effect of lithological variations of mine roof on chock...

Journal of Scientific & Industrial ResearchVol. 65, September 2006, pp. 702-712

Effect of lithological variations of mine roof on chock shield support usingnumerical modeling technique

A K Verma 1 and D Debi='Department of Mining Engineering, BESU Shibpur, Howrah 721 302

2Department of Mining Engineering, Indian Institute of Technology, Kharagpur 711 103

Received 30 August 2005; accepted 10 May 2006

Interaction between chock shield supports, the most popular powered supports in Indian longwall mines, and surroundingcoal measure strata is analyzed using finite element models. Thickness and material properties of the main roof, theimmediate roof and the coal seam are varied to simulate various geological conditions of Indian coal measure strata.Contact/gap elements are inserted in between the main roof and overburden layer to allow strata separation. Nonlinearmaterial properties are applied with plastic corrections based on Drucker-Prager yield criterion. This paper illustrates effectsof lithological variations on shield load, abutment stress, yield zone and longwall face convergence.

Keywords: Chock-shield, Drucker-Prager, Finite element method, Longwall, Regression

IPC Code: E21 C41128

IntroductionLongwall method is the most productive and

efficient underground coal mining method especiallyfor deep coal seams in USA, China, South Africa andEuropean countries, whereas in India, it faced withboth successes and failures. Several longwall faces inIndia had collapsed due to adverse geologicalconditions and/or inadequate capacity of poweredsupports I. In some cases, entire panel and equipmentshad to be abandoned due to severity of damageoccurred in the form of roof and face collapse", Inmost of the mines (250 m deep), 4x550 to 4x800 toncapacity chock shields are used to support the roofs.

Lithology of Indian Coal Measure RocksRock lithologies above coal seam are significant

factors in mine stability, particularly with respect totheir strength properties. In Indian coal measurerocks, two particular roof rock sections viz.immediate roof and main roof, have been wellaccepted. Generally, composition of these roofs hasgreat influence on pressure distribution at a longwallface. Excessive growth of main roof cantilever cancause severe periodic weighting while the weakimmediate roof can create roof fall problems in

*Author for correspondenceE-mail: [email protected]

between the canopy and the coal face:'. Indian coalmeasure strata accounts for more than 90% ofsandstone and shale, whereas 50-85% of coal measurerocks comprise of sandstone. Immediate roof,consisting of weak and soft carbonaceous shale, sandyshale, laminations of shale in argillaceous sandstone,soft and fractured intercalation, etc., caves in shortlyafter the roof support advances, without forming anystep or overhang",

Main objective of this study is to evaluate theloading behavior of powered support and conditionsof longwall face for different geological variations ofmain roof, immediate roof and coal seam. Studyprovides a comprehensive understanding of loadingpattern on powered support in terms of lithologicalvariations of roof strata especially for identifying thecause of support yielding.

Materials and MethodsCoal measures strata in Jharia and Raniganj

Coalfields of India were modeled to analyze loadingbehavior of chock shields. Typical dimensions ofIndian longwall panel are: width, 120; length,850-1000; depth, 50-300; and average height ofextraction, 2.4 m. Behavior of coal measure stratawere thoroughly studied and explained through ageneralized borehole litholog (Fig. 1). A number of

VERMA & DEB: EFFECT OF LITHOLOGICAL VARIATIONS OF MINE ROOF ON CHOCK SHIELD SUPPORT 703

(m)50

2

Shale & Sst

?Intercalation ofShale & Sst

/;'////////

7.23Sandy shale

Shale 6.56

Shaly Sst 2.9

Sst 5.67

?Shale laminee & Sst

4.2Intercalation 2.4

Coal 0Very fine Sst

-30

Fig. I-Typical geological borehole column

Table I-Variation of input parameters

n, n2 n3 n4 n5

I I 2 0 I5 5 5 0.75 210 x 8 x x

layers of rock were either grouped into compositelayers or simply taken as a unit rock layer, dependingupon the nature, thickness and physico-mechanicalproperties of rock. For Finite Element Model (FEM),material properties and thickness of roof strata andcoal seam are varied to generalize the region of coalmeasure strata.

In FEMs, 4x550 ton capacity Gullick-Dobsonpowered support is used for the analysis of interactionbetween support and surrounding rock strata.Hydraulic legs of powered support are assumed to beelastic-perfectly plastic material upon yielding. Allrock and coal strata are modeled with nonlinearmaterial based on Drucker-Prager yield criterion.

In this study, six lithological variations of thegeological strata above the longwall panels of Jhariaand Raniganj Coalfields are proposed. The thickness

of immediate and main roofs are taken as the ratio ofthe height of extraction (Table 1, Fig. 2).

where, E", = Modulus of elasticity of main roof, Ec =Modulus of elasticity of coal seam, e, = Modulus ofelasticity of immediate roof, T", = Thickness of mainroof, Tc = Thickness of coal seam, r,= Thickness ofimmediate roof, and MCM = Mean Coal Modulus.

Based on the three variations of thickness (4.8m,12m and 19.2m) and modulus of elasticity of the mainroof, the linguistic hedges such as "thin", "thick" and"thicker" roof strata are defined as below 8 m,between 8-15 m and above 15 m respectively.Similarly, "weak", "hard" and "massive" roof stratawill be signified by a ratio of Ell/Ec as 1-3, 3-8 andmore than 8, respectively.

Two-Dimensional FEM AnalysisThe 2D FEM elasto-plastic analysis are performed

using structural analysis software ANSYS. In thisstudy, 2D plain strain condition is simulated. Mesh isdeveloped by quadratic 6-noded triangular elements(Fig. 3a), which describe structural characteristics oflongwall mine viz. longwall face, chock-shieldsupport, mainlimmediate roof, overhang length ofmain roof, gob, contact planes, and various rocklayers. Zoomed view of the longwall face (Fig. 3b)describes structural characteristics of chock-shieldsupport such as canopy, base, hydraulic legs,lemniscate links, pins and gob shield. Domain forFEM is the vertical cross-section along center axis ofthe longwall panel. Boundary of the model is takenfar away from region of interest (longwall face) so asto minimize the end effects around longwall face area.Total width of the model (200 m) consists of 100 m ofintact coal in front of face and rest 100 m consists ofworking area, gob and the chain pillars at mineboundary. Total height of the model (82.4 m) consists50 m above the coal seam and 30 m below the seam.Additional 50 m of overburden load is applieduniformly on top of the model as an external load tosimulate a depth of 100 m from surface to the rooflevel. FEM model consists of 21,129 elements and43,097 nodes. Separation of main roof withsuperincumbent strata is simulated by 2-noded point-to-point contact elements; total of 96 contact elementsare used. Normal stiffness (KII), shear stiffness (K,)

704 J SCI IND RES VOL 65 SEPTEMBER 2006

Main roof

A~~~mmmrFractured roof

Tm = 2,5,8Tc

L: =0,0.75Tc

Tc = 2.4m

Fig. 2---Geometrical (thickness) variations of roof at longwall face

IE 100m ---~)I( 100m ~I

50m

24~TIi~:,II·i"IIIIII~~~~30m

a

.s: ,",' .s: " .,J))/"

",' .... :. • .\_ .. :_~i:-

bgoaf

Fig. 3--a) Finite element model of longwall panel along central axis; b) Close view of FEM model of longwall face

VERMA & DEB: EFFECf OF LITHOLOGICAL VARIATIONS OF MINE ROOF ON CHOCK SHIELD SUPPORT 705

Table 2-Mechanical properties of rock layers

Thickness p E C <p 0No. Strata

kg/m3 GPa MPa 0 0m

I Shale & sandstone (Sst) ? 2330 6.37 1 32 42 Intercalation of shale & Sst 2 1800 4.5 6 25 33 Sandy shale 7.23 2000 4 6 20 34 Shale 6.56 1500 5 JO 30 3.55 Shaly Sst 2.9 2180 2.45 1.5 20 2.56 Sst 5.67 2330 4.4 8 30 3.57 Shale laminee & Sst Till 2330 EIII 1.2 35 4.5

8 Intercalation T; 2130 E; 1.2 20 3

9 Coal 2.4 1360 Ec 2 25 3JO Very fine Sst 2 2500 JO I 15 2

P : Density; E: Modulus of elasticity; C: Cohesion; <p: Friction angle; 0: Dilation/flow angle; ?: changing quantity.

and coefficient of friction are taken as 10 MN/m, 2MN/m and 0.3 respectively for each model.

Periodic weighting interval' varies (21-22 m) fromone seam to the other at Jhanjra longwall project inRaniganj coalfield. Hence, in this study, main roofoverhang length (22 m) is modeled behind the faceline to simulate non-periodic/periodic weightingcondition depending on the lithological strata.However, overhang length depends upon thickness ofthe roof rock layer and physico-mechanical propertiesof rocks, thickness of the roof rock layer and others.Gravity load is applied by assigning the materials'unit weights. For the boundary conditions, thehorizontal displacements are constrained on twovertical sides and the vertical displacements arerestrained at the bottom of the model.

Material Model and ParametersRock Layers

Drucker-Prager failure criterion is used for thematerial yields with a smooth failure surface", Everyrock layer is assumed to be isotropic, homogeneous,and elasto-plastic material. The material propertiesused for each rock layer (Table 2) are determined fromlaboratory tests, field tests or by referring to relevantgeological data. Rock mass properties, mainly com-pressive strength, cohesion and elastic modulus, arereduced by a factor of 1/5 of the values obtained fromlaboratory tests':". Weak and strong coal is consideredwith elastic modulus of 1 GPa and 2 GPa respectively.Cohesion and friction angle of roof strata are varied toanalyze the effect of loading on powered support.

Components of Hydraulic Powered SupportExcept hydraulic legs, all other components are

assigned hard steel properties. Legs are assigned non-

40 b c35.4 .*-_---;~--~'\

Yield stress

K = Stiffness orElastic modulusof hydraulic leg

0.01 0.02 0.04 0.05 0.060.03

Strain, m/m

Fig. 4--Stress-strain characteristics of hydraulic leg

linear elastic properties based on load-deformationcharacteristics of the chock shield support (Fig. 4).There are three different phases of loading conditions(oa, ab and be) on shield legs. In initial phase (oa),shield leg has a near zero modulus of elasticity tosimulate the first-stage deflection (Vo) caused bygravitational load of overburden strata and pre-existing deformation of surrounding strata (Po) beforeshield is advanced and reset with the roof. From insituFEM models, 6 mm of initial roof to floorconvergence is obtained. In second phase (ab),operating phase, an equivalent elastic modulus orstiffness of the shield leg is assigned. In third phase(be), perfectly plastic behavior is modeled once theload on shield exceeds its yielding capacity. Yieldstress for 550-ton capacity powered support is foundto be 35.4 MPa per leg. Elastic modulus of phase (ab)is calculated on the basis of yielding load (Pili) andcorresponding maximum leg convergence (UIII) in theshield leg, as given in the following equation":

K = ~( P,,, - Po )A VIII -vo

r


where, K = Elastic modulus of hydraulic leg; A =Equivalent cross-sectional area of shield leg,0.0381 m2; L = Length of shield leg, m; Pill = Yieldload, 34.5 MPa; Po= Load at Uo, 1 MPa; UIIl=

Maximum leg convergence corresponding to Pill ,

0.026m; and U0 = Leg-convergence correspondingto po. 6mm.

Based on above relation, elastic stiffness ofhydraulic leg in phase ab is estimated to be 1475 MPafor the front leg and 1470 MPa for the rear leg for anextraction height of 2.4 m. Modulus of elasticity forboth lemniscate links is 206.8GPa and the equivalentcross section area of the upper and lower lemniscatelinks are measured as 0.143m2 and 0.012m2respectively.

Goaf/GobCaved rock mass that forms gob plays an important

role in supporting the overburden strata. The degreeof compaction of caved rock mass affects the capacityof gob to support overburden. Theoretically, ifcompaction occurs fully in the gob area, then verticalstress will reach to insitu stress, which is usuallyachieved at a distance of 1I3rd of the mining depthfrom faceline'". In this study, for a depth of 100m;insitu stress would be achieved at a distance of about33 m from faceline. To account for increasingcompaction of gob, three zones are assigned based onthree compacting conditions (loosely packed, packedand well packed gob regions). Stress-straincharacteristic of gob material is taken to be linear(Table 3) 9.

Concepts of Longwall Support LoadLoading at longwall face is manifested by main

roof as a cantilever beam behind the poweredsupports. Length of cantilever depends upon strength,thickness, and elasto-plastic properties. As cantileverlength increases, load on the face and poweredsupport increases. At a particular length, whencantilever load exceeds its strength, cantilever breaks.

Table3---'Material property of gob material

Gob type p, kg/m3 E,MPa v, Poissionratio

Loosely1440 100 0.40

packedPacked 1920 200 0.30Well

2400 500 0.25packed

Two concepts of loading on longwall support aregiven as:

Detached Roof Rock Mass (Dead Weight)As the length of cantilever beam grows to a

sufficient length, it may break ahead of the longwallface. Measurement of shield pressure data shows thatmain roof can break about 3-6 m ahead of the face".In case of weak and thick main roof strata, a deadweight is applied on top of the powered support oncethe roof breaks ahead on the face line (Fig. 5).Powered support capacity in this situation is estimatedby the largest roof block liable to need support. Incase of weak and thin roof, length of roof cantileverbeam is short. For massive and thicker roof, theinterval of cantilever beam is very high.

Deflection of Roof StrataRoof cantilever can grow to a considerable length

if the rock strata is hard-massive and thick-thicker.Then roof beams will deflect and may separate witheach other (Fig. 6). The load carrying capacity of theroof strata does not diminish owing to the deflection.The roof strata are self-supportive meaning that it cansupport the overburden load. At this time, loading onthe shield is low. However, when the roof stratabreaks ahead of the face, it caused tremendous loadon the shield and sometimes air blast also occurred.This phenomenon is sometime referred asuncontrollable periodic weighting.

Results and DiscussionLoad distribution among supporting elements

(Powered support, Goaf and Coal face) must complywith conservation of energy requirements in such away that the combined work of coal, poweredsupports, and goaf must equal the loading imposed byoverlying rock massl2. This means that if coal bed isvery strong and part of the overburden load can besupported by goaf, the loading on powered supportswill be considerably less. On the contrary, if coalbedis weak and deformable and goaf is newly formedthen load on powered support will be higher. Resultsobtained from FEM analysis also support thisconcept.

Abutment Pressure (AP) and Load on CanopyAP distributions in unmined coal are characterized

by peak stresses near faceline, which decreases inmaznitude with distance towards the solid coalb

(Fig. 7). AP results are defined in terms of unit lessquantity abutment stress ratio (ASR), which is the


Roof bre ak line

Fig. 5--Detached roof mass ahead of coal face

Deflection ofroofstrata

Fig. 6--Detlection of roof strata

Distance from face, m

-9 -8 -7 -6 -5 -4 -3 -2 -1 o 2 3 4 5

'" 6Q.

~.n 4(/)

~ 2U3Cii 0(J:eQl -2>

-4

-108+-~~--~--~--~--~--L---L-~~~---4-T-J--~---J~~~~1:'- Plastic ot\ ~ Supported_:

: _Z;s>~ .. ~.: .. _ .. ..l?n~.._ ..l..: • : Unsupported :I I II I zone II I I I-..-..-..-..-..- ..- ..-..-..-..-..~ ..-..-.. ..- .._ .._ .._ .._ .....•....I I I

Insitu stress lecel : I : I

_··_··_··_··_··_··_··_··_··_··_··4··_··_··+ t .. _ .. _ .. + ..I Centroid of IZero stress lexel ': : I

: : leg thrust :I I b a I: '~~K

Elastic zone

Fig. 7-Vertical stress distribution at longwall face

ratio of induced stresses to the insitu stress (2.05MPa). Peak AP is 2-3 m inside the solid coal; this isbecause the face coal has yielded due to high stressand low confinement. AP increases with the increaseof both strength and thickness of the main roof. But

thickness of the strata is significant in determining theself-supporting capability of roof strata.

Among distinct four zones of stress variation(Fig. 7), in the elastic zone, stress uniformly andmonotonically decreased inside the coal seam. Plastic

ri!

708 J SCI INO RES VOL 65 SEPTEMBER 2006

zone is formed from the longwall face to 2 m insidethe coal seam. Vertical stress is reduced as thecoal face yielded owing to spalling and reached toinsitu level at around 30 m inside the coal seam.The peak vertical stress of 5.5 MPa is occurred insidethe coal seam. Unsupported zone is the area betweenface and canopy tip. In this study, it is taken as 0.445m. From FEM results, tensile vertical stress occurs inthe unsupported zone even for thick and hard roof.Thus, span of unsupported should be kept theminimum depending on the working condition.Supported zone is above the canopy supporting therock mass above it.Typically, tensile or low compressive stress is

developed at the tip of canopy than at the rear, sincethe tip portion is considerably distant from the leg andrequires proportionally less force to maintain momentequilibrium. But when canopy tip is not in contactwith the roof then loading of canopy takes place littleinside the canopy tip. If distance between zero canopyloading to the centroid of leg thrust is 'b', and inbetween centroid of leg thrust and the maximumloading point (rear) is 'a', then the ratio of 'b' to 'a' iscalled canopy ratio (Fig. 7), which is almost constant(equal to 2: 1) for each type of roof strata.

Yield Zones in the Face AreaFor every model, the face and the immediate roof

area have yielded the span of yield zone at longwallface that varied (2.00-2.54 m) depending upon mainroof thickness and material properties (Fig. 8). Yieldzone is calculated based on the stress state ratiodefined as trial stress to stress on yield. If stress stateratio is greater than or equal to unity then yieldingoccurs after last iterative load cycle. If the ratio is lessthan one then stress state is in elastic regime. Higherstress state ratio signifies intensive yielding.As a result, strain-energy accumulated in the rock

layer. The situation become dangerous when geologi-cal anomalies are like fault or shear plane intersectwith the layer. Then a big chunk of solid roof willdetach and will sit over a number of powered supportscausing them to solidify. Apart from angle of internalfriction, all other material properties of main andimmediate roof are same (Fig. 8b). Yield zone in roofis extended on top of the unsupported zone but confi-ned due to stronger overlying main roof. This yieldingbehavior is conducive for periodic breaking of imme-diate roof behind the support, and also induce inherentmicro-macro crack to propagate up to shearlbeddingplane and finally detached from roof layer.

The thicker main roof strata interact with coal seamand powered support (Fig. 8c). The roof is 19.2 mthick and weak. The vertical extent of yield zone ismore than the previous cases but, it is not extendedupto the next overlying strata. It is observed that asthe modulus of main roof increases, the yield patternremains the same with less intensity.

Loads on Hydraulic LegLoad on hydraulic legs are the results of interaction

between powered supports and surrounding strata.Variation of the load on the hydraulic legs is animportant manifestation of strata behavior. Fig. 9shows the vertical and horizontal displacements of theleg top as well as the change in inclination of legsafter loading for thicker and weak main roof havingno- (a-a), weak- (a-b) and hard-immediate (a-c) roofs.Horizontal displacement in case of no-immediate roofis more than others, but vertical displacement in caseof weak-immediate roof is more. The more horizontaldisplacement suggests that more loads from goaf sidesare performed due to horizontal forces generated byoverhang main roof, whereas high verticaldisplacement is mainly due to dead weight. In thiscase, stiffer immediate roof is imparting minimumvertical and horizontal displacement to the hydrauliclegs. It suggests that for this particular length ofoverhang main roof, deflection of main roof is less.So for massive and thicker strata, overhang length cangrow longer behind the powered support and impartmore horizontal load as compared to weak immediateroof. In most of the mine, setting pressure is set closeto 80% of the total capacity. Under variouslithological conditions, weak and thicker roofexerts more load on the legs than other cases(Table 4).

In general, load on shield increases with increasingthickness and stiffness of the main roof. This isexpected since much of overburden load will betransferred to the coal seam rather than to the poweredsupport. It is found that, as the main roof overhanglength grows behind faceline, load on the coalface andon the powered support will increase. However, incase of no-immediate roof conditions, thick (over10 m) and massive sandstone type of main roof wouldbe self-supportive and may not exert high pressure onthe coal face, during non-periodic time. However,when overhang length is sufficiently grown behindthe shield; much load will be applied on the coal faceand on the powered support.


Leg Pressure Monitored DataIn the study mine, leg pressures are measured

several times daily for each leg. These data arecollected over a period of twelve (12) days after thefirst periodic weighting; this is equivalent to faceadvancement of approx 18 m. Although a continuousdata monitoring is more representative of stratabehavior, but this data may provide some insight andvalidate FEM and analysis results. Chock shield(capacity, 4x550 ton) was installed in the mine at acover depth of approx 100 m. Other parameters oflongwall panel are as follows: T" 1.S 1 m; Till,

Main roof

Immediate roof

Ec = 1. C = 2MPa

17.48 m; e; 1.00 GPa; E;, 1.77 GPa; and E""4.89 GPa. Fig. 10 shows the measured front and rearleg pressure data of the longwall panel.

In this panel, the main roof rock can be classifiedas thicker and moderately hard whereas immediateroof as weak strata. FEM results suggest that this typeof roof causes high load (31.0-35.4 MPa) on theshield in the front leg. Daily average rear leg pressurefor the same period remains almost the same at31MPa. It is also found that for these roof conditionsthe ratio of front leg pressure to the yield pressuremay be 0.8-1.0, which is supporti ve to the field data

Main roof EEm = 1, T,. = 8, C = 1.2MPa, <p = 35°c Tc

--~----------7-

:1 = 1, C = 1.2MPa. <p = 20° /c /

I/

~::::::::::::::=:--------:;.;:' ..-...-._._._.-_./

Immediate roof

stress state ratioa

-.60263312345679--

b

stress state ratio

o -.577829123456--

710 J scr IND RES VOL 65 SEPTEMBER 2006

Main roof

cEm = 1 T,. = 8 C = 1 2MPa m = 350E 1 1: 1 • 1 't"C C

stress state ratioo -.444645.1.2.3.4____-_~-----.l....-.-.---.---- :

f

Fig. '8--Yield zone around longwall face for a) Hard immediate roof, b) Weak immediate roof, and c) No immediate roof

Horizontal Displacement, mm

-2 -1.5 -1 -0.5 a 0.5E _49~ ~ -_-__~_-_~_~_~_~_~__~_~_~_=_=_~__~~~~0~ -50 ,:

~ -51 ~ ,~'~ -52 Hydraulic leg---. ",',,''J:c

~ -53.~ -54 NJ = 0.159° "Cl NJ = 0.226° "OJ -55 " NJ = 0 2270~-~ b

~ -57 a: no-immediate roof. b : weak-immediate roof.c : strong-immediate roof

NJ = Change in leg inclination towards face after loading

Fig. 9--Resultant displacement of hydraulic leg

obtained from that panel. Most of the FEM modelsalso show lower pressure (5-10% lower) at the rearleg as compared to front leg.

Statistical Analysis of FEM ResultsData obtained from 54 FEMs are analyzed using

multivariate regression analysis. From an engineeringviewpoint, all third and higher-order interactioneffects between input parameters can be consideredinsignificant or not interpretable. So the factorialdesign is constructed with 5 main effects (n I' n 2 '

Table 4-Roof types for which leg pressure exceedsyield limit

Roof Main Roof Immediate Load on legType Type Thickness roof type MPa

I Weak Thick 31.052 Weak Thicker 33.703 Hard Thick 32.384 Massive Thicker 31.995 Weak Thick Weak 32.086 Weak Thick Hard 29.617 Weak Thicker Weak 42.498 Weak Thicker Hard 39.45

-rr nand ns) and second-order interaction'" 3 ' 4 '

between these parameters. Dependent variables suchas ratio of front leg pressure to yield pressure(FLPIYP), ratio of rear leg pressure to yield pressure(RLPIYP) and ASR are standardized into a scale of 0and 1 for statistical analysis (Table 5). Similarly, allmain effects are also standardized between scale of 0to 1.

Leg pressure monitored data shows that RLP is lessthan FLP for the given geomining condition. Leg

VERMA & DEB: EFFECT OF LITHOLOGICAL VARIA nONS OF MINE ROOF ON CHOCK SHIELD SUPPORT 711

27 I-+- Front Leg --- RearLeg -Yield Limit I25~~~--~======~======~==~1 2 3 4 5 6 7 8 9 10 11 12

Days

Fig. ](}-Field measured data of leg pressure from the study mine

pressure is made dimensionless by dividing it with YPof the leg. Regression analysis is carried out usingSPSS software package 13. Backward regression modelis used to obtain the best-fit model for FLP/YP,RLP/YP, and ASR based on the F test and t tests ofeach coefficient as follows.

ASR = 0.186 + 0.4367r, - 0.2887r4 + 0.7207r5- 0.3 107r,7r3+ 0.2857r27r4+ 0.5847r37r4

Above equations provide solutions in standardizedscale ranging 0 to 1. In order to convert them in actualvalues, following mapping function is used:

Yoctua! = YmodelX (Max - Mill) +Mill

where, Yaetua'= Converted value of output parameters,and Ymodel= Output from the statistical model as givenin the above equations.

From statistical analysis, coefficient and themagnitude of t-statistic of T dTe are found higher thanEdEc suggesting that former is more significant thanthe later for determination of load on the shield.Strength of coal and stiffness of main roof are themost significant parameters for estimation of ASR.Based on above equations, leg pressures and ASR areestimated for the study mine as follows: esti matedFLP, 30.97 MPa; estimated RLP, 30.53 MPa;estimated AP, 6.67 MPa; estimated ASR, 3.25; fielddata of FLP (av), 32.74 MPa; and field data of RLP(av), 31 MPa. It can be seen that both estimated front

Table 5--Range of input and parameters in statistical analysis

Input Range Output RangeParameters Min Max parameters Min Ma

x

Em/EC 7r, 1 10 FLPIYP 0.40 1.0

Ej/Ec 7r2 I 5 RLP/YP 0.44 1.0

TmlTC 7r3 2 8 ASR 2.42 4.21

t.tr. 7r4 0 0.75

Ec/MCM 1 2 .'7r5

and rear leg pressures are comparable with theaverage pressures obtained from the field-monitoreddata.

ConclusionsIn general, self-supporting nature of the roof rock

is playing a pivotal role in determination of shieldload. Load on hydraulic legs is high for thicker andweak main roof because thicker and weak main rooftends to exert load as dead weight on the shield. Themagnitude of ASR is less for such roof condition. Onthe contrary, ASR will be high for thicker andmassive roof. For an overhung length of 22 m, shieldlegs are yielded in the presence of a weak roof.However, shield leg does not yield for hard roofmeaning that 22 m overhung length is not sufficient tocause fracture in the roof. Hence, this roof acts asself-supportive and can carry overburden load. On theother hand, for weak main roof, the 22 m overhunglength is sufficient to cause periodic weighting. Ingeneral, rear legs of a chock type powered supportwill carry lower load than the front legs due to lowerstiffness of the rear hydraulic leg. In addition,convergence of canopy tip will directly affect thefront leg pressure rather than rear leg. Multivariateregression analysis suggests that the thickness of mainis the most significant parameter for the determinationof load on the shield. On the other hand, strength ofcoal and stiffness of main roof are the significantparameters for determining peak abutment stress.

References1 Ghosh A K, Why longwall in India has not succeeded as in

other developing countries like China, J Ins! Engr (I) - MN,84 (2003) 1-4.

2 Sarkar S K, Mechanized Longwall Milling the IndianExperience (Oxford and IBH publishing Co. Pvt. Ltd, India)1998,240.

3 Hararny K Y, Magers J A & McDonnell J P, Mining understrong roof, Proc tit lilt COllI' Ground Control ill Milling

edited by J A Hudson (Pergamon Press, Oxford) 1993,491-522.

9 Deb D, Ma J & Chugh Y p, A numerical analysis of theeffects of weak floor strata on longwall face ground control,1 Coal Sci & Engg, 7 (2001) 1-8.

10 Pappas D M & Mark C, Behavior of simulated longwall gobmaterial, in Report of Investigations, RI-9458 (USBM, USDepartment of Interior, USA) 1993.

II Deb D, Development of the longwall strata control,maintainance system (LosCoMS), Ph D Thesis, University ofAlabama, Tuscaloosa, AL, USA, 1997.

12 Barczak T M, Design and operation of powered supports forlongwall mining, Engg & Mining J, June (1993) 28-32.

13 SPSS: Statistical Package for Social Sciences lnc (Chicago,Illinois, USA) 1993.


(West Virginia University, Morgantown, USA) 1988, 171-194.

4 Das S K, Observations and classification of roof stratabehavior over longwall coal mining panels in India, Int 1Rock Mech & Mining Sci, 37 (2000) 585-597.

5 Kundoo S, Thakur B K, Chatterjee S, Bhattacharya R M &Karmakar N C, Periodic strata loading characteristics in amechanized longwall caving face - A case study, Proc Nat SympSustainable Mining Technology: Present and Future (AnnaUniversity, Chennai, India), 14-15 march 2002, 250-261.

6 ANSYS, ANSYS Users Manual (ANSYS Inc, USA), 6.1(2002).

7 Peng S S & Chiang H S, Longwall Mining (John Willy &Sons, New York) 1984,708.

8 Pariseau W G, Application of finite element analysis tomining engineering, in Comprehensive Rock Mechanics,

effect of lithological variations of mine roof on chock...

Documents