechanism, abutment stress and verticalstresses were examined for a mine using continuum mechanics1modeling [12, 13].Massive sandstones, L A IVA 3D finite element code called COSFLOW is used in this studya unique feature of CoSFLOW is the incorporation of Cosserat con-Case 1+double strength of all sandstonestinuum theory in its formulation [ 12]. In the Cosserat model, inter-5(strainCase 1+ softening slope of 3 GPa and residual strengthlayer interfaces goints, bedding planes)are considered to beof 0. 4 MPasmeared across the mass, i.e the effects of interfaces are incorpo-An important feature of the Cosserat model is that it incorporates may cause severe geotechnical and safety problems such as faceinstability, roof guttering, windblast. Thus it is imperative thatbending rigidity of individual layers in its formulation and this roof caveability is appropriately assessed. cosFLow was used tomakes it different from other conventional implicit modelsanalyze the caveability of immediate sandstone roofs at the mineA model was developed with a plan area of approximately2 Model deve9 km. It was necessary to use such a large area to minimize theboundary effects. the plan view of the model can be seen inExistence of easily caving main roof and maximum recovery of Fig. 2. The plan view consists of 11418 elements The mine plansthe top coal are the pre-requisites for a successful longwall top coal for a mine formed a basis for the geometry of the model. The meshcaving. The roof strata that do not cave during longwall extraction was then rotated to align with the principal horizontal stress20001000中国煤化工CNMHGFig. 2. Plan view of the model (left) and close up view of theM Khanal et al / Mining Science and Technology( China)21(2011)787-796Rock mass properties obtained from laboratory studyOther stratigraphic untDensity (kg/m)aT(MPa)UCS(MPa)E(GPa)K(GPa)G(GPa)C(MPa)219207.17353.335.003.843.000.7815.623.42.831.04003.33SS40218700.583.09Table 3Coal mass properties used in the model.Stratigraphic unitDensity(kg/maT (MPa)UCS(MPa)E(GPa)K(GPa)G(GPa)C(MPa)1529.010712.003.007.50Bottom coal529.01073.00125Case 1, 10-20 m from tailgateCase 2, 110-120 m from tailgateCase 1, 110-120 m from tailgateCase 3. 110-120 m from tailgateCase 1, 240-250 m from tailgate20406080100120140Distance(m)Distance(m)(a) Case 1(different location)(b)Cases 2 and 3160Case 1, 110-120 m from tailgateCase 3, 110-120 m from tailgateCase 1, 110-120 m from tailgateCase5.110-120mCase 2, 110-120 m from tailgate120Case4.110-120m0204060801001201401601802006080100120140160180200(c) Cases I and 2(d)Cases 1, 2, 3, 4 and 5Fig. 4. Comparison of chock convergences for different casesdirection. The close up view of the region(fine mesh area)whereIn the model, the top coal is extracted using the forward diago-the chocks were installed is shown in Fig. 2.nal method as shown in Fig. 1. When the"n"excavation step of theIn this study first 750 m retreat along a 250 m wide panel is bottom coal is being extracted the fractured coal generated in theconsidered. the coal seam was divided into bottom and top parts. top coal at"n-1"excavation step is also being collected as shownThe bottom part was assigned a uniform height of 3.5 m and the in the figuretop coal was assigned with remaining height of the coal seam.A number of detailed 3D COSFLOW simulations were conductedThe bottom part was extracted using the normal longwall extrac- to assess the caveability, chock capacity and behavior of top coaltion method. The top coal was extracted in such a way that it mim- for the high capacity longwall top coal panels at the mine. Thecs the real top coal caving mechanism, i.e, top coals located models were prescribed roller boundaries on the four sides andbehind the chocks were excavated as soon as the chocks were the base. Initial stress field equal to the in situ stress wasadvancedprescribedThe longwall panel was extracted in steps to minimize theIn this study, 1100 t chocks and 250 m wide panels weredynamic response of the model. First 500 m was extracted with selected. The 1100 t chocks capacities were represented by thecoarser but gradually fining steps Next 50 m was extracted with corresponding stiffness in the model. Each model was discretisedfurther fining steps. Then, the last 200 m was extracted with uni- using approximately 1.5 million finite elements. The parallel ver-form steps of 0.8 m. The chocks are 1.75 m wide and 4. 8 m in sion of CoSFLoW wa中国煤化 Tvas split int32length. These steps are shown by dark lines running parallel to different regionsthe excavation face in the figures discussed in Section 3.HCNMHGM. Khanal et al /Mining Science and Technology (China)21(2011)787-79640m100mThicknessThickness(b) ThicknessFig. 5. Fracturing and thickness distribution of Ss40( Case 5)across the longwall face.Thickness(m)Where oT is the tensile strength, UCS the unconfined cosive stress, e the elastic modulus, the friction angle, K the stiffness,G the bulk modulus. C the cohesive strength and y the dilationangle.In Case 2, the coal strength was doubled compared to Case 1The elastic modulus was same as in Case 1. In Case 3, in addition550+80mto doubling the strength the elastic modulus of the coal seamwas also doubled. In Case 4 strengths of all sandstone units weredoubled. Case 5 was a strain softening model with residual coalstrength of 0.4 MPa and the slope of the softening curve of 3 GPa.The chocks were modeled as finite elements. the deformationat each node of the elements was noted and averaged out to calcu-late the average vertical displacements of the elements. The ele-ment expansion and shrinkage were also noted and considered inthe averaging process. The average deformation of the elementswas considered as chock convergenceFig. 6. Thickness distribution of S540 along the mining direction.3. Results and discussioe, The constitutive model employed for the rock blocks was thestic perfectly plastic Mohr-Coulomb model. The softening/hard- 3.1. Chock convergenceening model was also used to compare with the standard modelsThe parameters which are a function of plastic strain wereThe comparison of chock convergences for a 250 m wide panelassumed based on the experience. The constitutive model used with Case 1 properties at different locations along the mining widthfor the joints were the standard Mohr-Coulomb slip model a typ- is shown in fig. 4(top left). the chock convergence plots are ana-ical core log of the model is shown in Fig3yzed only in the fine mesh area as shown in Fig. 2. In other wordsA number of models were simulated to study the effects of the abscissa values start from 550 from the start line, i. e, 20 m invariation on strength of sandstones and coal. Table 1 presents the fine mesh area is equivalent to the 570 m(550+ 20 m)fromthe various rock mass strength properties used in the CoSFLoW the start line. It can be seen from the figure that the middle partsimulationsCase 1 represents the properties shown in Tables 2 and 3. In this the chocks locatedis sagging and yielding higher chock convergence compared tocase all the sandstones were massive without any bedding planes: be observed in the mhowever, planes of weaknesses were introduced in-between SS40 an effect of coal sea中国煤化工 ht) also shand IllA, and lIlA and SS503 results. Case 3 hasCNMHGCase 2 and Casewaus than Case 2M. Khanal et al /Mining Science and Technology(China)21(2011)787-796791at this 80 m location is thickest(about 29 m thick) compared toits thicknesses at other locations. this observation infers that themine geology (thickness)is a major factor to cause the change inthe fracturing location of Ss40 bringing it very closes to the faceline resulting in higher chock loading. Fig. 6 shows the thicknessof SS40 along the mining direction, which clearly indicates thatsS40 attains a maximum thickness at this location In most of thecases the thickness of SS40 varies between 22 m and 26 m(lessthan 28)yielding relatively less chock convergence values. Thisindicates that at about 29 m thickness ss40 may act criticallyexerting excessive load and inducing rapid chock convergence. Itis worthwhile to note that in the case with strain softening coaleven the upper lying sandstone units (SS50 and SS50)seems tofracture right above the chock positions(see Fig. 10 and discussedin Section 3.3) thus exerting excessive load on the chocks andyielding relatively much higher chock convergence as noted inFig 43. 2. Top coal caving behaviorTop coal of a thick seam was discretised into seven elementallayers. These different elemental layers were necessary to investigate the nature and evolution of failure profile within the top coal.It was assumed thaof bottom coon top coal as well as overlaying strata on fracture. For the consid-ered geological condition, top coal, SS40, SS50 and SS60 can be af-ted during the excavation of the bottom coal. The yield pletaken along the vertical cross-section passing through the miningface(i.e, perpendicular to the mining direction). On the plot exceptfor the case of zero yield (i.., coal never yielded and remained elas-tic), every other values indicate that the coal has undergone one oranother form of yielding(i. e, either tensile or shear or combinationof both). On the figure red color shows the yielded fractured coalFig 7. Comparison of yielding behavior of the top coal at 614 m from the start line and blue color shows the intact coal Fig. 7 shows the yield of topfor different cases(red color shows the yielded or fractured coal and blue color the coal at different excavation stages. As observed from the pictures,almost all of the top coal is yielded. the bottom part of the top coalhas yielded in every excavation step for the considered cases. Ex-From the figure it seems the modulus of the coal layer does not have cept in Case 2 and Case 3, top part of the top coal is also predictedmuch effect on the convergence. Similarly the figure(bottom left) to fracture in all other cases. In Case 2 and Case 3, top two out ofshows the effect of coal seam strength in chock convergence Case seven elemental layers can be seen to be only partially fractured.2 has double the coal strength than Case 1. It is interesting to note It is worthwhile to note that in Case 2 and Case 3 coal massthat the coal strength affects the loadings on the chock with fewer strength is assumed to be 24 MPa which represents an extremelyfluctuations and relatively lesser convergence compared to the rare case. Comparison between Case 2 and Case 3 suggests thatweaker coal seam(Case 1). Fig. 4(bottom right)presents the effect although the coal seam strength is same, the elastic modulus canof strength of sandstones( Case 1 and Case 4)and the strain soften- play some role in fragmentation mechanism, i.e., coal with highering models(Case 5). Compared to the standard models, strain soft- stiffness will possibly attract higher stresses with the same amountening models show some peaks with higher convergences of deformation of the surrounding rocks thus favorably assisting insuggesting that the top coal with the strain softening characteristics the fracturing process.Case 5)may have caused breaking of the overlying sandstone unit In most of the cases, except strain softening model Case 5, top(SS40)just above the chocks or in close proximity to the chocks. It is coal at the right hand top corner is not yielded. Recalling theinteresting to note that in almost all the models(standard and strength of different cases, Case 1 has weakest sandstones, Case 2strain softening)maximum peaks are obtained at around 80- has stronger coal seam compared to Case 1 and Case 3 has double900 m and 140-145 m longwall retreat distances. This maythe elastic modulus compared to Case 2. Case 4 is strongest of allattributed to the local variation in mine geologythe cases. Case 5 has a strain softening model In general, the topIn the convergence graphs discussed above, chock convergence coal can be seen to yield in all the cases indicating favorable LTCCcan be seen to attain maximum values at some specific locations: conditions.this is especially true for the case with strain softening coal.To further investigate the fracture evolution pattern, the topFig 5 shows the fracturing of SS40(for the strain softening coal, coal has been examined at different distances from the mining facevertical cross-sections taken across the longwall face at a number distances from the mining face at different excavation steps. Theseof locations. The left pictures show fracturing pattern and the right vertical cross-sections are again aligned perpendicular to the min-pictures show thickness distributions. As can be seen from the ing direction, i. e parallel to the mining face. It can be seen that theigure SS40 seems to be breaking ahead of the mining face at top coal is completely yielded above the chock and is partiallylocations 40 m and 100 m as marked in( fig. 4): whereas at loca- yielded up to 2, 4 m中国煤化工 it abutment distion marked 80 m in(Fig 4). the failure of SS40 can be seen to tance increases fror-racture reducesoccur very close to the face almost right above the chocks. SS40The model resultsCN MHG yield at aboutKhanal et aL/ Mining Science and Technology (China)21(2011)787-796Step A=4.8 m into the goaf when the longwall hasretreated to 598 m from the start lineFE DCBtep B=2, 4 m into the goaf when the longwall hasretreated to 598 m from the start lineStep D-0.8 m ahead of the face when the longwallStep E=2.4 m ahead of the face when the longwallhas retreated to 598 m from the start lineStep En4.8 m ahead of the face when the longwalStep BFig. 8. Top coal yield at different distances from the face for Case4 (red color shows the yielded/fractured coal and blue color the intact coal).4.0 m ahead of the face. this is further verified by fig 9 where all coal may facilitate the fracturing of the top parts of top coal. Exceptthe top coal above the coal seam can be seen to yield. The top coal zero all values indicate failureahead of the face can be seen to partially yield up to a certain dis-In the LtCC, top cotance as shown in the magnified picture.of deformation fract中国煤化工hrkIt has to be noted here that these quasi-static simulations have mass under the initialCNMHGthe continuumsome limitations. The dynamic failure of the lower parts of the top code like, COSFLow, thenu uie supports is notM Khanal et aL/Mining Science and Technology(China)21(2011)787-796793ock shear yield00-1000-500500-1000-50001500-1000-5000300-250-200-150-100Fig. 9, Vertical cross-section of the coal seam(including top coal) along mining direction(left to right) for Case 4. (a)590 m from the start line. (b)670 m from the start line(c)750 m from the start line(zoom-in is shown in the right picture).60080010001200140000800100012001400600800100012001400(a) ss40 unit600800100012001400600800100012001400600800100012001400(b)ss50 unit600800100012001400600800100012001400(c)ss60 unit中国煤化工Fig 10. Comparison of yield for different cases at different excavation steps as noted in the respective figures forCNMHGmIayer or SSO)SS60 units (except zero all values indicate failure)HKhanal et aL /Mining Science and Technology( China)21(2011)787-796Case 3Case 4Case50Distance from abutment (m)Distance from abutment (m)Distance from abutment(m)Distance from abutment(m)(a)646 m from the start line (b)744 m from the start line (c)656 m from the start line (d)644 m from the start line (e)692 m from the start line可2020Distance from abutment(m)Distance from abutment(mDistance from abutment(m)Distance from abutment(m)Distance from abutment(m)( f707 m from the start lin(h)709 m from the st(595 m from the start line ()707 m from the start lineFig. 11. Influence of rock mass strength(including coal strength)on the abutment stress in(Pa)for a 250 m wide panel with 1100 t chocks( BC-bottom coal and TC-topMiddle of theMiddle of the100020000100015002000reoord-5450Fig 12. Vertical stress in(Pa) for 250 m wide panel for Case 1 and 1100 t chocks at the middle of bottom and top coal layers, 630 m from the start iineconsidered in the paper. The dispersion can be well studied with behavior of overlying strata specially the sandstone layers SS40,coupled finite-discrete element codeSS50 and SS60 during longwall excavation.3.3. Strata caving behaviorFig. 10 shows a comparison of fracture for different cases at718 m from the start litsuggest that the S$40中国煤化工. The picturesIn addition to chock convergence and top coal caving analyses, ure extend up to theculty and fail-CN MH Ge nature andthe models were also used to better understand the deformation strength of sandstones tne seventy or the tailure may be different.M. Khanal et al/Mining Sctence and Technology(China)21(2011)787-796Compared to Case 4, other cases have weaker sandstones: henceThe model predictions indicated that the chock convergenceseverity of failure is higher for other cases. The failure pattern ofwas likely to be between 20 and 30 mm in average whereasSS50 and SS60 shows that the failure is contained within the miningit could reach up to maximum of 130 mm if chocks were leftface however extended beyond the chain pillars. The strongeststanding for a long time, i.e, several hours. thus it would besandstones on Case 4 show the failure lagging relatively behindnecessary to maintain a critical minimum retreat rate tothe mining face and less extended on the side pillars. The SS50mitigate the possibility of face instabilityand SS60 would show the arch type failure on the horizontal(2)Top-coal caving behaviorTo investigate the top coal fracture evolution pattern seven3. 4. Abutment stresselemental layers of the top coal starting from the top ofthe top coal to the bottom of the top coal for different casesAbutment pressure may be considered as a main parameterwere analyzed. in most of the cases, the pictures suggest thataffecting the caveability of top coal [1,] and the abutment pressurethe top coal would break easily.in front of the LTCC face is affected by the coal strength, seam thick(3)Strata caving behaviorness and ratio of top coal drawing. Fig. 11 presents the distributionThe yield patterns of all the cases can be seen to be similar.of maximum abutment stress along the centre of the 250 m wideThis infers that almost all of them could follow a similar fail-panel with 1100 t chocks for Case 4 and Case 5 at a particular exca-ure pattern, similar cave in condition and could show similarvation step. The pictures on the top row show the abutment pres-trend in chock convergences.sure measured for the bottom coal and the bottom row shows theA number of following important observations could beabutment pressure measured for the top coal. The"H"on top rightmade from strata caving analysis: failure patterns of thecomer of each figure shows the location of maximum abutmentrespective sandstones were similar. Failure of SS40 is pre-stress while excavation. Being the model having strongest sanddicted to extend up to the mining face and beyond the min-stones Case 4 shows the highest abutment stress for the bottoming face occasionally with the strain softening coal. Inand top coal. This can be attributed to the strength and massive natcomparison to the fracturing in S$40 and SS50, the fractureure of sandstones present in the model. the similar observation hasfront in SS60(in some cases SS50 as well)seemed to takebeen noted for the chock convergence(discussed in Section 3. 1).arch shapes and lag substantially behind the face line.Case 4 shows the maximum abutment stress of approximately(4)Abutment pressure65 MPa for the bottom coal and 52 MPa for the top coal, which isThe maximum abutment pressures were measured at midmuch higher than Case 1. For other cases except, strain softeningdle of the bottom and top coal layers. It was found that thecase( Case 5). the bottom and top coal abutment pressures werebottom coal abutment pressure could lie between 50 andnoted between 50 and 65 MPa and 34 and 53 MPa, respectively65 MPa and top coal abutment pressureween 30The stiffness and strength of the coal also seem to have effect onand 53 MPa for the standard cases( Cases 1ase4 withthe abutment stresses similar to the observation noted in top coalthe strongest sandstones showed the higabutmentyield. The Case 2 has stronger coal strength but low stiffness com-stress for the bottom and top coals.pared to Case 3, as a result Case 2 show lower abutment stress onThe stronger and more massive the roof strata the highertop and bottom coal seam compared to the Case 3.would be the abutment stress the front abutment stress isThe stronger and more massive the roof strata the higher will bepredicted to attain as high as six times the pre-mining stressthe abutment stress. At this depth of approximately 400 m thein the bottom coal and five times the pre-mining stress inin situ vertical stress can be expected to be around 10 MPa. thusthe top coalit can be seen that the front abutment stress can reach as high assix times the pre-mining stress for the bottom coal and five timesthe pre-mining stress for the top coal.Acknowledgment3.5. Vertical stressThe authors would like to acknowledge the singareni CollieriesFig. 12 presents a plot showing the distribution of vertical stressCompany Ltd for providing the data and permission to publish thefor 250 m wide panel for Case 1 with 1100t chocks. the figure is paper.plotted at the distance of 630 m from the start line the plot indicatesthat the vertical stress in the chain pillar can be more than 40 MPa, References1. e, more than four times the in situ pre- mining stress. This stresstured)at present. The vertical stress above the chock is lower than日m个可 caving mining Chinaestimate can be used in the design of chain pillars. As seen from I1 Zhong M]. Theory andwhite awgwall mining New York: wiley: 1983M. wright B, KraemerApplication of longwall top coal caving to Australian operations. CSIRO-ACARPby the chocks. It has to be noted here that once the lower part ofreport C11040: 2003the top coal fractures the stress distribution on the top coal will bedeformation and failure during top coal caving. Int J Rock Mech Min Sc4. Conclusion999:36:551-8.[61 Alehossein H, Poulsen BA Stress analysis of longwall top coal caving. Int]Rock(1)Chock convergence[71 Humphries P, Poulsen B, Ren T. Longwall top coal caving applicatiThe chock convergence was not uniform due to the nature of 18] Humphries P. Poulsen B. Sliva R. Geological assessment and numericalhe sandstones present in the mining zone. As anticipatedthe assumption of massive strata compared to bedded strata中国煤化工yielded higher chock convergence. the coal seam strengthfor modelling topdid not seem to affect the chock convergences.CNMHGM. Khanal et aL/ Mining Science and Technology(China)21(2011)787-796[10] Poulsen B, Evaluation of software code UDEC for modelling top coal caving in [12] Cosserat E, Cosserat F Theorie des corps deformable. Paris: Hermann: 1909Australian environment. CSIRO exploration and mining report 1115F[13] Adhikary DP, Dyskin AV A continuum model of layered rock masses withNumer Anal Meth Geomech111 Xie GX. Chang JC, Yang K. Investigations in to stress shell characteristics of1998:22(4):245-61ounding rock in fully mechanized top- coal caving face. Int J Rock Mechsci2009:46:172-81中国煤化工CNMHG