'. •*' <',• / " ^ 
 -SV"*' ;, j( *' r , > 
 
 1 ' '' 
 
 
 
 ^i^S: 
 
 
 .?i^ 
 
VO ,t' 
 
 . fL 
 
 
 .^■^ .*'. 
 
 
 "<?> * o » - <5, 
 
 
 ' % 
 
 '< -^^o* o. 
 
 
 
 <^. -""O^ .V 
 
 
 .°: -^..^^ 
 
 
 • -^^ A^ ' /^V^'<-_ '■^<'. c-?;^ »'^lfe'. -^^ 05- ' »: 
 
 V *""* .« 
 
 
 ^*V.' 
 
 
 
 ^_ 
 
 
 ^^ "^^ '^^m.^ . '^^^\ 
 
 
 
 
 
 
 
 
 V 
 
 ^- 'V .^^*' /aK\ ' ^^..^^ ' .^;(M>:'» 'V ./ /^% ^u^a'' ' /Mik^ X c/^ ^' 
 
 
 ^^.-■•. -^ 
 
 .^ 
 
 0^ '^ ^^w-,' . 
 
 ^ ^% ^ 
 
 -^^^^^ 
 
 
 -J^ .ft, -ti 
 
 v^^ 
 
 • \.^^ . 
 
 
 < o 
 
 • ^<^'- -q, ^^TT.-^-.o^ 
 
 -^a^ 
 
 ,-^^\^:J^^\. ./..^:'>o ..■^^\*^:^/%. / -"'• ' 
 
 
 .'< -^^o^ ^ 
 
 
 ♦ -9." 
 
 
 
 
 ■'-^0^ 
 
 
 
 ^q,. ".V,." 
 
 
 
 
 
 
 
 * -r %*> ".^ 
 
 
 • '-^^0^ I 
 
 <> * » o ' . ■^ 
 
 
 
 
 ^°-n^. • 
 
 0° 
 
 
 ♦ ay 'Hi. 
 
 ■ o " » » '^ 
 
 o, *.Tvr' A 
 
 V '^ 
 
 ". '>bV* :iM^^^ '^-Ac^i °. 
 
 O 0^ 
 
 i^ .''■••* < 
 
 
 <?■ * o » o 
 
 
 V* 
 
 '^ A-* .*JU/(W* -^ V 'J, "J"' '•^^ A'-" . ».1».(T»».' ♦ -^ V i>_ "' <*•>, A. 
 
 •"- v./ i^M% 'U..-^^ y^iA'. \.J^ /^fe\ ^^..^^^ .'i^iA'o v./ *■ 
 
 .0' .»LVL% V 
 
 (* ^^ % 
 
 ' A 
 
 
 
 
 ■J. A^ » 
 
 .^,. »^-^-: %,^* . 
 
 
 
 'bv 
 
 ^°-"^. 
 
 
 
 
 ^bV" 
 
 
 , " « « » 
 
 
 ^'•- \,.<^^\'»i^^\..//ate^ %.'^^ iMk'^ ''^^..<^' :^^-^^ %.A^\^'AM:''^^..^ 
 
 
 
 
 
 G^ ^jj ♦;»w*', X 
 
 v"^* 
 
 "o>9' 
 
 ' .^'\ 
 
 -•^.-'-V 
 
 ^°-v.. 
 
 
 *bv* 
 
 .■^ ♦J! 
 
 
 V-0^ 
 
 
 *^ .0 
 
 'bV 
 
 ^'aO 
 
 ' A O 
 
 
 i> « • • , ^^. 
 
 
 G^ -^o *.T^.T^ A 
 
 0° .1^:- °o 
 
 
 • %/ »' 
 
 
 
 
 
 'bV 
 
 
 o««, -^-o 
 
 
 .' ^^^ 
 
 ■^' ** 
 
 X* ■^<» "• * A*- 
 
 • %<^^ 
 
 /'^ 
 
 ..■K >. 
 
 .-& A 
 
 ^bv" 
 
 .0 ^- ^ 
 
 •^^ 
 -^^0^ 
 
 ^•1°^ 
 
 .' <^K 
 
 
 
 •$- A^ ♦ 
 
 
 •^ -^^-^^^ 
 
 
 
 
 
 A*^ ..^"« <t 
 
 
 ^ "e 
 
 ' r^ A^ ' 
 
 r. ■tAC,'^ o, 
 
 
 ■'■ ^ A^ 
 
 "It^^^ 
 
 .*^\.--,^'' 
 
 ^^-^^^ ". 
 

 
 
 <^ *»»o° ^ 
 
 
 v-^^ 
 
 
 
 
 v.^^ 
 
 
 ^^^.J> 
 
 'j^'i^ 
 
 
 
 qv , " o 
 
 • %.^^ ^^^ 
 
 
 
 
 
 
 J 
 
 ,0 ■ ->_ 
 
 J-^-^K • 
 
 ';^o^ 
 
 
 A^" * 
 
 *bv" 
 
 
 
 
 '^0^ 
 
 <!> * o « o ■■ <5,> 
 
 
 
 ,-i.* v_^;^'/'*'? 
 
 
 >. ';'>^o- ^^'' '^a^T^r.^^ .Q> '"'i^J^TWy J-' ^^'-^r^^.^* cO 
 
 
 
 
 -55 °<. 
 
 
 
 
 c"""". c 
 
 1'', •* 
 
 v-0^ 
 
 V 
 
 
 •^^^^^ 
 
 
 
 4 
 
 > 
 
 ^^''.'--'^:% 
 
 
 
 
 % 
 
 ^^. .<^ 
 
 ''^^ ^ ^ ^^^(>:^i\^^ .v-^^ ■;~eiii^; .^ 
 
 
 ^0^ 
 
 .'' ,0^ 
 
 
 
 e 
 
 . ^"^"-o 
 
 " / "^o. 
 
 
 j.--^^^ 
 
 
 ^o' 
 
 
 ^* V -^ 
 
 ■ *w* * 
 
 
 
 
 ** ■v ^<> • 
 
 • 'tt. A^" /^V^"'. %. .c,'^" 
 
 
 
 \/i^To- ,^^' 
 
 
 
 ^bV" 
 
 ^°-n^ 
 
 
 
 
 
 
 O 
 
 ^^ 
 
 
 
 ^ av »j 
 
 
 .... V 
 
 
 -^hv^ :^m^^ -^^6^ f''^»". -^of 
 
 
 v'^-v'!..%J?^*y^!.%7^^'i^^^ 
 
 
 4 o 
 
 
 
 
 
 <?, -ONO-- «, 
 
 
 
 V 
 
 %. 
 
 
 
 .«' ,^ 
 
 
 
 
 4' "^ '^is* a"^ "^ 
 
 ^0^ \D, *?^T^ A <r 
 
 
 
 -^-r 
 
 -^^0^ 
 
 <^°^ 
 
 ,'. o 
 
 
 
 
 
 ■^t 
 
 
 .e yi-,^% -e 
 
 
 />V _ o « o ^ *^ 
 
 ,0^ ^3^ ■^7Y^^ A 
 
 '•: %,** .•• 
 
 b aT .^", ^ 
 
 
 X.^"^ 
 ^/^S 
 
 
 o,' O ",,.'■ J. 
 
 rr.'\o^ 
 
 *bV^ 
 
 '. -^^0^ », 
 
 
 ' ^^ 
 
 <i5°^ 
 
 
 
 " '^bv* :^^^- ^-^o" r^^M" ^ov^ 
 
 b),'„ ^o .c,-^ 
 
 "q,.^^*^^\o^ 
 
 ^^•^^^ 
 
 < o 
 
 ^°-v.. 
 
 
Bureau of Mines Information Circular/1986 
 
 Cutter Roof Failure: An Overview 
 of the Causes and l\/lethods for 
 Control 
 
 By John L. Hill III 
 
 UNITED STATES DEPARTMENT OF THE INTERIOR 
 
Information Circular 9094 
 
 Cutter Roof Failure: An Overview 
 of the Causes and IVIethods for 
 Control 
 
 By John L. Hill III 
 
 UNITED STATES DEPARTMENT OF THE INTERIOR 
 Donald Paul Hodel, Secretary 
 
 BUREAU OF MINES 
 Robert C. Horton, Director 
 

 Library of Congress Cataloging in Publication Data 
 
 Hill, John L. 
 
 Cutter roof failure: An overview of the causes and methods for control. 
 
 Ilnformation circular / United States Department of the Interior, Bureau of Mines; 9094 i 
 
 Bibliography: p, 26 
 
 Supt. of Docs, no.: I 28.27: 
 
 1. Mine roof control. 2. Rock deformation. 
 I. Title. II. Series: Information circular (United States. Bureau of Mines 1; 9094 
 
 -^FN2a5J[J4_ ITN2881 622 s |622'.26| 86-600247 
 
CONTENTS 
 
 Page Page 
 
 Abstract 1 Using prediction of failure in mine design 19 
 
 Introduction 2 Selecting control measures 19 
 
 Background 4 Artificial support 19 
 
 Factors in the propagation of cutter roof failure . . 5 Angle bolting 20 
 
 Openings in a rock mass 5 Truss bolting 21 
 
 Single openings 6 Other supports 22 
 
 Multiple openings 7 Mine design changes 22 
 
 Stress environment 7 Sacrifice entries 22 
 
 Regional stresses 8 Pillar softening and yield pillars 23 
 
 Stress concentrations beneath stream valleys . 9 Roof slotting 24 
 
 Rock mass characteristics 13 Entry reorientation 25 
 
 Rock strength and stiffness 13 Indirect control measures 26 
 
 Minor geologic structures 14 Conclusions and recommendations 26 
 
 References 26 
 
 ILLUSTRATIONS 
 
 1. Cutter roof failure sequence 2 
 
 2. Initial propagation of cutter roof failure to weak bedding plane 3 
 
 3. Severe cutter roof failure 3 
 
 4. Massive roof fail 4 
 
 5. Stress concentration diagrams of rectangular openings with width-to-height ratios of 1, 2, and 3 under 
 
 gravitational loading conditions 6 
 
 6. Stress concentration diagram of six adjacent openings under gravitational loading conditions 7 
 
 7. Stress concentration diagrams of rectangular openings with width-to-height ratio of 3 and loading 
 
 conditions of horizontal-to-vertical-stress ratios of 1, 2, and 3 8 
 
 8. Qualitative example of influence of horizontal-to-vertical-stress ratio on angle of failure propagation 8 
 
 9. Analysis of elemental components of angle of failure as related to stress environment and rock properties . 8 
 
 10. Horizontal in situ compressive stress measurements in the United States 9 
 
 11. Partial mine map from western Pennsylvania, showing preferential orientation of roof falls 10 
 
 12. Mine map from mine in southern West Virginia, showing correlation of roof fall locations with centers of 
 
 overlying stream valleys 11 
 
 13. Values of stress for critical points of an opening beneath a valley versus an opening beneath a hill, for an 
 
 increasing height of the hill 12 
 
 14. Stress concentration in roof-rib corner versus elasticity of coalbed 13 
 
 15. Stress values in roof-rib corner and midspan of roof versus ratio of thickness of the two immediate roof 
 
 members 14 
 
 16. Qualitative interpretation of cutter failure propagation in thinly bedded, single roof rock type 14 
 
 17. Clastic dike in coal pillar 15 
 
 18. Clastic dike in midspan of roof rock of crosscut 16 
 
 19. Geologic and roof failure map of Main A of Greenwich North Mine 17 
 
 20. Clastic dike in-mine occurrences in Eastern United States 17 
 
 21. Section of mine in southern West Virginia, showing location of coalbed roll and local roof falls 18 
 
 22. Cross section of roll A-A' from figure 21 18 
 
 23. Stress values for critical points of an entry versus severity of pitch of an overlying roll 19 
 
 24. Decision process diagram for determining the cause of cutter roof failure and selecting control measures . . 20 
 
 25. Angle bolting hardware 21 
 
 26. Angle bolt installation 21 
 
 27. Two basic designs of roof bolt trusses 21 
 
 28. Plan view of truss installation 22 
 
 29. Plan view of mine entry showing placement of cribbing adjacent to clastic dike to deter cutter failure 
 
 formation 23 
 
 30. Contemporary versions of caving chambers for use in three-entry gateroad configurations 23 
 
 31. Plan for placement of auger holes for pillar-softening concept 23 
 
 32. Cross-sectional view of rib-slotting method 24 
 
 33. Placement of holes for roof-slotting method 24 
 
 34. Apparent values of horizontal in situ stress for the two extremes of entry orientation versus orientation of 
 
 actual principal in situ stress 25 
 
 TABLES 
 
 1 . Maximum shear stress values and analysis of failure 9 
 
 'I Horizontal in situ stress measurements 10 
 
UNIT OF MEASURE ABBREVIATIONS USED IN THIS REPORT 
 
 deg 
 
 degree 
 
 ft 
 
 foot 
 
 in 
 
 inch 
 
 Ib/ft^ 
 
 pound per square foot 
 
 Ib/ft^ 
 
 pound per cubic foot 
 
 pet 
 
 percent 
 
 psi 
 
 pound per square inch 
 
CUTTER ROOF FAILURE: 
 AN OVERVIEW OF THE CAUSES AND METHODS FOR CONTROL 
 
 By John L. Hill IIP 
 
 ABSTRACT 
 
 The Bureau of Mines is conducting research on the causes and methods for control 
 of cutter roof failure in underground coal mines. This hazardous ground control 
 problem exposes miners to the danger of falling roof rock and frequently results in 
 massive roof failure. This report outlines the probable causes of cutter roof failure, 
 which are proposed based on field investigations, numerical model analysis, and 
 in-mine observations. Traditional methods of control are presented, as well as 
 innovative methods based on mining concepts developed during earlier years of coal 
 mining history. The report can be useful to a mine operator for assessing the causes of 
 cutter roof failure on a site-by-site basis and for predicting the probability of its 
 occurrence. A process is presented for selecting an optimum control method that 
 includes both traditional and innovative control techniques for each of the various 
 causes of cutter roof failure. 
 
 'Geologist, Pittsburgh Research Center. Bureau of Mines, Pittsburgh, PA. 
 
INTRODUCTION 
 
 Ground control research conducted by the Bureau of 
 Mines is designed to develop technology that will aid in 
 reducing the frequency of accidents associated with poor 
 ground control conditions. Cutter roof failure often poses a 
 safety hazard to miners and causes delays in production 
 while massive roof falls are cleaned up and unstable roof 
 is resupported. Implementation of control measures 
 reduces the threat of injury to miners and prevents 
 production delays. 
 
 The definition of cutter roof failure used in this report 
 separates this unique type of failure from other ground 
 control problems, to aid in the analysis of the causes and 
 in the development of control methods. The definition is as 
 follows: 
 
 Cutter roof failure in mine roof rock is a failure 
 process that initially begins as a fracture plane in the roof 
 rock parallel to, and located at, the roof-rib intersection. 
 The fracture propagates upward into the roof over the 
 mine opening at an angle usually steeper than 60° from 
 the horizontal. 
 
 The mappable extent of cutter failure may range in 
 length from only a few feet to several hundred feet and 
 may traverse entry-crosscut intersections without change 
 in direction. Once initiated along the roof-rib line, a cutter 
 may propagate away from the roof-rib line of one side of a 
 room, cross the roof span, and continue along the other 
 side of the room. It is important to note that this definition 
 excludes a similar type of failure (sometimes referred to as 
 "kink roof) that has comparable characteristics but 
 occurs in the center of the entry. (The reasons for this 
 distinction are discussed in the "Background" section.) 
 Researchers refer to the cutter failure discussed in this 
 report as "classic" cutter roof failure; except for short 
 deviations across intersections and entries, initial failure 
 is confined to the roof-rib intersection. 
 
 The cutter failure sequence is illustrated in figure 1. 
 An in-mine example of figures lA and IB can be seen in 
 figure 2, which shows roof rock in a crosscut as viewed 
 from an intersecting entry. The initial vertical fracture 
 propagated from the roof-rib line until it reached a weak 
 bedding plane. At that point, cribbing was installed to 
 prevent collapse. Figure 3 is another example of cutter 
 failure as it may appear when the vertical fracture 
 propagates to a much higher level in the roof If adequate 
 support is not installed and failure is allowed to progress, 
 massive roof failure will ensue, as shown in figure 4, 
 
 Cutter roof failure has long been known to be most 
 prevalent in the Appalachian Coalfields. However, 
 ongoing reconnaissance research by the Bureau is 
 revealing that cutter roof failure occurs in each of the 
 major coal basins of the United States where underground 
 mining is practiced. In each case, its occurrence may 
 appear to be unpredictable. An entire mine may be 
 experiencing ideal ground control conditions when sud- 
 denly a working section encounters this failure with 
 seemingly no explanation. For other mines, cutter failure 
 is a chronic condition, and when conditions become only 
 
 -7 Roof bolts^ 
 
 V - Direction of cutter 
 \ (fracture) propagation 
 
 ^' ---Nearly vertical crack 
 ^ (fracture)^^ . 
 
 Entry or crosscut 
 
 Underclay 
 
 A INITIAL CUTTER FAILURE 
 
 Fracture propagating across 
 the entry or crosscut at 
 anchorage horizon (bed 
 separation occurs at this 
 weak zone) 
 
 B, PROPAGATION OF CUTTER ALONG WEAK ZONE 
 
 Undercloy 
 
 C, CUTTER FALL 
 
 Figure 1. — Cutter roof failure sequence. Not to scale. 
 
 slightly worse than normal, entire working sections may 
 have to be abandoned, resulting in the sterilization of 
 large tracts of coal reserves. In either case, control of 
 cutter roof failure is no easy task. Simply changing bolt 
 length or placing a crib in a strategic location may have no 
 effect at all. Thus, based on the nature of cutter roof 
 failure, and the problems with trying to control its 
 occurrence. Three components of the issue become evident 
 and are the subject of this report: (1) causes, (2) prediction, 
 and (3) control. 
 
Figure 2.— Initial propagation of cutter roof failure to a weak bedding plane In roof rock above a crosscut. 
 
 Figure 3. — Severe cutter roof failure supported by fiber cribbing. 
 
Figure 4. — Massive roof fall resulting from unsupported cutter roof failure. 
 
 BACKGROUND 
 
 Theoretical explanations of the causes of cutter roof 
 failure exist; however, in-mine verification of these 
 theories through instrumentation and mapping have been 
 far from comprehensive. The publications discussed in 
 this section are representative of the present understand- 
 ing of cutter roof failure and applicable control measures, 
 and each of the concepts discussed here will be developed 
 further in following sections. 
 
 In 1948, R. W. Roley {43Y published an article with 
 reference to a particular type of roof failure, which he 
 referred to as "pressure-cutting." Many of the charateris- 
 tics that Roley described are the same ones used to 
 describe the current term "cutter roof failure." Although 
 cutter roof failure has been found to be most prevalent in 
 the Appalachian Coalfields, Roley originally described 
 this type of failure in the Illinois Basin. For an 
 explanation of its occurrence, he borrowed from D. W. 
 Phillips {41 ), who cited abnormally high lateral pressures 
 or stress conditions, of regional magnitude, causing shear 
 forces in the roof rock at the rib line to exceed the shear 
 strength of the rock. 
 
 Rock type as an important factor in the development 
 of cutter roof failure was discussed by Thomas (45) in 
 1950. He observed that the immediate roof rock type must 
 be competent with respect to the overlying roof rock; 
 otherwise, failure will occur in the center of the entry. In 
 addition, Thomas promoted the use of roof bolting as 
 
 ^Italicized numbers in parentheses refer to items in the list of references 
 at the end of this report. 
 
 opposed to timbering for successful support of the roof, and 
 proposed angle bolting as an effective control measure, 
 borrowed from the lead and zinc mining industry (46). 
 Thomas identified rock tjrpe as a cause for localized 
 occurrences of cutter roof failure. For local occurrences 
 where no change in roof rock type was found, a possible 
 cause was suggested in 1961 when Lang (28) demon- 
 strated that stress concentrations exist beneath stream 
 valleys creating an unstable environment for mining. 
 
 A first step toward assessing the influence of high 
 lateral stress and roof rock competency on the formation of 
 cutter roof failure was taken by Wang (47-48), using 
 two-dimensional, finite-element analysis. His results 
 mirrored observations that had been made in the field; 
 with this correlation, it was suspected that regionally 
 high horizontal stress and stress induced by overlying 
 stream valleys and overlying structural features such as 
 paleochennels were causing failure. With the aid of 
 computer modeling, he analyzed the effect of pillar 
 strength on the distribution of stress in the upper corners 
 of the entry, with a result that pointed at a possible 
 method of control. Wang found that by weakening the 
 pillar to some specific depth from the pillar skin, shear 
 stresses in these critical areas were reduced and the 
 threat of cutter failure could be lessened. A limited 
 amount of in-mine testing produced inconclusive results 
 (34). 
 
 In-mine verification of regionally high lateral stress 
 was necessary to determine its influence on the propaga- 
 tion of cutter roof failure. Agapito (IT and Aggson (2-4) 
 
both reported on an investigation conducted in several 
 mines of southern West Virginia. The mines were located 
 in the Beckley Coalbed, and overcoring techniques were 
 employed to determine the horizontal in situ stress. The 
 measurements were found to be much greater in 
 magnitude than would be expected if the horizontal stress 
 was attributable only to the Poisson ratio effect of the 
 overburden. A two-dimensional, finite-element model 
 representative of the minesites was analyzed, and a close 
 correlation was found between in-mine observations of 
 cutter roof failure and the failure modeled by the 
 computer. Three important findings were made through 
 this method of in-mine verification: 
 
 1. The horizontal stress field in the region was found 
 to be relatively uniform in magnitude and orientation. 
 
 2. The angle of the failure surface with the horizontal 
 is dependent, at least partly, upon the relative magni- 
 tudes of the horizontal and vertical stresses. 
 
 3. Pillars designed to completely yield reduce roof 
 stresses by 15 pet. 
 
 Each of these significant findings is elaborated upon later 
 in the text. 
 
 Kripakov (27) conducted a similar study. Using 
 in-mine, in situ stress measurements and two- 
 dimensional, finite-element analysis, he assessed the 
 applicability of Wang's pillar-softening concepts (34, 47). 
 According to the model, it was found that, for the 
 particular stress state of the mine, the pillar-softening 
 concept showed promise, provided certain modifications 
 were made. Kripakov (27) recommended that the longwall 
 method of mining be used in areas of severe cutter roof 
 failure, which would allow for less exposed roof area in 
 
 need of support and for the possible use of sacrifice entries 
 on advance of panels. His recommendations also included 
 continued use of truss bolting, which has also been 
 recommended by others (5, 19, 23, 29-30). 
 
 These in-mine investigations and subsequent analy- 
 ses accounted for only the immediate rock properties and 
 minor rock structure such as bedding. Although Wang 
 conducted computer analysis of the influence of 
 paleochannels, in-mine stress analysis of these features 
 has not been conducted. The influenc^e of geologic 
 structures, such as the presence of clastic dikes, (clay 
 veins), has recently been explored by lannacchione (23) 
 and Hill (19). Through detailed in-mine geologic mapping, 
 these two independent studies found that the presence of 
 clastic dikes contributed to the instability of roof rock, 
 resulting in the initiation of cutter roof failure. Although 
 only three mines were considered in these investigations, 
 it is highly probable that clastic dikes influence the 
 formation of cutter failure in many other mines, and 
 geology should certainly not be ignored. The contribution 
 of clastic dikes to the failure mechanism is discussed in 
 the section "Minor Geologic Structures." 
 
 With respect to current research in cutter roof failure, 
 mine operators are reemploying techniques of sacrifice 
 entries, yielding pillars, and reorientation of headings in 
 an attempt to solve their cutter failure problems. This 
 trend in the industry toward alternative control measures 
 shows promise for success. The following explanations for 
 the causes of cutter failure and methods for control 
 provide a systematic approach to aid the operator in 
 determining which technique is best suited to a particular 
 situation. 
 
 FACTORS IN THE PROPAGATION OF CUTTER ROOF FAILURE 
 
 Prior to applying various control measures to a 
 particular ground control problem, the mechanisms 
 responsible for its initiation and subsequent propagation 
 should be understood. The problem of cutter roof failure is 
 an excellent example of this, because many trial and error 
 methods have been unsuccessful in controlling it. Two 
 interdependent variables, cited in the preceding section, 
 are critical factors in the formation of cutter roof failure: 
 the stress environment and rock type. The stress 
 environment is further influenced by the addition of 
 adjacent entries, the introduction of forward and lateral 
 abutment pressures induced by retreat mining, and the 
 stresses created by the simultaneous extraction of 
 multiple seams. An additional variable, equally impor- 
 tant, is that of opening dimensions, which directly affect 
 the distribution of stress concentrations around the 
 periphery of the opening. 
 
 To help the reader understand how each of these 
 variables contributes individually and corporately to the 
 formation of cutter roof failure, this section is organized 
 on the basis of increasing complexity. Initially, the rock 
 mass is modeled as a homogenous, isotropic, linearly 
 elastic medium, with only a single opening. This is not an 
 accurate representation of the actual mining environ- 
 ment, and it is not designed to give any specific indication 
 of the characteristics of failure, but it does provide a 
 simplified means of demonstrating how stress concentra- 
 tions develop along the periphery of mine openings. Once 
 the concept of stress concentrations around mine openings 
 
 has been introduced, the two most important variables in 
 cutter roof failure formation are discussed in detail, the in 
 situ stress environment and rock mass characteristics. 
 
 OPENINGS IN A ROCK MASS 
 
 Once an opening is created in a rock mass, stress 
 concentrations form around the periphery of the opening, 
 with locations and magnitudes that are a function of the 
 shape of the opening, the stress environment (in situ 
 stress), and the characteristics of the rock mass (including 
 the mechanical properties of the rock). If the in situ stress 
 is a function of gravitational stresses alone (excluding 
 shear effects and idealizing the medium as a 
 homogeneous, isotropic, elastic field), the stress magni- 
 tudes can be roughly estimated if the surface topography 
 above the area is relatively flat. In this case, the vertical 
 stress is a result of the weight of the overburden and is 
 calculated as shown in equation 1. 
 
 a, = dD, (1) 
 
 where u,, = vertical stress, lb/ft', 
 
 d = average density of overburden, lb/ft\ 
 and D = vertical depth, ft. 
 
 The horizontal stress is also simplified if plain strain 
 conditions area enforced, thereby becoming a function of 
 
the vertical stress and Poisson ratio for the given rock 
 type, which reduces to equation 2, after Phillips {41). 
 
 (2) 
 
 O'h - (Tv > 
 
 1-v 
 
 where ah = horizontal stress, Ih/ft', 
 
 a^ = vertical stress, Ib/ft^ 
 and V = Poisson ratio for the rock, unitless. 
 
 For this discussion of the influence of opening 
 dimensions on the formation of stress concentrations 
 around the periphery of mine openings, the in situ stress 
 state results from gravitational effects. Isotropic material 
 and the effects of linearly elastic properties can be 
 assumed, and the deformation of the rock may be analyzed 
 as a plain strain system. Since the vast majority of coal 
 mine entries in the United States are rectangular in cross 
 section, the basic rectangular shape is the only shape 
 considered in this discussion. It is shown later in the 
 section "Mine Design Changes," that changes from 
 rectangular to other shapes may prove to be an effective 
 control measure. 
 
 Variations within the fundamental rectangular shape 
 can have an effect on stress concentrations along the 
 periphery of an opening for a given stress environment, 
 thus influencing failure propagation. For this reason, 
 isolated single openings will be considered first. Later, the 
 interaction of multiple-entry configurations both in 
 single and multiple-seam scenarios will be discussed. 
 
 Single Openings 
 
 Figure 5 illustrates an estimation of the stress 
 distribution around the periphery of rectangular openings 
 having width-to-height (W/H) ratios ranging in value 
 from 1 to 3. The figure was constructed from a 
 finite-element analysis using the conditions outlined at 
 the beginning of this section. In general, the models can be 
 interpreted by taking particular note of the areas that 
 have the greatest density of contour lines. These areas of 
 high stress concentrations are the most probable areas for 
 failure. The stress distribution is illustrated by two 
 different contour lines: (Da solid line representing equal 
 values of the ratio of the maximum secondary principal 
 stress, at the points through which the contour passes, to 
 the maximum stress applied to the model and (2) a dashed 
 line representing the ratio of the minimum secondary 
 principal stress to the maximum stress applied to the 
 model. 
 
 It will be shown that the location and orientation of 
 the failure plane is actually a function of the shear stress 
 that develops as a result of the difference between the 
 principal stresses. The numbers in figure 5 at the entry 
 opening corners, and midspan of roof and floor are 
 numerical approximations of the ratio of maximum 
 secondary principal stress for that immediate area of the 
 model to the maximum applied stress. The values 
 calculated for these areas are not to be directly applied to 
 the in-mine environment but are provided only to give an 
 indication of the increasing magnitude of the stress 
 concentrations as the shape of the opening changes. 
 
 Figure 5 demonstrates that for increasing W/H ratios, 
 the stress increases in the corners of the entry, while the 
 midspan is essentially under no stress. The opening most 
 representative of presently operating underground coal 
 mines is shown in figure 5C with a W/H ratio of 3 
 
 -(Th 
 
 A. -p- = l 
 
 — <r\y 
 
 "<rh 
 
 Figure 5. — Stress concentration diagrams of rectangular 
 openings with width-to-height ratios of 1 (A), 2 (B), and 3 (C) 
 under gravitational loading conditions. (Values are ratios, 
 representing the major principal stress, for the critical point, 
 divided by the maximum stress applied to the model.) 
 
(although many mines have higher W/H ratios). However, 
 it is important to note that the stress concentrations 
 around an actual coal mine entry will be influenced by the 
 rock properties and stress environment. For the opening 
 modeled here, the central roof portion of the entry 
 approaches a tension value of stress, and both the central 
 rib area and entry corners are subjected to compressive 
 stress in excess of the applied vertical stress. Several 
 researchers have demonstrated similar characteristics of 
 openings under specific load conditions, using other 
 techniques (20, 39). 
 
 Multiple Openings 
 
 Since the excavation of a single opening redefines the 
 state of stress in a rock mass, it follows that the ex- 
 cavation of more than one opening further redistributes 
 stresses. Such a redistribution of stresses occurs in (1) the 
 mining of multiple adjacent entries in a single seam and 
 (2) multiple-seam mining. With the first case, many mines 
 that drive several parallel entries experience cutter roof 
 failure in entries adjacent to the solid coal. The cause of 
 this failure has often been attributed to so-called 
 abutment pressures {36). Figure 6 is a finite-element 
 analysis of six openings with a W/H ratio of 3, using the 
 conditions previously described. The analysis demon- 
 strates that the stresses in the upper corners of an entry 
 are greatest when that entry is in the center of the panel. 
 Theories on the use of pillars for support of roof calculate 
 similar results but the in situ stress can vary from the 
 ideal case of gravitational loading alone, which can alter 
 the location of stress concentrations around openings. In 
 most cases, abnormally high in situ stress is thought to be 
 controlling the occurrence of failure in these outer entries. 
 Under the second scenario, many mines in a 
 multiple-seam configuration have experienced cutter roof 
 failure. The most common occurrence has been in the 
 overlying workings when entries being driven over the 
 solid suddenly cross over entries in the lower seam or a 
 gob area. Several researchers (16-18, 40) have concluded 
 that superimposing of mine workings (i.e., pillar directly 
 above pillar, entry directly above entry, crosscut directly 
 above crosscut) decreases the stress concentrations around 
 the openings and consequently the risk of failure. 
 Although this recommendaton has improved mining 
 conditions, it too has limitations. In practice it was found 
 that even when workings were superimposed, for inter- 
 burden of less than 110 ft (in the Appalachian region) (18), 
 stress concentrations developed causing roof instability. 
 
 The analyses of stress distributions around the 
 periphery of mine openings generally indicate that with 
 the enlargement of the width dimension of an opening the 
 magnitude of the stress concentrations increases. Addi- 
 tionally, with the introduction of adjacent openings the 
 magnitude of the stress concentrations increases even 
 more. 
 
 STRESS ENVIRONMENT 
 
 Prior to the excavation of an opening in a rock mass, 
 states of stress exist in the rock, which are functions of 
 gravitational and tectonic forces, thermal stresses, gas 
 pressures, and material and rheologic properties of the 
 strata. In the United States, in situ thermal stresses have 
 negligible influence on the stresses experienced in coal 
 mining, and modeling has shown that gas pressures. 
 
 -6 15 /3.44 \ -5.54 ; 3.08 , -6.36 -5.54 344_;-6.l5 
 r^^ -6.56^ 
 
 Figure 6. — Stress concentration diagram of six adjacent 
 openings under gravitational loading conditions. (Values are 
 ratios, representing the major principal stress, for the critical 
 point, divided by the maximum stress applied to the model.) 
 
 likewise, have little effect on the formation of cutter 
 failure (44). However, forces of tectonic origin and forces 
 that are a function of differential loading (as in the stress 
 state beneath stream valleys) have significant influence 
 on the stability of underground openings. In either case, 
 the stress environment surrounding the underground 
 excavation is such that the magnitude of the horizontal 
 stress is some value greater than what would be 
 calculated on the assumption of uniform gravitational 
 loading alone. 
 
 A simple analysis of the influence of this increased 
 value of horizontal stress is illustrated in figure 7, which 
 shows the results of a finite-element model of an opening 
 with a W/H ratio of 3, in a two-dimensional, linearly 
 elastic, isotropic, homogeneous substance, subjected to an 
 increasing ratio of horizontal to vertical stress (cth/o'v)- It 
 should be recognized immediately that this is the worst 
 case scenario from figure 5; i.e., under gravitational 
 loading alone, the opening with a W/H ratio of 3 had the 
 least desired stress concentrations of the three openings 
 presented in figure 5 (although the model still does not 
 directly represent the actual in-mine conditions, owing to 
 the absence of realistic rock properties). A familiar trend 
 is also seen in figure 7, in that as the horizontal-to- 
 vertical-stress ratio increases (from 7A to 7C) the 
 magnitude of the major principal stress concentration 
 along the upper corners of the entry likewise increases. 
 
 Aggson (4) conducted an analysis of the influence of 
 the ratio of horizontal to vertical stress, using actual rock 
 properties and calculating the maximum shear stress in 
 various models to determine the characteristics of failure. 
 He calculated the maximum shear stress because cutter 
 roof failure most likely initiates when this value exceeds 
 the shear strength of the immediate roof strata. Figure 8 
 qualitatively demonstrates his results, showing the effect 
 of different vertical-to-horizontal-stress ratios on the 
 angle the failure plane makes with the vertical. The 
 orientation of the predicted shear fracture rotates out over 
 the opening as the horizontal stress component increases. 
 Kripakov (27) conducted similar analyses, modeling the 
 mine entries of the Kitt Mine of northern West Virginia 
 using in situ stress measurements and actual rock 
 properties. Figure 9 illustrates the general conditions of 
 Kripakov's model and the manner in which the different 
 stress values are resolved at the corner of the entry. The 
 elemental components of the shear stress are resolved 
 from one-half the difference between the two principal 
 stresses at any point in the roof. Table 1 lists the various 
 values corresponding to figure 9 for different entry widths 
 
; — o-h 
 
 M M I I M n I M I M 
 
 — o-h 
 
 Figure 7. — Stress concentration diagrams of rectangular 
 openings with width-to-height ratio of 3 and loading conditions 
 of horlzontal-to-vertical-stress ratios of 1 {A), 2 (B) and 3 (C). 
 (Values are ratios, representing the major principal stress, for 
 the critical point, divided by the maximum stress applied to the 
 model.) 
 
 and other changes to the basic entry shape, which are 
 discussed in the section "Mine Design Changes." 
 
 In each of the models developed for this discussion of 
 stress concentrations around mine openings, and in the 
 models used by Aggson and Kripakov, the rock mass is a 
 continuum without separation along bedding planes. 
 However, although the in situ stress plays a major role in 
 the propagation of cutter roof failure, the characteristics 
 of the rock mass also contribute to the end result and are 
 necessary input for assessing the formation of cutter roof 
 
 o-h- 
 
 Plane of maximum shear [a\^<ay) 
 Poisson effect ^ , 
 
 /-Tension fracture ' 
 
 rCooT 
 
 A , HORIZONTAL STRESS LESS THAN VERTICAL STRESS 
 
 o-h- 
 
 Plone of maximum shear — ~- — 
 
 (ah^cTy' — ^ --■ — 
 
 Hydrostatic — : _ 
 
 B , HORIZONTAL STRESS EQUAL TO VERTICAL STRESS 
 
 C7y 
 
 I I I I I I I it I t M 
 
 Plane of maximum shear 
 
 (cTh > CTy ) 
 
 [Coal 
 
 o-h 
 
 C, HORIZONTAL STRESS GREATER THAN VERTICAL STRESS 
 
 Figure 8. — Qualitative example of influence of horizontal-to- 
 vertical-stress ratio on angle of failure propagation. (Courtesy N. 
 P. Kripakov) 
 
 Plane of max 
 shear stress 
 
 Figure 9. — Analysis of elemental components of angle of 
 failure as related to stress environment and rock properties. 
 [Adapted from Kripakov (27)] 
 
 failure. Rock properties are discussed later, in the section 
 "Rock Mass Characteristics." 
 
 Regional Stresses 
 
 Basically, there are two stress cases that have a direct 
 influence on the formation of cutter roof failure in coal 
 mines: tectonic stress and differential gravitational stress, 
 both of which can often be identified by recognizing 
 patterns of failure from a minewide perspective. Control 
 measures and evasive measured are different for each 
 case. Stresses that are a result of tectonic forces often 
 display characteristics that suggegt regional influence and 
 are discussed as regional stresses. Stresses that are a 
 
Table 1. — Maximum shear stress values and analysis of failure 
 
 Condition 
 
 Stress (cr), psi 
 
 IWaximum 
 
 Minimum 
 
 Angle from horizontal 
 
 to maximum principal 
 
 stress (28), deg 
 
 Maximum 
 
 shear. 
 
 psi 
 
 Angle from vertical 
 to plane of maximum 
 shear stress (if), deg 
 
 1,120 
 1,099 
 1,022 
 
 
 19.52 
 19.72 
 21.45 
 
 919 
 744 
 693 
 
 
 41.57 
 44.47 
 44.38 
 
 1,109 
 
 1,055 
 
 975 
 
 
 20.54 
 24.41 
 28,42 
 
 As mined: 
 
 17-ft entry width -930 
 
 16-ft entry width -924 
 
 Pillar softening, 3-tt depth at roof line . . -814 
 
 Slot at roof line: 
 
 3 by 12 in 68 
 
 3 by 24 in 11 
 
 3 by 36 in 42 
 
 Slot at pillar midheight: 
 
 3 by 12 in -868 
 
 3 by 24 in -638 
 
 3 by 36 in ^19 
 
 Source: Adapted from Kripakov (27). 
 
 -3,170 
 -3,121 
 -2,857 
 
 -1 ,770 
 -1 ,476 
 -1 ,343 
 
 -3,085 
 -2,748 
 -2,368 
 
 64.52 
 64.72 
 66.45 
 
 86.57 
 89.47 
 89.38 
 
 65.54 
 69.41 
 73.42 
 
 result of differential gravitational loading are discussed 
 under the heading "Stress Concentrations Beneath 
 Stream Valleys." 
 
 Stresses of tremendous magnitude have been known 
 to exist in the Earth's crust ever since the earliest 
 geologists recognized that mountains are the result of 
 massive regional uplift. The fact that the horizontal 
 component of these stresses is rather uniform with respect 
 to orientation, over vast areas of continents, can be easily 
 demonstrated by the curvilinear trend of the Appalachian 
 Mountains or most other mountain chains. In coal mines 
 of the United States, the influence of these regionally high 
 in situ horizontal stresses has been recognized since at 
 least the 1930's (42) and is currently the focus of much 
 research. 
 
 Probably the most conspicuous and most often cited 
 sjrmptom of regionally high horizontal stresses in coal 
 mines is that cutter roof failure occurs most frequently in 
 one particular orientation (i.e., in either the main 
 headings or the crosscuts but not both). References to this 
 phenomenon in the United States often make note of the 
 fact that this type of unidirectional roof failure occurs 
 mainly in northerly oriented headings (7-8, 25, 29). In 
 many cases, it has been found that this northerly direction 
 is subperpendicular to the major principal horizontal in 
 situ stress. Figure 10 demonstrates the regional presence 
 of excessively high horizontal in situ stress across the 
 United States as compiled by Aggson (2). Table 2 gives the 
 actual values of the stress for the various locations. Of 
 particular interest is the fact that most of the readings 
 revealed a horizontal biaxial stress field with a maximum 
 principal stress greater in magnitude than would be 
 calculated in the usual manner (equation 2). [Engelder 
 (13) has compiled more recent in situ stress data for the 
 northeastern United States.] 
 
 Figure 11 is a map of a portion of a mine in western 
 Pennsylvania showing the locations of roof falls. The fact 
 that the majority of roof falls occur along the same 
 orientation suggests the influence of regionally high 
 horizontal stresses where the difference in magnitude 
 between the minimum and maximum secondary principal 
 stresses is relatively large. Recognition of patterns of roof 
 failure such as this can aid in determining whether the 
 failure is a local or regional phenomenon and can 
 eventually aid in selecting a control measure. 
 
 In other instances where the in situ horizontal stress 
 state is greater than the expected value from gravitation- 
 al loading, there is little difference between the magni- 
 tudes of the principal horizontal stresses. An example of 
 this was reported on separately by Aggson (3) and 
 lannacchione (23). Aggson conducted in situ horizontal 
 
 Sell*, pal 
 
 Figure 10. — Horizontal in situ compressive stress measure- 
 ments in the United States. Stress fields are drawn to pressure 
 scale (in legend), not to mileage scale. [Adapted from Aggson (2)] 
 
 stress measurements by overcoring and found the horizon- 
 tal stress to be at least three times the vertical stress at 
 the Kitt Mine in northern West Virginia. However, only a 
 22-pct difference was found between the minor and major 
 principal horizontal in situ stresses. lannacchione found 
 little difference in the frequency of failure in crosscuts 
 versus entries, through in-mine mapping that reconfirmed 
 that the orientation of the entries was not influencing the 
 formation of failure. Kripakov (27) also reported on cutter 
 roof failure at the Kitt Mine. His research indicated that 
 modification to the pillars, such as cutting slots in the 
 pillar near the roof, would be necessary to effectively 
 control the cutter problem (discussed further in the 
 section "Mine Design Changes"). In short, it is very 
 difficult to control cutter roof failure when the cause is 
 regionally high horizontal stresses and especially when 
 there is little difference in magnitude between the 
 maximum and minimum horizontal stresses. 
 
 Stress Concentration Beneath Stream Valleys 
 
 Many mines that normally have relatively good 
 ground conditions occasionally encounter severe condi- 
 tions in the form of cutter roof failure. These localized 
 occurrences can be the result of stress concentrations 
 beneath stream valleys, mining-induced stresses, or rock 
 type. In addition, localized high-stress conditions exist in 
 
10 
 
 Table 2. — Horizontal in situ stress measurements, surface and underground sites 
 
 Site' 
 
 Location 
 
 Direction 
 of P 
 
 Magnitude 
 of P, psi 
 
 (vlagnitude 
 
 Average deptli 
 
 of Q. psi 
 
 of measurement^ft 
 
 941 
 
 18.1 
 
 285 
 
 1.8 
 
 1,191 
 
 33 
 
 1.385 
 
 8.6 
 
 335 
 
 4.8 
 
 1,113 
 
 61.9 
 
 516 
 
 1.2 
 
 791 
 
 151.2 
 
 1,397 
 
 4.7 
 
 1,519 
 
 4.9 
 
 777 
 
 4 
 
 519 
 
 4.5 
 
 1,491 
 
 4.7 
 
 171 
 
 10 
 
 551 
 
 925 
 
 2,500 
 
 2,300 
 
 2,937 
 
 3,200 
 
 1,595 
 
 1,000 
 
 1,053 
 
 6,200 
 
 4,966 
 
 5,300 
 
 2,283 
 
 3,127 
 
 2,898 
 
 1,060 
 
 404 
 
 1,600 
 
 705 
 
 850 
 
 345 
 
 1,250 
 
 160 
 
 1,570 
 
 1,466 
 
 700 
 
 Surface: 
 
 1. .. 
 
 2. .. 
 
 3. .. 
 
 4. .. 
 
 5. .. 
 
 9. 
 10. 
 11. 
 12. 
 13. 
 14. 
 
 Lithonia, GA 
 
 Dougiasville, GA 
 
 Mt. Airy, NC 
 
 Rapidan, VA 
 
 St. Peters, PA 
 
 West Cfielnesford, IVIA 
 
 Proctor, VT 
 
 Barre, VT 
 
 Graniteville, IVIO 
 
 St. Cloud, (WIN 
 
 Carthage, MO 
 
 Troy, OK 
 
 Marble Falls, TX 
 
 Green River, WY 
 
 Underground: 
 
 15 Immel Mine, Knoxville, TN 
 
 16 Limestone Mine, Barberton, OH 
 
 17 Mather Mine, Ishpeming, Ml 
 
 18 Fletcher Mine, Bunker, MO 
 
 19 Homestake Mine, Lead, SD 
 
 20 Crescent Mine, Wallace, ID 
 
 21 Henderson Mine, Empire, CO 
 
 22 Sunnyside. Mine, Sunnyside, UT 
 
 23 Allied Chemical Mine, Green River, WY 
 
 24 Big Island Mine, Green River, WY 
 
 25 Rainier Mesa, NV, test site 
 
 26 Lakeshore Mine, Casa Grande, AR 
 
 27 Beckiey No. 1 Mine, Bolt, WV 
 
 49° E 
 64» W 
 87° E 
 
 6° E 
 14° E 
 56° E 
 
 4° W 
 14° E 
 77° E 
 50° E 
 
 2° E 
 84° W 
 33° W 
 42° E 
 
 N 58° E 
 
 N 77° E 
 
 N 82° W 
 
 N 17° W 
 
 N 38 
 
 N 27 
 
 N 15' 
 
 N 31' 
 
 N 23' 
 
 N 38' 
 
 N 46° W 
 
 N 69° E 
 
 N 69° E 
 
 1,639 
 
 512 
 
 2,464 
 
 1,678 
 
 820 
 
 2,133 
 
 1,328 
 
 1,734 
 
 3,190 
 
 2,205 
 
 1,066 
 
 1,075 
 
 2,219 
 
 415 
 
 3,007 
 4,000 
 3,822 
 3,682 
 2,778 
 6,258 
 3,398 
 3,718 
 1,781 
 1,054 
 972 
 502 
 2,973 
 
 'Numbers correspond to those on figure 10. 
 
 NOTE. — P and O are the maximum and minimum secondary principal stresses. 
 Source: Adapted from Aggson (2). 
 
 LEGEND 
 
 ' Roof falls 
 
 
 
 L 
 
 800 
 
 Scale, ft 
 
 Figure 1 1 . — Partial mine map from western Pennsylvania, showing preferential orientation of roof falls. 
 
11 
 
 Streams 
 
 Figure 12.— Mine map from mine in southern. West Virginia, siiowing correlation of roof fall locations with centers of overlying 
 stream valleys. 
 
 areas of close proximity to anomalous geologic structures 
 such as paleochannels, rolls, and clastic dikes. For the 
 purpose of emphasis, the only topic discussed in this 
 section is the effect of surface topography on underground 
 opening stability. The influence of anomalous geologic 
 structures is discussed under the heading "Minor Geologic 
 Structures." 
 
 Figure 12 is a map of a mine in southern West 
 Virginia showing roof falls and overburden. The pattern of 
 roof falls beneath the stream valleys is typical of 
 valley-stress-induced failure. All other factors in the mine 
 remained relatively constant, such as rock type and 
 opening dimensions; thus, differential gravitational stress- 
 es were deduced as the cause of failure. For many other 
 
12 
 
 mines, similar correlations can be found between stream 
 valleys and roof failure, although failure may occur in 
 wider or narrower zones throughout the mine. In still 
 other cases, mining beneath a valley may create no 
 adverse mining conditions. From the limited amount of 
 research conducted on this phenomenon, the following 
 variables are thought to influence the occurrence of cutter 
 roof failure associated with stream valleys: rock type, 
 opening dimensions, percent extraction, depth beneath 
 the valley, relief of surface topography, gradient of valley 
 walls, availability of flowing water, and magnitudes and 
 orientation of the principal in situ stresses that influence 
 differential gravitational loading. It is also important to 
 note that, depending on the variables just listed, failure 
 may occur in ways other than cutter roof 
 
 It has long been known that mining under low cover 
 beneath stream valleys creates a potential for bad ground 
 conditions. Little in-mine research has been conducted to 
 analyze the cause, as a guide for developing control 
 measures, although it has been estimated that as much as 
 90 pet of all roof falls in the northern Appalachian Basin 
 occur beneath valleys (36-37). Ferguson (14-15) described 
 unstable ground conditions associated with valley stresses 
 in almost all surface engineering work within valleys in 
 the Appalachian Plateau region. Laboratory investiga- 
 tions have revealed that a stream valley can be modeled 
 as a V-notch in a horizontal, thick plate subjected to 
 gravitational loading. Lang (28) demonstrated the effect 
 of a stream valley on the in situ stress environment using 
 a V-notch in a photoelastic gelatin model, as did 
 Worotnicki (50), who also used an electric analog model. 
 These models showed that immediately beneath a 
 V-notch, and for a substantial depth (as much as 600 ft 
 beneath it), the horizontal stress is actually greater than 
 the vertical stress (compared with the case of uniform 
 gravitational loading beneath relatively flat topography). 
 If other stresses are applied, the stress concentrations are 
 modified by the notch. 
 
 Wang (48) used a two-dimensional, finite-element 
 model to compare the stress concentrations around an 
 opening beneath a stream valley with the stress concen- 
 trations around an opening beneath an adjacent hill. For 
 this analysis, the only applied stress was gravitational 
 loading, and the mine openings had a W/H ratio of 2. The 
 coalbed of the model was attributed a different modulus of 
 elasticity from the rest of the model to more accurately 
 represent the in situ environment. Figure 13 is a 
 comparison of the results for each of three sets of openings 
 as the height of the hill increased from 240 to 355 to 615 ft. 
 (The depth of the opening beneath the valley remained 
 constant at 115 ft.) The most important feature of this 
 graph is that the opening beneath the valley has higher 
 compressive stresses in the corners and lower tensile 
 stresses in the midspan of the roof than does the opening 
 beneath the hill. 
 
 Moebs [as cited by Enever (12)] conducted a study on 
 the frequency of roof falls as related to their lateral 
 distance from the center of valleys, compiling data from 
 several mines in the northern Appalachian Basin. Based 
 on his data, Moebs developed an empirical method of 
 determining the likelihood of roof instability associated 
 with a stream valley as related to the slope of the valley 
 walls and the depth of the mine below the valley. 
 Equation 3 represents his empirical method. 
 
 550 
 
 Pillar 
 compression 
 
 Pillar 
 compression 
 
 KEY 
 Opening under hill 
 
 Opening under valley 
 
 Roof corner 
 
 -Midspan 
 tension 
 
 Midspan 
 tension 
 
 Roof corner 
 ompression 
 
 300 
 
 500 
 
 700 
 
 900 
 
 HEIGHT OF HILL, ft 
 
 Figure 13. — Values of stress for critical points of an opening 
 beneath a valley versus an opening beneath a hill, for an 
 increasing height of the hill. [Adapted from Wang (48)] 
 
 where 
 
 and 
 
 F = an empirical index, unitless, 
 
 A = the mean slope angle of one of the valley 
 
 walls from the horizontal, deg, 
 A = the mean slope angle of the other valley 
 
 wall from the horizontal, deg, 
 D = the depth of the coalbed at the point of 
 
 interest, beneath the valley wall, ft. 
 
 F (fall factor) = ( A + A' ) 
 D 
 
 (3) 
 
 It can be seen from the equation that as the steepness of 
 the valley walls increases, the values of F correspondingly 
 increase, representing an increasing risk of roof falls. 
 However, the F value decreases for greater depths. 
 For the small set of mines analyzed by Moebs, a particular 
 value for F was obtained and a greater F value in any area 
 usually indicated imminent failure. Moebs did not 
 specifically identify this "critical" value since it has not 
 been shown to be universal in application. Another factor 
 discussed by Moebs in a separate publication (37) is the 
 fact that the highest frequency of falls at each mine 
 occurred beneath streams oriented in a northerly direc- 
 tion. As mentioned earlier, for the northern Appalachian 
 Basin this northerly direction is subperpendicular to the 
 major principal horizontal in situ stress. Thus, it is 
 assumed that stream valleys with a northerly orientation 
 increase the magnitude of the horizontal in situ stress. 
 
 This phenomenon of the influence of the stream 
 valleys on the already high horizontal stress field was also 
 seen in Australia by Enever (11), where in situ stress 
 measurements revealed not only an increase in magni- 
 
'13 
 
 tude but also a reorientation of the principal stresses. The 
 research showed that a ratio of depth of cover to the 
 maximum surface relief for a particular valley (D/R, 
 where D is the depth and R is the maximum surface relief 
 with compatible units) was the most reliable empirical 
 relationship for determining the likelihood of valley- 
 stress-induced failure of underground workings. The 
 value of D/R is calculated at any point within the valley to 
 determine the likelihood of failure beneath that point. 
 Enever concluded by suggesting that D/R ratios of less 
 than 0.5 indicate a strong possibility of encountering 
 adverse roof conditions, and that this ratio would need to 
 be adjusted in the presence of regionally high horizontal 
 stresses. 
 
 Both Moebs and Enever admit that their empirical 
 methods are unable to treat the in situ stress state as a 
 separate variable. Additionally, a second missing variable 
 may be one that would take into account drastic changes 
 in the cross-sectional profile of the valley as taken 
 perpendicular to the trend of the valley. It is important to 
 note that these empirical relations are based solely on 
 observations of physical conditions and do not include the 
 influence of the failure mechanisms. Further, neither of 
 the equations has been tested against a statistically 
 significant population of failure occurrences. For this 
 reason, these methods are presented only as a means of 
 demonstrating the problem to the reader. For predicting 
 failure, these relations would most likely need to be 
 adjusted from mine to mine; e.g., they should also include 
 any pertinent information concerning failure in the area 
 of question, such as variables of available water, rock 
 type, and geologic anomalies. 
 
 ROCK MASS CHARACTERISTICS 
 
 Rock mass characterization is a vast subject area 
 concerned with the comprehensive description of rock 
 masses. The characterization of rock mass has been 
 tackled by many researchers through the use of classifica- 
 tion systems [of which Bieniawski (6, pp. 97-132) has 
 listed the most popular presently being used]. Each of the 
 various systems attempts to classify a rock mass either 
 qualitatively or quantitatively into groups exhibiting 
 similar behavior. A number of parameters are utilized in 
 the various classification schemes (e.g., compressive 
 strength of the intact rock and joint spacing), with each 
 variable being weighted in terms of its overall importance 
 with respect to support. The descriptions and value of 
 these systems in mine design are not discussed in this 
 report; however, the two most common threads in these 
 systems are significant factors in the propagation of cutter 
 roof failure: rock properties (including elasticity) and 
 minor geologic structures (such as joints, clastic dikes, and 
 facies changes). 
 
 In the previous discussions of the effect of opening 
 dimensions and stress environment on cutter roof prop- 
 agation, all other variables were held constant at some 
 value of either a commonly encountered condition or 
 convenience with respect to modeling. In each case, it was 
 useful either to omit the variable of rock mass characteris- 
 tics by using a constant of purely elastic uniform 
 homogeneity, or to only partially incorporate it by using 
 the elastic properties of the rock. For coal measure rocks, 
 this is obviously not an accurate description. In this 
 discussion of rock mass characterization, the in situ stress 
 state is held constant at a value of gravitational loading, 
 and the opening dimensions have a W/H ratio of 3. 
 
 OO 
 
 cn >_ 
 
 COUJ 
 
 25 
 
 — Harder coal 
 
 Softer coal—* 
 
 1 1 
 
 1 I 
 
 20 
 
 - 1 c^ 
 
 1 1 
 
 1 1 
 
 I/I 
 
 1/2 1/4 1/6 1/8 1/10 
 
 RATIO OF MODULUS OF ELASTICITY OF COALBED |Ec) 
 TO MODULUS OF ELASTICITY OF ROOF ROCK (Epl 
 
 Figure 14. — Stress concentration in roof-rib corner versus 
 elasticity of coalbed. (Stress concentration values are ratios, 
 representing the major principal stress, for the critical point, 
 divided by the maximum stress applied to the model.) [Adapted 
 from Wang (47)] 
 
 However, as the influence of separate rock mass charac- 
 teristics is analyzed, it will become clear that the geologic 
 environment places limitations on opening dimensions 
 and at times creates local stress anomalies. 
 
 Rock Strength And Stiffness 
 
 Wang {48) used finite-element analysis to investigate 
 the effects of rock stiffness (or elasticity) on the stress 
 distribution around single mine openings. For the first 
 analysis, he looked at a single opening in a coal seam 
 bounded by a uniform shale above and below, from which 
 he concluded the following: 
 
 The stresses in the mine roof are highly 
 dependent on the relative values of the elastic 
 moduli of the roof materials closest to the surface 
 of the opening. Where the roof is a single 
 material, the compressive and shear stresses at 
 the roof-rib intersection tend to decrease and the 
 tensile stress at midspan of the roof tends to 
 increase as the roof material becomes stiffer 
 elastically in respect to the coal seam .... 
 
 The converse of this statement also holds true: As the coal 
 seam becomes stiffer elastically in respect to the roof 
 material, the compressive and shear stresses at the 
 roof-rib intersection tend to increase and the tensile stress 
 at midspan of the roof tends to decrease. Figure 14 
 illustrates how the magnitude of the stress concentration 
 changes as coal elasticity changes. 
 
 Wang went on to analyze single mine openings in 
 multilayered material, which had primarily been ana- 
 lyzed with beam equations in the past. Obert (39) used the 
 beam method of analysis to estimate limits of stable roof 
 spans across single and multiple entries in multilayered 
 material. The results of Wang's work, however, illustrate 
 with simplicity the qualitative interpretation of opening 
 stability for multilayered roof (48): 
 
 ... for a multicomponent roof, both the shear 
 and compressive stresses at the roof-rib intersec- 
 tion and the midspan tensile stress increase in 
 value when the elastically stiffer roof material is 
 closest to the surface of the opening. 
 
14 
 
 600 
 
 500 
 
 400 
 
 300 
 
 200 
 
 100 
 
 1 
 
 KEY 
 • Tensile strength 
 A Compressive strength 
 
 J I I 1 L 
 
 1/6 3/12 1/3 5/12 1/7 2/3 
 TI/(TI + T2) 
 
 5/6 
 
 900 
 
 800 
 
 700 
 
 600 
 
 500 
 
 400 
 
 Bock to in SITU 
 sheor stress vtilut 
 
 So: 
 
 O UJ 
 
 Of?, 
 
 Z3 W 
 
 Figure 15. — Stress values in roof-rib corner and midspan of 
 roof versus ratio of thickness of the two immediate roof 
 members. T^ Is the roof rock layer closest to the opening, and Tj 
 is the layer immediately above T,. [Adapted from Wang (48)] 
 
 Wang further defined the stability of openings in 
 multilayered material by analyzing the thickness of the 
 various roof members, the results of which are illustrated 
 in figure 15. The figure demonstrates that the threat of 
 cutter failure is greatest for thick, weak layers of rock that 
 overlie a strong immediate roof 
 
 When only rock strength is considered as the 
 controlling factor in the propagation of cutter roof failure, 
 the same worst case scenario can be drawn from both the 
 finite-element and beam methods of analysis. In 1950, 
 Thomas {45) reported the same findings as Wang's based 
 on underground observations of the cutter roof failure 
 process and described the worst case scenario as follows: 
 
 . . . the conditions necessary to produce a 
 "cutter" are: (DA relatively strong immediate 
 roof that may be thinly laminated, but the 
 cementation between the laminations must not 
 break down easily, and (2) a series of weaker 
 strata that tend to sag and slowly load the 
 immediate roof below it. 
 In addition, as the thickness of the strong immediate roof 
 member decreases, the effect of loading upon this member 
 by the overlying weaker member increases. 
 
 The work of Aggson (4) and Kripakov {27) is discussed 
 in the section "Stress Environment" with respect to the 
 orientation of the cutter fracture plane in the roof as a 
 function of the in situ stress. However, the actual 
 calculations for exact determination of the location and 
 orientation of the fracture plane for a particular in-mine 
 condition are extremely complex; the required data for 
 calculating these values are the elastic properties of the 
 rock, the stress concentrations around the periphery of the 
 opening (under the given in situ stress environment), and 
 the manner in which these stresses are redistributed as 
 the fracture propagates. This information is useful in 
 calculating the failed roof rock load on artificial support. 
 As an example, when trusses are used to support failed 
 roof it is beneficial to know how much dead weight is to be 
 suspended in order to use a sufficiently large gauge of 
 steel. 
 
 Minor Geologic Structures 
 
 A minor geologic structure inherent to coal measure 
 
 Figure 16. — Qualitative interpretation of cutter failure prop- 
 agation in thinly bedded, single roof rock type. 
 
 rocks, introduced in the discussion on rock strength and 
 stifftiess, is the naturally occurring bedding planes of 
 sedimentary rocks, which commonly separate rocks of 
 differing material properties. These discontinuities divide 
 the roof rock into separate beams, which allow for shear 
 displacement as the roof sags into the mine opening 
 (provided the rock does not sag beyond its elastic limit). 
 The fewer the number of bedding planes, the greater the 
 horizontal shear stiffness of the roof material. From 
 available beam equations, it has been shown that as the 
 shear stiffness of the roof increases, the predicted point of 
 failure moves closer to the ribline {24). Conversely, as the 
 number of bedding planes in the roof increases, the shear 
 stiffness decreases and the predicted point of failure 
 moves toward the center of the entry. Thus, according to 
 beam equations, if horizontal shear stiffness were the only 
 controlling factor in the occurrence of cutter roof failure 
 for a specific site, the worst case scenario would be a 
 massive roof rock unit devoid of bedding planes. However, 
 roof rock, of only one rock t5T)e, with many weak bedding 
 planes has also been observed to fail in the cutter manner 
 {29-30). 
 
 Figure 16 qualitatively demonstrates the possible 
 mechanisms behind the failure of thinly laminated rock 
 that does separate between bedding planes as a result of 
 differential deformation along individual beds. When the 
 mine opening is initially excavated, the maximum shear 
 stress is in the roof rock layer closest to the opening 
 (labeled 1 in figure 16); as a result, this unit undergoes 
 slightly more deformation than does the overlying layer, 
 causing a gap to form between the layers. The shear stress 
 causes this lowest layer to fail, and a redistribution of 
 stress occurs, creating an unstable environment for the 
 next layer up. This process continues until an equilibrium 
 is reached, usually in the form of massive roof failure. 
 This type of roof failure has been frequently observed 
 underground, and Aggson (2) has identified and described 
 a similar failure mechanism in floor heave. 
 
 Another minor geologic structure influence would 
 appear to be coal cleat. Thomas {45) indicated that cutter 
 roof failure occurred most frequently in entries oriented 
 parallel with the face cleat, implying some inherent 
 relation between cleat and the formation of cutter failure. 
 However, present trends in cutter roof failure do not show 
 preference to entries oriented parallel with the face cleat; 
 in fact, cutter failure generally occurs more frequently in 
 headings parallel with the butt cleat of the coal. In either 
 case, coal cleat has not been established as an instigator of 
 cutter roof failure. The only apparent influence it may 
 
15 
 
 Figure 17. — Clastic dii<e in coal pillar. Dilie is outlined with white chalk. 
 
 have results when coal is left as an immediate roof 
 member; in this case, cleat has an effect on the elasticity of 
 the immediate roof Discordant joints (not bedding planes) 
 in roof rock other than coal also appear to have little or no 
 effect on the formation of cutter roof failure. 
 
 Clastic dikes (or clay veins) have recently been cited 
 as significant contributors to the propagation of cutter roof 
 failure (19, 23). Figure 17 is an example of a clastic dike in 
 
 a coal pillar; the shape of these minor geologic structures 
 may vary significantly over the length of the dike, which 
 may or may not cut entirely through the coal from the roof 
 to the floor. The width may range from as thin as a 
 film-like trace to as thick as SVi ft or larger, and the 
 material that infills the dike ranges from claystone to clay 
 matrix with inclusions of coal, shale, sandstone, etc. The 
 dikes are frequently associated with a minor fault fracture 
 
16 
 
 Figure 18. — Clastic dike In midspan of roof rock of crosscut, showing how the roof is divided into two separate cantilever beams. 
 
 in the coal and slickensides in the roof, which have been 
 observed to extend as much as 25 ft above the coalbed. 
 Chase (9) discusses the various features and theories of 
 formation of clastic dikes with recommendations on 
 support methods. His work shows that clastic dikes are 
 associated with many forms of roof failure in addition to 
 the cutter type. Figure 18 is an example of the manner in 
 which a clastic dike can affect roof rock stability. In the 
 case shown, the clastic dike divides the roof span into two 
 separate cantilever beams. When the roof is left unsup- 
 ported, the shear stress in the corners of the entry 
 increases because of the moment imposed by the weight of 
 the unsupported beam. Depending upon the stress 
 environment and the rock type, the failure that results 
 may take on the form of cutter roof failure. 
 
 At the Kitt Mine, lannacchione (23) conducted 
 in-mine mapping of minor geologic structures and 
 deformation due to pressure release. He found that clastic 
 dikes frequently formed the boundaries of cutter roof falls 
 and otherwise destabilized roof conditions, allowing cutter 
 failure to form. In a study by Hill (19), geologic mapping 
 was conducted of the Greenwich North and South Mines 
 in Indiana County, PA, and cutter failure was observed to 
 form at the point where clastic dikes intersected the 
 ribline. And, as in lannacchione's work, clastic dikes were 
 also found to form the boundaries of many cutter roof falls. 
 Figure 19 is a portion of the mapping conducted at the 
 Greenwich North Mine, illustrating the high frequency of 
 occurrence of clastic dikes and their association with 
 cutter roof failure at this mine. 
 
 Figure 20 is a map of the Eastern United States, 
 showing each of the major coal basins and the distribution 
 of clastic dikes by their occurrence in mines. A survey of 
 
 mines currently being conducted throughout these basins 
 is revealing a similar distribution pattern of mines having 
 a high frequency of cutter roof failure. The conclusion 
 drawn from this correlation is not that clastic dikes are 
 the cause of cutter roof failure but rather that they act as a 
 discontinuity from which failure can initiate. 
 
 Other minor geologic structures that affect the 
 formation of cutter roof failure are paleochannels and rolls, 
 somewhat larger in size than clastic dikes. These 
 sedimentary and compactional features can be areas of 
 abrupt change in rock type and frequently influence stress 
 concentrations around nearby openings. Several geolo- 
 gists have described these features in detail, discussing 
 their effects on opening stability and suggesting methods 
 of support (21-22, 26, 35-36). Severe cutter roof failure 
 conditions have also been observed in crosscuts driven 
 subparallel to the strike of a roll in a mine in southern 
 West Virginia. Figure 21 is a map of the section of the 
 mine experiencing problems, and figure 22 illustrates the 
 attitude of the roll. In figure 21, note that failure was 
 probable where a crosscut was driven directly beneath and 
 parallel with the roll. 
 
 Wang (48) conducted finite-element analysis of rolls 
 overlying mine openings and found an increase in stress 
 concentrations at the corners of the entry. Additionally, 
 for entries directly beneath rolls, stress concentrations at 
 the roof midspan and corners were calculated for different 
 values of the pitch of the roll (or the angle the bedding 
 makes with the horizontal). Figure 23 is a graph of Wang's 
 results. 
 
 Geologic mapping of mines experiencing cutter roof 
 failure is imperative if the causes of failure are to be 
 discovered and controlled, but it has been conspicuous by 
 its absence in past investigations of cutter roof failure. 
 
17 
 
 LEGEND 
 Clastic dike ^:= Cutler 
 Fall area ^s Floor heave 
 
 Kettle bottom 
 
 Scale, 
 
 Figure 19. — Geologic and roof failure map of Main A of Greenwich North IVIine. 
 
 ^ LEGEND 
 
 J ^\J etc. Indicatestheoccurrenceof clay veins 
 and designates which coalbed is 
 
 ^ being extracted 
 
 [Zj,[Z],[Z],6'c. Indicates that clay veins were not 
 observed in extracted coolbed 
 
 
 « Desig- 
 nation 
 
 I 
 2. 
 
 3. 
 4 
 5. 
 6 
 7 
 8 
 9. 
 10 
 II 
 12 
 
 Mine location 
 Coal basin 
 Coalbed 
 
 Blue Creek 
 Campbell Creek 
 Chilton 
 Eagle 
 
 Eagle No. 2 
 Elkhorn, Lower 
 Freeport, Lower 
 Freeport, Upper 
 Harlan 
 Hortshorne 
 Hortshorne, Lower 
 Herrin 
 
 « Desig- 
 nation 
 
 13. 
 
 14. 
 
 15. 
 
 16. 
 
 17 
 
 18. 
 
 19. 
 20. 
 21 . 
 22 
 23. 
 24 
 
 Coalbed 
 
 High Splint 
 Kittanning, Lower 
 Kittanning, Upper 
 Pittsburgh 
 Pocahontas No. 3 . 
 Pocohontas No. 4 K 
 Pond Creek I 
 
 Powellton, Upper 
 Redstone 
 Sewell 
 Sewickley 
 Springfield 
 
 Figure 20. — Clastic dike in-mine occurrences in Eastern United States. (Courtesy F. E. Chase) 
 
18 
 
 g G D y □D^ODfl^ 
 
 g n D wouooacjo 
 p D ggDQ^^^3^ 
 
 EAGLE COALBEp 
 
 Mi 
 
 BBgR 
 
 ^i§ 
 
 gpaC 
 
 -"■-in 
 
 aaaDQQ 
 
 DDaa 
 
 I ^ 
 
 --""-f 
 
 f^= 
 
 / 
 
 LEGEND 
 .'' Sandstone washout 
 
 ra Roof fall 
 - Roll 
 
 200 
 
 Scole, ft 
 
 'J^'-V-'-- 
 
 
 
 nDdg 
 
 dS 
 a 
 
 □ ddDD 
 
 ^, , .DODDDaDD^anDDDDDDD 
 
 f-,nnnnr-' L't- nnnrfctdyy 
 
 fDDD 
 
 
 □ DD 
 ODD 
 
 m 
 
 [3Y]nDi:7Dac7^zi:?ODac7Dan 00000 c/jaaDDaDannDDC 
 
 NORTHEAST MAINS 
 
 Figure 21.— Section of mine in southern West Virginia, showing iocation of coaibed roii and iocai roof falls. 
 
 5=7.6 
 
 Shale roof 
 'Coal 
 
 """■Shale parting 
 Coal 
 
 
 
 - 
 
 
 
 5 
 
 1 
 
 
 1 
 
 
 
 25 
 
 
 50 
 
 
 Scale, 
 
 fT 
 
 
 Figure 22.— Cross section of roll A-A' from figure 21. 
 
19 
 
 1,600 
 1,200 
 
 in 
 a. 
 
 (O 800 
 
 UJ 
 
 a: 
 
 \- 
 
 400 - 
 
 
 
 Tension 
 
 Jl 
 
 -• - 
 
 Compression - 
 
 KEY 
 
 ■Roof midspan 
 Roof corner 
 
 0.10 
 
 0.15 
 
 0.20 
 
 0.25 
 
 PITCH OF ROLL 
 
 Figure 23.— Stress values for critical points of an entry 
 versus severity of pitch of an overlying roll. [Adapted from Wang 
 (48)] 
 
 USING PREDICTION OF FAILURE IN MINE DESIGN 
 
 For particular cases of cutter roof failure, the causes 
 can be determined, as demonstrated in the preceding 
 section. During the design stages of mining, one may 
 include the possibility of encountering cutter roof failure 
 by recognizing the extent to which the factors of 
 propagation may influence a particular area. On an 
 elementary level, stream valleys that overlie the coalbed 
 can be avoided, or a rough estimation of their influence 
 may be assessed through the use of the empirical formulas 
 previously presented. If information on the state of the 
 in situ stress is available, entries can be oriented so that 
 the in situ stress has the least effect on ground stability. 
 Additionally, a comprehensive analysis of the minesite 
 
 can be conducted if meaningful data on rock properties 
 and in situ stress can be obtained and then incorporated 
 into numerical models. 
 
 Any specific prediction of cutter roof failure, with 
 respect to location, is very difficult. However, if obvious 
 factors are accounted for in the design process, changes to 
 the design may not have to be implemented at a later date. 
 The following section on control measures includes 
 subsections on mine design changes and indirect control 
 measures. These can be integrated into the initial design 
 (when conditions warrant), effectively decreasing the risk 
 of exposing miners to cutter roof failure. 
 
 SELECTING CONTROL MEASURES 
 
 Before a control method for cutter roof failure is 
 selected, the cause or at least the pattern of failure should 
 be determined, if possible. For example, if it can be 
 determined that failure is occurring only in localized 
 areas of the mine, it is highly probable that control 
 methods will only need to be applied locally. On the other 
 hand, if failure is the result of regionally high, horizontal, 
 biaxial stresses, it may be more effective to reorient mine 
 headings as opposed to using supplemental support. 
 
 Figure 24 is a decision process diagram demonstrat- 
 ing a suggested method of analysis for determining the 
 cause of cutter roof failure at a particular mine. The 
 decision process is based on observations of underground 
 failure, geologic conditions, and surface topography. Once 
 the cause has been identified, the rock mass as a whole is 
 analyzed to determine the influence of characteristics of 
 the rock that may not be as readily apparent. Once a 
 
 control measure is implemented, careful monitoring is 
 conducted to assess its performance. If the performance is 
 poor, further investigation is necessary to determine the 
 mechanisms involved and further modifications should be 
 made to the support technique. 
 
 ARTIFICIAL SUPPORT 
 
 Support of cutter roof failure in the past has been 
 directed mainly toward the resupport of roof that has 
 already begun to fail. This technique has proven to be 
 unsatisfactory. The only resupport methods even margi- 
 nally successful are (1) extensive cribbing {19, 47) and (2) 
 a combination of additional bolting, trussing, and resin 
 injection (3, 5, 19); both of these methods are prohibitively 
 expensive, and cribbing often results in the loss of a 
 
20 
 
 Begin 
 
 
 
 No 
 
 Is failure 
 
 of the 
 
 
 
 cutter roof type? 
 
 
 
 |Yes 
 
 
 
 Mine wide 
 
 Is failure localized 
 
 or mine wide in 
 
 occurrence? 
 
 Local 
 
 
 
 
 
 
 1 
 
 r 
 
 
 
 
 i 
 
 
 
 
 Does failure show 
 a preferred 
 orientation? 
 
 Yes 
 
 
 Are areas of failure 
 beneath stream 
 
 valleys' 
 
 Yes 
 
 Design mine for 
 
 limited exposure 
 
 to these areas and 
 
 employ artificial 
 
 support. 
 
 
 
 
 
 
 
 No 
 
 
 i 
 
 
 Ino 
 
 
 
 
 
 
 Follow recommen - 
 dationsfor strongly 
 biaxial high horizon- 
 tal stresses. I.e., entry 
 orientation. 
 
 
 Is failure in areas 
 
 affected by multiple 
 
 seam mining? 
 
 Yes 
 
 Follow recommen- 
 dations for multiple 
 seam mining support. 
 
 - 
 
 
 
 i Continue 
 
 
 
 No 
 
 
 
 
 
 
 Analyze rock mass 
 characteristics to deter- 
 mine influence on 
 failure. Modify support 
 based on findings 
 
 
 Is failure mainly 
 isolated to entries 
 adjacent to the solid? 
 
 Yes 
 
 Consider design 
 
 options for reducing 
 
 number of adjacent 
 
 entries. 
 
 — 
 
 
 
 
 
 
 
 
 f 
 
 
 
 ,No 
 
 
 
 
 
 
 Implement control 
 method. 
 
 
 Continue 
 
 
 
 i 
 
 
 neg 
 
 f 
 dtive 
 
 Analyze results. 
 
 
 Figure 24. — Decision process diagram for determining the cause of cutler roof faiiure and selecting control measures. 
 
 passageway or at least an increase in airflow resistance. 
 Recent studies have concentrated on the prefailure control 
 of cutter roof iailure, with limited success, and a large 
 portion of the methods developed in these studies remain 
 to be field-tested. The following discussion of control 
 methods presents the use of artificial support and mine 
 design changes as attempts to control the occurrence of 
 cutter roof failure rather than to resupport failed roof 
 
 Angle Bolting 
 
 Angle bolting was one of the first artificial support 
 methods employed to inhibit the formation of cutter roof 
 failure. During the early 1950's, when roof bolting was 
 being promoted for the first time in the coal mining 
 industry, one of the major selling points for the use of 
 angle bolting was its effectiveness in controlling cutter 
 failure development (45-46). Figure 25 presents the two 
 
 most commonly employed types of hardware associated 
 with angle bolting. Figure 26 shows the plan of 
 installation that would be used when angle bolting is a 
 part of the normal roof control plan. When a tensioned 
 angled roof bolt is placed near the corners of the entry, the 
 shear stress in the corners is redistributed, effectivelely 
 decreasing the unsupported span of the entry. With a 
 reduction in the shear stress, the probability of failure is 
 reduced. Problems associated with angle bolting stem 
 mainly from the availability of equipment needed for 
 drilling inclined bolt holes (an angle of 45° with the 
 horizontal is recommended). The use of handheld 
 pneumatic drills for installation is awkwar/l and time 
 consuming. In addition, if the bolts are not installed 
 on-cycle, shortly after mining of the cut, the roof may sag 
 significantly, creating dangerously high shear stresses in 
 the corners of the entry. Dual-boom tilt-head bolters are 
 available that make the on-cycle installation of angle 
 bolts a relatively routine operation (5). 
 
21 
 
 A ROCKER NUT TYPE B, ANGLE IRON TYPE 
 
 Figure 25. — Angle bolting hardware. 
 
 
 ^~ _,„-^ — 
 
 ^^^-;^ 
 
 j— 
 
 
 .:^^_'^_'^ J2= ^ 
 
 - ,-v-- -->«^r-^ 
 
 =^ 
 
 fc=^ Shale — 
 xs 
 
 
 ^* 
 
 ^ 
 
 ^^^S[ 
 
 — ^^^ 4 
 
 r A 
 
 — 4'— 
 
 1 o' 
 
 1^ '^ 
 
 — 4'- +3'- 
 
 1 
 
 v////y///// 
 
 lo 
 
 A CROSS SECTION VIEW 
 
 10 
 
 I I I 
 
 Scale, ft 
 
 ----E H Y S & 
 
 4' 
 
 Shale 
 roof 
 
 A- — 
 
 B, PLAN VIEW 
 Figure 26. — Angle bolt installation. 
 
 Truss Bolting 
 
 Truss bolting is an artificial support technique 
 commonly recommended for cases of severe cutter roof 
 failure not significantly deterred by the installation of 
 angle bolts (5, 36). The two most commonly used designs of 
 truss bolts are shown in figure 27, along with an 
 indication of the theoretical components of compressive 
 force that make this method of support successful. 
 Controversy over the comparative effectiveness of the two 
 separate designs has arisen from the ability of the truss in 
 figure 27B to maintain equal tension in both the 
 
 A ANGLE BOLT TYPE 
 
 B, CONTINUOUS BOLT TYPE 
 
 Figure 27. — Two basic designs of roof bolt trusses. Arrows 
 shown in A represent compressive forces which are also true 
 forB. 
 
 crossmember and bolts, while the truss of figure 27A 
 allows for adjustments in these tensions after installation 
 (29-33). An advantage to the truss type of figure 27A is 
 that the crossmember does not have to be installed 
 immediately and can be used intermittently as conditions 
 warrant. The normal plan of installation for both types of 
 truss bolts is shown in figure 28A, but some mines have 
 gained amendments to their ground control plans allow- 
 ing them to use the angle bolt portion of the truss (shown 
 in figure 27A) on-cycle as a replacement for the outer two 
 bolts. The installation of the crossmember then qualifies 
 as supplementary support when hazardous conditions are 
 encountered (fig. 28B) (5). This has proven to be 
 exceptionally useful in areas where cutter roof initiates at 
 clastic dikes. In-mine monitoring of roof behavior near 
 clastic dikes at the Greenwich North Mine showed that 
 trusses were most successful when installed on-cycle (19). 
 Again, the obvious drawback to truss bolting is the need 
 
22 
 
 Entry 
 
 Entry 
 
 A SUPPLEMENTAL SUPPORT B, ON-CYCLE INSTALLATION 
 
 20 
 
 I I I 
 
 Scale, ft 
 Figure 28. — Plan view of truss installation. 
 
 for a bolting machine that can drill inclined bolt holes (the 
 recommended angle of installation is 45°). Installation 
 on-cycle is also recommended for the reasons discussed 
 under the heading of "Angle Bolting." 
 
 Other Supports 
 
 As mentioned earlier, cribing and posts are generally 
 used for the resupport of roof after cutter failure has 
 formed; however, in specific cases where the failure begins 
 at a clastic dike or some other minor geologic structure 
 and in situ stresses are not too high, strategic placement 
 of cribbing and posts has deterred cutter formation (fig. 
 29). At the Greenwich Mines, some success has been had 
 using this technique. Cribs and posts should be installed 
 shortly after mining, and the support should be placed in 
 such a way as not to yield significantly (e.g., a minimum of 
 cap pieces). 
 
 With respect to conventional bolting practices, cutter 
 failure fi-equently propagates to just above the anchor 
 horizon, resulting in massive roof failure. Changes in bolt 
 length most often result in only a change in the height to 
 which cutter failure propagates. However, for cases of 
 cutter failure that are not severe, some operators have 
 found a combination of bolt lengths across an entry to be 
 successful (e.g., bolts next to ribline are shorter than bolts 
 in center of entry). Another remedy has been to use 
 point-anchor-resin, tensioned, rebar-type bolts and to 
 ensure that these bolts are uniformly tensioned, upon 
 installation, throughout the entry. 
 
 MINE DESIGN CHANGES 
 
 Mine design changes, as control measures, frequently 
 take advantage of the elemental factors that cause cutter 
 roof failure. In the following examples, rock property data 
 and knowledge of the in situ stress state are valuable in 
 assessing the probability of success and developing an 
 implementation plan for the control measure. Obtaining 
 accurate measurements of the in situ stress state and 
 meaningful rock property values is difficult, but often 
 necessary. In the literature are several cases where mine 
 officials used mine design changes to control cutter roof 
 
 failure without the aid of rock mechanics data. However, 
 since a great deal of time and expense is invested in 
 making mine design changes, the benefits of having these 
 data are obvious. 
 
 Sacrifice Entries^ 
 
 Sacrifice entries have been used as a method of 
 ground control since the 18th century, and dating the 
 common use of sacrifice entries in the United States at 
 approximately 1935, Roberts (42) referred to this method 
 as the use of caving chambers and described it as follows: 
 
 In essence, the caving chamber is an auxiliary 
 road, driven parallel to the main roads, and kept 
 slightly in advance of them. At short periodic 
 intervals all supports are withdrawn from the 
 caving chamber, which is allowed to collapse, and 
 as result of the falls of roof in this chamber the 
 roof in the neighboring roads remains solid. 
 
 The premise for using sacrifice entries is that" 
 extremely high horizontal in situ stresses exist (or that 
 measurements have shown that high in situ stresses exist) 
 that are thought to be the primary cause of roof failure at 
 the site, and further, that by initiating failure in an entry 
 ahead of and parallel to future adjacent entries, the in situ 
 stress can be relieved to manageable levels, allowing 
 adjacent entries to advance without incident. Nicholls (38) 
 has discussed the use of this method in conjunction with 
 the problem of cutter roof failure in Australia and 
 reported limited success. 
 
 Figure 30 illustrates the use of contemporary caving 
 chambers in three-entry gateroads. Roof rock is mined in a 
 
 ^he author thanks Nicolas P. Kripakov, mining engineer, Denver 
 Research Center, Bureau of Mines, Denver, CO, for providing the results of 
 his finite-element analysis of the arch-sacrifice-entry method and for 
 relating his experience with the subject. 
 
 » 
 
 Clastic r^V 
 
 dike '-'^ 
 
 -Cribbing ^^ 
 
 Entry 
 
 
 
 20 
 
 Scale, ft 
 
 Figure 29. — Plan view of mine entry showing placement of 
 cribbing adjacent to clastic dike to deter cutter failure 
 formation. 
 
o-h =1,900 PS 
 
 23 
 
 Center -lo-center 
 dimension 
 
 Center- to- center 
 dimension, ft 
 
 Reduction in sheor stress 
 at entry corner, pet 
 
 Inside (1) 
 
 Outside (0) 
 
 50 
 40 
 30 
 
 5 2 
 
 1 1 3 
 24 6 
 
 3 5 
 5 7 
 
 10,1 
 
 A, CENTER ENTRY USED FOR STRESS RELIEF 
 
 I \ I 
 
 C 
 
 I i o^i^ 700 psi 
 
 Center -to -center 
 dimension, ft 
 
 Reduction in stiear stress 
 at entry corner (C), pet 
 
 50 
 40 
 30 
 
 82 
 
 156 
 28 3 
 
 crh=l900psi Center -to -center 
 
 ' dimension 
 
 e, OUTER 2 ENTRIES USED FOR STRESS RELIEF 
 
 Figure 30. — Contemporary versions of caving chambers for use in three-entry gateroad configurations. (Courtesy N. P. Kripakov) 
 
 rough arch outline to a height above the coalbed in either 
 the center entry (fig. 30A) or the two other entries (fig. 
 SOB), with only steel arches and steel lagging as support. 
 The entry (or entries) mined with the arches is driven 
 approximately 100 ft in advance of the other entries. 
 Approximately 18 in of space is left between the steel 
 arches and the newly exposed roof surface (an arbitrary 
 amount of space leaving sufficient room for expansion), 
 thus allowing the roof to cave onto the arches and relieve 
 in situ stresses. Upon abandonment of the gateroad, it 
 may be possible to retrieve the arches and lagging for 
 future use. 
 
 Kripakov conducted the finite-element analysis of the 
 two different scenarios using the stress values and rock 
 properties used to generate the results shown in table 1. 
 The results of the analysis are shown in figure 30, 
 demonstrating that reduction of pillar sizes increases the 
 effectiveness of the caving, as does the use of two arched 
 entries as opposed to one. Since Kripakov's finite-element 
 code generally treats rock as a continuum, the actual 
 reduction in shear stress may be much greater in the 
 actual mine environment because of shear displacement 
 along bedding planes. The basic principle in reducing the 
 shear stress occurring at an entry corner is reducing the 
 difference in the principal stresses. While the magnitude 
 of the stresses is important, the difference in magnitudes 
 can control failure. 
 
 Although sacrifice entries are not in wide use today, 
 their success in the past suggests some possibilities for 
 future use. From the differences between the old caving 
 chamber, as explained by Roberts, and the contemporary 
 sacrifice entry, which requires special equipment and 
 arches, it is obvious that companies must go to great effort 
 to incorporate this method as a regular practice. 
 
 Pillar Softening And Yield Pillars 
 
 Pillar softening is a technique developed by Wang 
 |-?7i for reducing the magnitude of stress concentrations in 
 the corners of entries. The concept is based on coal pillar 
 
 elasticity versus roof rock elasticity, discussed in the 
 section "Rock Mass Characteristics," and the effect of this 
 ratio on the magnitude of stresses in the corners of the 
 entry (fig. 14). By reducing the elasticity of the pillars to 
 some distance away from the entry, the stress concentra- 
 tion in the corners of the entry are effectively redis- 
 tributed. Using finite-element analysis, Wang found that 
 by drilling 6-in-diam holes into the pillars and face as an 
 entry was advanced, the elasticity of the pillar near the 
 rib and the stress in the corners of the entry could be 
 reduced. 
 
 Softenedl 
 zone I 
 
 KEY 
 ■ First advance 
 Second advance 
 
 Figure 31. — Plan for placement of auger holes for pillar- 
 softening concept. [Adapted from Maxwell (34)\ 
 
24 
 
 5 
 
 Scale, ft 
 
 Figure 32.— Cross-sectional view of rib-slotting method. 
 [Adapted from Kripakov (27)] 
 
 The pillar-softening concept was tested in Mine 32 of 
 the Bethlehem Mine Corp. (34) near Ebensburg, PA, an 
 area of the Appalachian Basin plagued by the problem of 
 cutter roof failure. Figure 31 illustrates the pattern of 
 holes used during the testing. An attempt to measure the 
 in situ horizontal stress was made, with inconclusive 
 results, and other measurements were made of pillar 
 stresses, material properties, roof strain, convergence, and 
 tilt. The results of the tests showed a reduction in stress in 
 the corners of the entry; however, there was no indication 
 that roof stability improved as a result of the softening. 
 
 Kripakov (27) further investigated Wang's method 
 and found that an optimum location for softening holes 
 was at the roof-coal interface, which resulted in an 
 effective decrease in stress concentration at the comers of 
 10 pet (table 1). Since the original pillar-softening test at 
 Mine 32 did not show a significant improvement in roof 
 conditions, Kripakov took the method a step further and 
 introduced the concept of rib slotting (fig. 32). Through 
 finite-element analysis, he found a 40-pct reduction in 
 stress concentrations in the corners of the entry when 
 horizontal 3-in-thick slots were created at the roof-coal 
 interface (table 1). The method has not been exhaustively 
 tested, and a means of efficiently cutting slots in the ribs 
 has yet to be presented. However, the initial analysis of 
 the method suggests it may be worthy of additional 
 testing. A great potential for limited use of this concept 
 exists for special cases, such as cutter failure in the outer 
 entries of a multiple-entry development. The pillar- 
 softening method may also be useful when a face area is to 
 be left idle for several days. In situations like this, the face 
 area essentially becomes an outer entry in a multiple- 
 entry scenario. Pillar softening in the area may deter 
 failure until mining is resumed and additional support 
 can be installed. 
 
 Another method of reducing the elasticity of a coal 
 pillar, the yield pillar design, has recently reemerged in 
 the coal mining industry as a viable ground control 
 method. While the method is being used only ex- 
 perimentally for controlling classic cutter roof failure, at 
 least three mines are using yield pillars for other ground 
 control problems. The basic concept is that pillars are 
 small enougb so they intentionally yield and transfer the 
 majority of roof loads to the abutments. The result is a 
 reduction of overall roof and floor stresses (49). Kripakov 
 (27) suggested yield pillars as a possible solution for 
 
 Slot - 
 
 A PLAN VIEW 
 
 10 r 
 
 h 
 
 10 
 Scale, ft 
 
 Slot holes 
 
 S, CROSS SECTION VIEW 
 
 Figure 33. — Placement of holes for roof-slotting method. 
 [Adapted from Maxwell (34)] 
 
 controlling cutter roof failure at the Kitt Min§, although 
 in-mine verification was never conducted. 
 
 Roof Slotting 
 
 In conjunction with the testing of the pillar-softening 
 method at Mine 32 of the Bethlehem Mine Corp., a second 
 method for controlling cutter roof failure was tested. 
 
25 
 
 which consisted of drilling a series of holes adjacent to the 
 entry to form a vertical slot in the roof rock above the 
 pillar (fig. 33). Entries treated by slotting prior to 
 development showed a marked improvement in roof 
 conditions over adjacent entries, and instrumentation 
 revealed that the slots did provide for relief of horizontal 
 stress. Unfortunately, because of the inability to use this 
 method efficiently during production, it is not very 
 practical. But the initial indications of success suggest 
 that future efforts aimed at developing a means for using 
 the method during production would be worthwhile. 
 
 Entry Reorientation 
 
 The practice of reorienting entries, in an attempt to 
 eliminate roof falls that occur in entries of a particular 
 orientation, has been a successful control method when 
 applied to problem ground conditions caused by a biaxial 
 horizontal stress field. The theoretical premise for the 
 success of this method lies in the relation of horizontal to 
 vertical stress, discussed under the heading of "Stress 
 Environment." If the horizontal stress is strongly biaxial, 
 the entries oriented perpendicular to the maximum 
 principal horizontal stress are under the greatest in- 
 fluence of the stress. By reorienting mine headings so that 
 both entries are perpendicular to the same least stress 
 value {fig. 34), the influence of the biaxial stress field can 
 be offset. However, in many cases, reorientation to obtain 
 least stress values is not an adequate solution. For these 
 cases, a solution can be found by orienting the main 
 headings parallel with the maximum principal horizontal 
 in situ stress and staggering pillars, thus isolating the 
 occurrence of roof failure to crosscuts. In addition, Aggson 
 (3) suggests a reduction in the width of crosscut entries to 
 further reduce the probability of failure. 
 
 In the Kitt Mine, only a 22-pct difference in 
 magnitude was found between the minor and major 
 principal horizontal stresses. In this case, reorienting 
 entry headings to an optimum angle from the principal 
 horizontal stresses would not decrease stress perpendicu- 
 lar to the entries enough to create entry stability. In fact, 
 the greatest reduction that could be realized by reorienta- 
 tion of the entries would be only an 11-pct reduction of the 
 major principal horizontal stress. 
 
 Entry reorientation has been used to control cutter 
 roof failure (and other types of failure) caused by a highly 
 biaxial, horizontal stress field in cases where the stress 
 field was of local occurrence and in other cases where it 
 was of regional occurrence. Local cases have usually been 
 associated with sedimentary or compactional features, 
 such as sandstone channels or rolls. Connelly ilO) refers to 
 the use of entry reorientation in Australia as a successful 
 method for offsetting the influence of "stone rolls." In the 
 case he cites, headings were reoriented to intersect the 
 rolls at an oblique angle, resulting in an immediate 
 improvement in roof conditions. In the United States, 
 similar conditions have been found, as in the West 
 Virginia mine shown in figure 21. Reorientation was also 
 used at this mine in an attempt to offset the influence of 
 the structure, and no further failure ensued. However, in 
 cases such as this, in-mine verification of a local biaxial 
 stress field has not been established and reorientation has 
 not been verified, through instrumentation, as the cause 
 of improvement. The Pennsylvania mine shown in figure 
 11 displays similar characteristics; in the section oriented 
 45 from the headings of the rest of the mine there is no 
 
 A 
 
 / 
 / 
 / 
 / 
 
 / 
 / 
 / 
 
 ^ // / /A 
 
 / 
 / 
 / 
 
 / 
 
 / 650 psi 
 
 A, 
 
 ORIENTED PERPENDICULAR TO 
 MAJOR PRINCIPAL STRESS 
 
 650 
 psi 
 
 1,300 
 psi 
 
 975 psi 
 as calculated 
 from av stress 
 
 B, ORIENTED 45* FROM MAJOR 
 PRINCIPAL STRESS 
 
 Figure 34. — Apparent values of horizontal in situ stress for 
 the two extremes of entry orientation versus orientation of 
 actual principal in situ stress. 
 
 roof failure. While these two cases are apparently 
 successful, in many other cases no deterrence of failure 
 was realized. 
 
 Some operators have reported on the implementation 
 and success of entry reorientation and its use in 
 controlling cutter roof failure associated with a regionally 
 high biaxial stress field. In some cases (7-8) in situ stress 
 measurements were not taken to quantify the stress field 
 but rather the decision to implement changes was made 
 based on in-mine observations of roof failure. In another 
 
26 
 
 case (29), attempts were made to quantify the stress field 
 before implementing successful reorientation changes, 
 although the results were inconclusive based on the fact 
 that the deepest overcoring measurement was taken only 
 6 ft from the mine opening. 
 
 Although defining the stress environment and other 
 factors of roof failure propagation is important, reasonable 
 engineering decisions can be made based on keen 
 observations, as was done in the cases cited above. The 
 value of having as much information as possible is 
 obvious, however, in light of the magnitude of the changes 
 made. 
 
 INDIRECT CONTROL MEASURES 
 
 Other ideas about controlling cutter roof failure 
 include the same kind of reasoning that suggests planing 
 
 around stream valley areas that may be susceptible to 
 failure. This type of reasoning leads to the conclusion that 
 specific mining methods may actually reduce the threat of 
 encountering severe cutter roof failure. Many mines 
 currently experiencing cutter roof failure are employing 
 the room-and-pillar mining method, which exposes a 
 tremendous amount of roof area that must be supported 
 over long periods of time. In regions where the threat of 
 cutter failure is such that control may be impractical or 
 impossible, the possibility of using the longwall mining 
 method should be assessed. For a given mined area, this 
 method exposes less roof to long-term support needs; 
 therefore, there is less potential for having to deal with 
 the problem. Additionally, gateroad systems lend them- 
 selves to short-term innovative kinds of control measures, 
 such as the sacrifice entry concept previously explained. 
 
 CONCLUSIONS AND RECOMMENDATIONS 
 
 This report outlines the most commonly cited theories 
 on the formation of cutter roof failure and the most 
 commonly suggested methods for controlling its occur- 
 rence. No unique solution exists for controlling or avoiding 
 cutter roof failure. Some operators have had limited 
 success in controlling failure propagation by systematical- 
 ly analyzing the patterns of failure in their mines and 
 then, through a process of elimination, selecting a control 
 method. The limitations on their success may stem from 
 the fact that although theories on cutter failure exist, few 
 have been verified through in-mine experimentation. 
 Likewise, when control measures have been implemented, 
 inadequate monitoring has resulted in an inability to 
 determine the overall effect of the measure on roof 
 stability. Continued research is needed by mine operators, 
 the Bureau, and other research organizations, so that 
 appropriate modifications can be made to existing theories 
 and control methods to increase their rate of success. 
 
 In addition to using the decision process diagram 
 illustrated in figure 24 (for the basic determination of 
 
 probable causes of failure and subsequently for the 
 selection of a control measure), mine operators are 
 encouraged to consider mining methods that reduce the 
 amount of exposed roof that must be supported for long 
 periods of time. Longwall mining meets this criterion in a 
 manner that provides fiexibility for the employment of 
 innovative control methods such as yield pillars, re- 
 orientation of entries, and sacrifice entries. In localized 
 areas of a relatively high probability of cutter roof failure, 
 the basic premining plan may be altered to take into 
 account these areas and either use them for barrier pillar 
 areas or only mine them upon retreat. 
 
 Until future advances are made in understanding the 
 cutter failure phenomenon, the present state of the art 
 does provide options for mining in areas where this type of 
 failure occurs frequently. However, it remains the 
 responsibility of individual mine operators to weigh the 
 cost of resupporting failed areas and the threat of injury to 
 miners against the investment made to select and 
 implement control methods. 
 
 REFERENCES 
 
 1. Agapito, J. F. T., J. R. Aggson, S. J. Mitchell, M. P. Hardy, 
 and W. N. Hoskins. Study of Ground Control Problems in Coal 
 Mines With High Horizontal Stresses. Paper in Proceedings of 
 Twenty First Symposium on Rock Mechanics: A State of The Art. 
 Univ. MO, Rolla, MO, 1980, pp. 820-828. 
 
 2. Aggson, J. R. Coal Mine Floor Heave in the Beckley 
 Coalbed, An Analysis. BuMines RI 8274, 1978, 32 pp. 
 
 3. How To Plan Ground Control. Coal Min. & 
 
 Process., v. 16, No. 12, 1979, pp. 70-73. 
 
 4. Stress-Induced Failures in Mine Roof. BuMines 
 
 RI 8338, 1979, 16 pp. 
 
 5. Barish, K. Truss Bolting On-Cycle in Jane Mine Lower 
 Freeport Seam. Paper in Proceedings, Fourth Conference on 
 Ground Control in Mining. WV Univ., Morgantown, WV, 1985, 
 pp. 1-10. 
 
 6. Bieniawski, Z. T. Rock Mechanics Design in Mining and 
 Tunneling. A. A. Balkema, 1984, 272 pp. 
 
 7. Blevins, C. T. Coping With High Lateral Stresses in an 
 Underground Illinois Coal Mine. Soc. Min. Eng. AIME preprint 
 82-156, 1982, 6 pp. 
 
 8. Blevins, C. T., and D. Dopp. Ground Control Experiences in 
 a High Horizontal Stress Field at Inland Steel Coal Mine No. 2. 
 Paper in Proceedings, Fourth Conference on Ground Control in 
 Mining. WV Univ., Morgantown, WV, 1985, pp. 227-233. 
 
 9. Chase, F. E. Clay Veins: Their Physical Characteristics, 
 Prediction, and Support. Paper in Proceedings, Fourth Confer- 
 ence on Ground Control in Mining. WV Univ., Morgantown, WV, 
 1985, pp. 212-219. 
 
 10. Connelly, M. A. The Uses of Geologic Structural Analyses 
 in Predicting Roof Conditions in Coal Mining. Paper in 
 Proceedings of the Symposium on Stress and Failure Around 
 Underground Openings. Univ. Sydney, Sydney, N.S.W., Austra- 
 lia, paper 13, 1967, pp. 1-2. 
 
 11. Enever, J. R., and J. McKay. Stress Measurements at 
 Nattai North Colliery and Their Interpretation in Terms of 
 Sedimentological and Topographic Features. CSIRO, Div. Ap- 
 plied Geomechanics, Mount Waverley, Victoria, Australia, Rep. 
 29, 1980, 12 pp. 
 
 12. Enever, J. R., J. Shepherd, and J. Huntington. An Initial 
 Mathematical Relationship Between Underground Working 
 
27 
 
 Conditions and Overlying Surface Topography With Reference to 
 the Western Coalfields, N.S.W. CSIRO, Div. Mineral Physics, 
 Mount Waverley, Victoria, Australia. Rep. 5, 1978, 6 pp. 
 
 13. Engelder, T. Is There a Genetic Relationship Between 
 Selected Regional Joints and Contemporary Stress Within the 
 Lithosphere of North America? Tectonics, v. 1, No. 2, 1982, pp. 
 161-177. 
 
 14. Ferguson, H. F. Geological Observations and Geotechnical 
 Effects of Valley Stress Relief in the Allegheny Plateaus. Pres. at 
 Annu. ASCE Meet., Los Angeles, CA, 1974, 31 pp.; available from 
 J. L. Hill III, BuMines, Pittsburgh, PA. 
 
 15. Valley Stress Release in the Allegheny 
 
 Plateau. Bull. Assoc. Eng. Geol., v. 4, No. 1, 1967, pp. 63-69. 
 
 16. Haycocks, C, M. Karmis, E. Barko, J. Chaman, S. Hudock, 
 B. Ehgartner, and S. Webster. Ground Control Mechanics in 
 Multiple-Seam Mining (grants G1105050 and G1115511, VPI), 
 BuMines OFR 7-84. 
 
 17. Haycocks, C, M. Karmis, and B. Ehgartner. Multiple 
 Seam Mine Design. Paper in State-of-the-Art of Ground Control 
 in Longwall Mining and Mining Subsidence. Soc. Min. Eng. 
 AIME, 1982, pp. 59-65. 
 
 18. Haycocks, C, M. Karmis, and E. Topuz. Optimizing 
 Productive Potential in Multi-Seam Underground Coal Mining. 
 Paper in Coal Conference & Expo VI. McGraw-Hill, 1981, 
 pp.151-164. 
 
 19. Hill, J. L. Ill, and E. R. Bauer. An Investigation of The 
 Causes of Cutter Roof Failure in a Central Pennsylvania Coal 
 Mine: A Case Study. Paper in Rock Mechanics in Productivity 
 and Protection, Twenty-Fifth Symposium on Rock Mechanics. 
 Soc. Min. Eng, AIME, 1984, pp. 603-614. 
 
 20. Hoek, E., and E. T. Brown. Underground Excavations in 
 Rock. Inst. Min. and Metall., London, 1980, pp. 110-243. 
 
 21. Hylbert, D. K. The Classification, Evaluation and Projec- 
 tion of Coal Mine Roof Rocks in Advance of Mining. Min. Eng. 
 (Littleton, CO), v. 30, 1978, pp. 1667-1676. 
 
 22. Developing Geological Structural Criteria for 
 
 Predicting Unstable Mine Roof Rocks (contract HO 133018, 
 Appalachian Coal Min. Inst., Moorhead State Univ.). BuMines 
 OFR 9-78, 1977, 249 pp.; NTIS PB 276-735/AS. 
 
 23. lannacchione, A. T., J. T. Popp, and J. A. Rulli. The 
 Occurrence and Characterization of Geologic Anomalies and 
 Cutter Roof Failure: Their Affect on Gateroad Stability. Paper in 
 Stability in Underground Mining 11. Soc. Min. Eng. AIME, 1984, 
 pp. 428-445. 
 
 24. Jeffrey, R. G. Jr., and J. J. K. Daemen. Analysis of 
 Rockbolt Reinforcement of Layered Rock Using Beam Equations. 
 Paper in Proceedings of the International Symposium on Rock 
 Bolting, A. A. Balkema, 1983, pp. 173-185. 
 
 25. Kent, H. B. Geologic Causes and Possible Preventions of 
 Roof Fall in Room-And-Pillar Coal Mines. PA Bur. Topogr. and 
 Geol. Surv., Inf Circ. 75, 1974, 17 pp. 
 
 26. Kertis, C. A, Reducing Hazards in Underground Coal 
 Mines Through the Recognition and Delineation of Coalbed 
 Discontinuities Caused by Ancient Channel Processes, BuMines 
 RI 8987, 1985, 23 pp. 
 
 27. Kripakov, N. P. Alternatives for Controlling Cutter Roof in 
 Coal Mines. Paper in Proceedings, Second Conference on Ground 
 Control in Mining. WV Univ., Morgantown, WY., 1982, pp, 
 142-151. 
 
 28. Lang, T. A, Theory and Practice of Rock Bolting, Trans, 
 Soc, Min. Eng. AIME, v, 223, 1962, pp, 333-348. 
 
 29. Lizak, J, B. and J. E. Semborski. Horizontal Stresses and 
 Their Impact on Roof Stability at the Nelms No, 2 Mine. Pres. at 
 
 4th Conf on Ground Control in Mining, WV Univ., Morgantown, 
 WV, July 22-24, 1985, 7 pp.; available from J. L. Hill III, 
 SuMines, Pittsburgh, PA. 
 
 30. Mallicoat, W. R. Truss Bolting With Point Resin Anchor- 
 age. Min. Congr. J., v. 64, June 1978, pp. 47-50. 
 
 31. Mangelsdorf, C. P. Current Trends in Roof Truss Hard- 
 ware. Paper in Proceedings, Second Conference on Ground 
 Control in Mining. WV Univ., Morgantown, WV, 1982, pp. 
 108-112. 
 
 32. Evaluation of Roof Trusses, Phase I (contract 
 
 GO166088, Univ. Pittsburgh). BuMines OFR 56-82, 1979, 111 
 pp.; NTIS PB 82-209768. 
 
 33. The Role of Friction in Roof Truss Behavior. 
 
 Trans. Soc. Min. Eng. AIME, v. 268, 1980, pp, 1869-1879, 
 
 34. Maxwell, B., G. Zink, and F, D, Wang, Improving Coal 
 Mine Roof Stability by Pillar Softening (contract GO133063, CO 
 Sch. Mines), BuMines OFR 7-78, 1977, 105 pp,; NTIS PB 276-474. 
 
 35. Moebs, N. N. Roof Rock Structures and Related Roof 
 Support Problems in the Pittsburgh Coalbed of Southwestern 
 Pennsylvania. BuMines RI 8230, 1977, 30 pp. 
 
 36. Moebs, N. N., and R. M. Stateham, The Diagnosis and 
 Reduction of Mine Roof Failure, Coal Min., v. 22, 1985, pt. 1, No. 
 2, 1985, pp. 52-55; pt. 2, No, 3, 1985, pp, 42-48. 
 
 37. Geologic Factors in Coal Mine Roof Stability — 
 
 A Progress Report. BuMines IC 8976, 1984, 27 pp. 
 
 38. Nicholls, B. Pillar Extraction on the Advance at Oakdale 
 Colliery. Paper in Proceedings of the First International 
 Conference on Stability in Coal Mining. Miller Freeman Publ., 
 1978, pp. 182-196. 
 
 39. Obert, L., W. I. Duvall, and R. H. Merrill. Design of 
 Underground Openings in Competent Rock. BuMines B 587, 
 1960, 36 pp. 
 
 40. Peng, S. S., and U. Chandra. Getting The Most From 
 Multiple Seam Reserves. Coal Min. & Process., v. 17, No. 11, 
 1980, pp. 78-84. 
 
 41. Phillips, D. W. American Coal Mining. Colliery Guardian, 
 V. 175, No. 4523, 1947, pp. 37-43. 
 
 42. Roberts, A. A Review of the Problem of Strata Control in 
 Bord and Pillar Working. Colliery Guardian, v. 170, No. 4404, 
 1945, pp. 663-668. 
 
 43. Roley, R. W. "Pressure-Cutting": A Phenomenon of 
 Coal-Mine Roof Failures. Mechanization, v. 12, No. 12, 1948, pp. 
 69-74. 
 
 44. Su, W. H., and S. S. Peng. Cutter Roof and Its Causes. Soc. 
 Min. Eng. AIME preprint 85-131, 1985, 5 pp. 
 
 45. Thomas, E. Conventional Timbering Versus Suspension 
 Supports. BuMines B, 489, 1950, pp, 175-181, 
 
 46. Thomas, E., A. J. Barry, and A. Metcalf Suspension 
 Support Progress Report. BuMines IC 7533, 1949, 13 pp. 
 
 47. Wang, F, D., D. M, Ropchan, and M, C, Sun. Proposed 
 Technique for Improving Coal-Mine Roof Stability by Pillar 
 Softening. Trans. Soc. Min. Eng, AIME, v, 255, 1974, pp, 59-63. 
 
 48. Structural Analysis of a Coal Mine Opening in 
 
 Elastic Multilayered Material. BuMines RI 7845, 1974, 36 pp, 
 
 49, Wilson, A, H. The Effect of Yield Zones on the Control of 
 Ground. Paper in Sixth International Strata Control Conference. 
 Natl. Coal Board, Burton-on-Trent, Staffordshire, England, Sept. 
 1977, pp. 52-93. 
 
 50. Worotnicki, G, Effect of Topography on Ground Stresses. 
 Presented at Rock Mechanics Symp., Univ. Sydney, Sydney, 
 N.S.W., Australia, Feb. 19-20, 1969, 12 pp.; available from J. L. 
 Hill III, BuMines, Pittsburgh, PA. 
 
 U.S. GOVERNMENT PRINTING OFFICE: 1986—605-017/40,051 
 
 IImT.-BU.OF MINtS,P6H.,PA. 28352 
 
U.S. Department of the Interior 
 Bureau of Mines— Prod, and Distr. 
 Cochrans Mill Road 
 P.O. Box 18070 
 Pittsburgh. Pa. 15236 
 
 AN EQUAL OPPORTUNITY EMPLOYER 
 
 OFFICIALSUSINESS 
 PENALTY FOn PRIVATE USE, 1300 
 
 I I Do not wish to receive this 
 material, please remove 
 from your mailing list. 
 
 I I Address change. Please 
 correct as indicated. 
 
 U5. 0FnClALM A«' 
 
 uspostapeJ I 
 
 PtNAlTY 
 FOff 
 FRIWU 
 USE $300 
 
 0.7 5 
 
 F'liETkii 
 50 
 
 S04171 
 
 LIBRARY OF CONGRESS RCU-J 
 CATALnniNG IN PUBLICATIONS DIV 
 WASHINGTON DC 20540 
 
 W5 
 

 
 
 
 .4^ 
 
 
 t^ A^ ■ ':f^V/^•^ ""^^^ .c.-^* »'afe'» ^^ A^ ' »:. 
 
 
 -^^.^^ 
 
 
 ^°-^^. 
 
 
 
 1^ ..- V ->> "•* A*- -:^ 
 
 
 .♦^"V 
 
 y "r*- 
 
 » o 
 
 
 
 
 
 
 "V 'o^C 
 
 
 
 ^<^^ 
 
 
 
 
 
 "^-ov* ;^^'< -^^o? f^'^M': "^o^ '^^^^^\ '"t..^ 
 
 ^^-^^K 
 
 
 
 ?^°--^. 
 
 v-o^ 
 
 V 
 
 
 
 /^\ ^^K^° '^'^^^ °^^^: </'\ ^^»»° '*-^'"^^- 
 
 > V 
 
 o > 
 
 -*-. 
 
 4/ <<> 
 
 ^°-n^. 
 
 
 T.S* .'V 
 
 y V 
 
 
 
 .*' 
 
 -0 j^ 
 
 
 ♦ o 
 
 ^-•V, 
 
 
 
 • >. 
 
 
 ^*^ ..-•- 
 
 
 -^^^^ 
 
 
 
 
 r^" «•„ c^ 
 
 
 <*-._ ' 
 
 .-^o. 
 
 ' .^^''^.. \ 
 
 
 
 ' o » - 
 "■. -^ A^^ 
 
 
 
 
 
 o 
 
 
 ./'^ • 
 
 ^oV" 
 
 *0 -^^ ".. 
 
 '-^tff 
 
 
 A^ 
 
 
 
 
 
 
 
 -^^ . " • , 
 
 
 
 s» a"^ ^ oVJlvP^o .4?' '^. -Mis* av "*:<> V<<yMw* «?■ "^.0. .50ill^, aV- -^ 
 
 V »»»o- 
 
 
 
 oV V,. 
 
 
 
 
 ■" -^hi^ 
 
 ^^-^^-^^ " 
 
 
 %, *'7!,T^ A <. -0.1* .0 
 
 ■* ■^ V '^\r' ^j. A'- . ----•n™. - 
 
 
 
 ^oV^ 
 
 .-?■'■ ^o. *»'Trr.' .0 
 
 * 
 / 
 
 ^°-nK • 
 
 
 
 
 ^* . '^ 
 
 
 
 <<> 
 
 "q./^-'\o'^ 
 
 
 ^^0^ 
 
 ^^•^°- 
 
 
 
 
 » o 
 o 
 
 
 
 fk'.%.,<.^ 
 
 / v*^\^^' "<v*^'*%°'' v*^*/' "°^*^-'y .. v*^* 
 
 • % .-^"^ 
 
'L% "-t. 
 
 .^ '-^^0^ o 
 
 
 
 C° .'^1'. °o 
 
 
 
 
 .■:* 
 
 
 
 o-o, "*-o 
 
 <i. * 
 
 
 
 :. -^^o* : 
 
 
 
 
 < -n^o^ ^ 
 
 •- %. c^^ > 
 
 
 
 
 '<f> *». = ' -i 
 
 ' \<^ 
 
 ^* A 
 
 
 
 V '•^'* .^^'^ % *^.' .0^ ^^ -: 
 
 
 ^^°^ 
 
 
 ". o 
 
 
 
 
 
 
 v^^ 
 
 
 v.<^ 
 
 
 
 
 
 :- -n^-o^ ■.', 
 
 % 
 
 s^.^.\ 
 
 
 :-- V.s^^/^^'-. %,^^',:^ 
 
 v*^-':/ 
 
 
 
 
 .*^..' 
 
 ■^^ .0 
 
 ^°-n^. 
 
 
 
 '"^v^^\ .^°^»'°- /^;^^%\ .^°/^i>- .^^^>^' '-^ ^° •^<^'' 
 
 
 
 
 .* ^0 
 
 
 
 
 
 
 ^*\" 
 
 
 » O -J 
 
 \*^<>\..V^^-> ^°^*^\o^ \'^^\/ V^%°' \'^-\ 
 
 ^mi 
 *^%^-*',* 
 
 
 
 UEKT 
 BOOKBINDING JI-''^ 4 ' ,^ 
 
 
 
 .j^r,' ^ 1^^ •<#• "^ 
 
 
 
 
 
LIBRARY °'' 
 
 CONGRESS 
 
 • V,*'