fringe cracks: key structures for the interpretation of the …jte2/references/link124.pdf ·...

ABSTRACT

Vertical joints in Devonian clastic sedimen-tary rocks of the Finger Lakes area of NewYork State are ornamented with arrays offringe cracks that reveal the complex defor-mational history of the Appalachian plateaudetachment sheet during the Alleghanianorogeny. Three types of fringe cracks weremapped: gradual twist hackles, abrupt twisthackles, and kinks. Gradual twist hackles arecurviplanar en echelon fringe cracks thatpropagate with an overall vertical directionwithin the bed hosting the parent crack andare found in all clastic lithologies of the de-tachment sheet. Abrupt twist hackles propa-gate as planar features in thick shale bedsabove or below the siltstone beds hosting par-ent joints. Kinks propagate horizontally asplanar surfaces from the tips of parent jointsin siltstone beds. The breakdown of the parentjoint into either gradual or abrupt twist hack-les depends on the orientation and magnitudeof the remote stress field, internal fluid pres-sure, and the elastic properties of the bed. Thetwist angle of gradual twist hackles is larger incoarser clastic beds, indicating that stress andinternal pressure are more important param-eters than elastic properties in controllingbreakdown. Assuming that the vertical stressaxis (Sv ) equals 78 MPa at 3 km burial depth,the difference in twist angle between sand-stone and shale beds is used to estimate themaximum horizontal stress difference in theshale beds as SH – Sh ≈≈ 2.5 MPa when SH – Sh≈≈ 12 MPa in sandstone beds.

The twist angle of the fringe cracks and theabutting relationships of parent joints give an

indication of the overall change in stress fieldorientation within the detachment sheet dur-ing Alleghanian tectonics. These parent jointsindicate a regional clockwise stress rotation ofAlleghanian age concordant with the twist an-gle of fringe cracks throughout the westernpart of the study area. A counterclockwisetwist angle in the eastern portion indicates alocal stress attributed to drag where no saltwas available to detach the eastern edge of theplateau sheet. The clockwise change in stressorientation is consistent with the rotation instress orientation found in the anthracite beltof the Pennsylvania Valley and Ridge, but isopposite to the sense of rotation in the south-western portion of the detachment sheet(western Pennsylvania and West Virginia).The two regional rotation domains are sepa-rated by the Juniata culmination.

INTRODUCTION

In outcrops of the anthracite belt in easternPennsylvania, gradual twist hackles, a set offringe cracks growing in an en echelon arrange-ment following breakdown of a parent joint intosmoothly curving segments (Fig. 1), all propa-gated to a plane rotated clockwise from the par-ent joint in map view (Fischer et al., 1991). Thereis the possibility that these clockwise twist hack-les are further evidence for the clockwise sense ofchange in stress orientation attributed to progres-sive Alleghanian deformation within both thenortheastern Appalachian plateau (e.g., Engelderand Geiser, 1980) and the Appalachian Valleyand Ridge east of the Susquehanna River (e.g.,Nickelsen, 1979; Gray and Mitra, 1993). Thequestion for us is whether fringe cracks areorganized across a geologic province to the ex-tent that they provide a structural record of thetectonic history of a region.

One purpose of this paper is to document theareal distribution of fringe cracks in Devonian

clastic rocks of the Appalachian plateau detach-ment sheet in New York State. Very few parentjoints within the detachment sheet are boundedby fringe cracks. However, those parent jointswith fringe cracks leave behind a record that is ofgreat value in deciphering the tectonic history ofthe Appalachian plateau detachment sheet duringthe Alleghanian orogeny.

Fringe Cracks

Twist Hackles.Early descriptions of joint sur-faces divided them into a planar portion, a rim ofconchoidal fractures, and a fringe (Woodworth,1896). With the evolution of terminology overtime, these three parts of a joint surface are nowdescribed as the main joint face, with its character-istic plumose structure, conchoidal ridges (ribmarks of Kulander and Dean, 1985), and thefringe, with its en echelon fringe cracks (Hodgson,1961a, 1961b). The boundary between the mainjoint face and its fringe cracks may be either anabrupt transition known as a shoulder or asmoothly curving transition. Later descriptions offringe cracks implicitly recognized the geneticsimilarity between the abrupt and smooth transi-tions by referring to both types of en echeloncracks as twist hackles (Kulander and Dean,1985). The fringe is called a gradual twist hackle(Fig. 1) if individual cracks emerge from the tipline of the parent joint face in a smooth, uninter-rupted manner, whereas the fringe is known as anabrupt twist hackle (Fig. 2) if a series of planar enechelon cracks abut the joint tip line (Kulander et al., 1979, 1990; Kulander and Dean, 1995).Other names for these structures include hacklezone, hackle marks, dilatant fringe cracks, fingers,and fracture lances (for review of terminology, seePurslow, 1986; Kulander and Dean, 1995).

Gradual twist hackles result from a continuousbreakdown of the parent joint, whereas abrupttwist hackles stem from a discontinuous break-down. Pollard et al. (1982) provided a mechani-

219

Fringe cracks: Key structures for the interpretation of the progressive Alleghanian deformation of the Appalachian plateau

Amgad I. Younes*Department of Geosciences, Pennsylvania State University, University Park, Pennsylvania 16802Terry Engelder†

GSA Bulletin,February 1999; v. 111; no. 2; p. 219–239; 19 figures; 1 table.

}

*Present address: Department of Geology, RoyalHolloway, University of London, Egham, SurreyTW20 OEX, United Kingdom.

†E-mail: [email protected].

220 Geological Society of America Bulletin, February 1999

Length

Spacing

Overall propagation directions Parent joint

Fringecracks

Fringe cracks

Parentjoint

Hostbed

Parenttip line

Gradualtwisthackle

Si l ts tone

Width

Figure 1. Schematic of a gradual twist hackle. The parent joint and twist hackle are carried within a single host bed. The number of fringecracks decreases away from the parent joint. The spacing of the cracks changes from small at the tip line of the parent joint to large at the edgeof the twist zone.

Map view of thesiltstone-shaleboundary

1 m

Front view ofthe shale bed

Abrupt twisthackle

SiltstoneShaleParent joint

Figure 2. Schematic of an abrupt twist hackle. These fringe cracks can propagate downward or upward from a siltstone bed into adjacent beds.They have a characteristic sequence of large and small cracks, indicative of suppressed growth where crack stress shadows limit the growth ofsome cracks.

Geological Society of America Bulletin, February 1999 221

cal explanation for the two types of twist hacklesand suggested that twist angle is a function ofchange in remote stress orientation, stress magni-tude, and elastic properties. There is generalagreement that the breakdown of the parent joint

into fringe cracks is a consequence of a stressfield with principal components that are neitherorthogonal nor parallel to the tip line of the par-ent joint (e.g., Kulander et al., 1979; Pollard et al.,1982; Bankwitz and Bankwitz, 1984).

Kinks. The tip line of the parent joint can bedecorated with a single fringe crack known as atilt or kink (Fig. 3). A kink is a planar crack thatpropagates laterally at some angle from the edgeof a parent joint (Cottrell and Rice, 1980). Suchfeatures are associated with systematic jointsin Arches National Park, Utah (Cruikshank et al., 1991), and in the granites of the SierraNevada, California (Segall and Pollard, 1983).Kinks form after parent joints have been ar-rested. If the parent joint had not stopped priorto kink formation, there would have been asmooth curving or hooking of the crack path(Olson and Pollard, 1989).

In this paper the term fringe crack refers to anyout-of-plane crack that emanates from the tip lineof a parent joint (Fig. 4). The parent joint is typi-cally a planar, persistent, long crack that often be-longs to a systematic joint set. After propagatingsome distance, the parent joint may encounter astress field with principal components that are nei-ther parallel nor perpendicular to its plane. De-pending on its orientation relative to the remotestress, the parent joint may break down at its tipline to form the en echelon cracks of a twist hackle(Pollard et al., 1982) or it may tilt or deviate fromits path to form either a hook or kink (Cruikshanket al., 1991; Olson and Pollard, 1989). In this pa-per we apply the term fringe crack more broadly

than Hodgson (1961a), who referred only to atwist hackle when using the term.

Conditions for Development of Fringe Cracks

Joint propagation occurs in response to threeprincipal loading modes (e.g., Lawn, 1993) thatrepresent the configuration of the stresses at thejoint tip line (Fig. 5A). If a joint propagates in itsplane driven by a tensile stress perpendicular tothat plane, it propagates under mode I, or openingmode, loading. As long as the tensile stress re-mains normal to the joint, long planar joints formwithout breakdown at the tip line. In addition tomode I loading, the joint plane may be subject toa shear traction. This condition is known asmixed-mode loading. If the shear couple is di-rected parallel to propagation direction and nor-mal to the tip line, the additional loading is modeII; if the shear couple is perpendicular to thepropagation direction and parallel to the tip line,the additional loading is mode III. The applica-tion of shear tractions near the joint tip line forcesthe joint to deviate, following a path determinedby the sense and orientation of the shear couple.Hooks, abrupt kinks, and twist hackles are allmanifestations of mixed-mode loading at thejoint tip line.

Several criteria predict the joint path undermixed-mode loading (Broek, 1991; Lawn, 1993).For all these criteria, the joint follows the propa-gation path that minimizes the shear stresses act-ing on the joint tip and that maximizes the tensile

Mode I & II Loading Mode I & III Loading

Continuouspropagation

Discontinuouspropagation

Gradual twist hackle Abrupt twist hackleKink

(Figs. 1 and 8) (Figs. 2 and 9)(Figs. 3 and 10)

Continuouspropagation

Discontinuouspropagation

Curve

Parent joint

Mode I Loading

Figure 4. Flow chart indicating the loading conditions for a parent joint and four types of fringe cracks.

Kink angle (β)

Parent jointKink

SiltstoneKink

Figure 3. Schematic of two kinks propagat-ing from the lateral tip lines of a parent joint.These fringe cracks form as the joint tip line issubjected to mixed mode I and II loading.

FRINGE CRACKS IN THE APPALACHIAN PLATEAU


stress at the joint tip. The out-of-plane propaga-tion at a joint tip is controlled by the orientationof the local stress field, which is not necessarilythe same as the orientation of the remote stresses.Local stresses can arise through the interaction ofneighboring cracks, elastic mismatches, or the in-teraction of a remote stress and the existing joint.

Kinks. Kinks form as the parent joint tilts toaccommodate a component of mode II loading atthe joint tip (e.g., Cottrell and Rice, 1980; He andHutchinson, 1989). Kinks grow because the ori-entation of the remote stress field changed afterinitial joint propagation was arrested, and thus

later subjected the parent joint to mixed-modeloading (Figs. 3 and 5, B and C). After initiation,kink growth is in the direction of the maximumhorizontal stress; the kink angle,β, indicates theangular change in the orientation of the remotestress field. This condition reflects a temporalchange in the orientation of the remote stressfield (Engelder and Geiser, 1980). The tip line ofa systematic joint may also curve smoothly out ofplane because of local stress conditions such asthe interaction between joints (Olson and Pollard,1989) or the interaction of a joint with an inclu-sion (McConaughy and Engelder, 1999). This

condition represents a spatial change in orienta-tion of the local stress field.

Mixed-mode loading is the consequence of ei-ther a temporal rotation of remote stresses(Cooke and Pollard, 1996) or a spatial variationin the orientation of the local stress field aroundlocal structures (Muller and Pollard, 1977; Olsonand Pollard, 1989). In this paper we focus ontemporal changes in the remote stress-field orien-tation because it is this interpretation, not a spa-tial variation of local stress, that gives us an ex-planation for fringe cracks displaying the samesense of stress rotation over sizable regions of

YOUNES AND ENGELDER

Segmented fringe crackTilted fringe crack (A kink)Parent joint

Surface morphology

Propagation direction

Right lateral shearat the front tip of

the joint

Right lateral shearat the top edge of

the joint

Mixed modes I & IIMixed modes I & III

Mixed modes I & II Mixed modes I & III

Joint plane

Joint tip line

Mode I loading Mode II loading Mode III loading

C

A

B

Parent joint tip lines

Parent joint

Figure 5. (A) Joint loading modes. Theblack arrows indicate the propagation direc-tion of the crack tip, and the white arrows in-dicate the extension or shear. (B) Fringe crackgeometries. The addition of mode II loadingcauses the joint to kink or curve, whereas theaddition of mode III loading causes the jointfront to break down into several smallerfringe crack segments. Here the tilted fringecrack is a kink. (C) The edge of a parent jointdecorated with both twist hackles and a kink.


NY

PAOH

WV

Albany

Harrisburg

0 50 100

km

50

Catskill Mts.

BuffaloExtent of

Silurian saltStudyArea

Ithaca

Elmira

CortlandFillmoreGlen

NEW YORK

PENNSYLVANIA

81

88

N0 10 km

Stony Brook TaughannockFalls

Letchworth

Tully

390

Bath

Bing-hamton

WatkinsGlen

390Stuben Co.

Allegeny Co.

Broome Co.

77˚42˚

App

alac

hian

Valle

y&

Ridge

Appala

chian

Plat

eau

Cayuga

Lake

Seneca

Lake

Figure 6. Location map of the Finger Lakes district in New York State. The extent of the Silurian salt is shaded and outlined by dashes in thestates of New York (NY), Pennsylvania (PA), Ohio (OH), and West Virginia (WV).

the Appalachian plateau. Structures such assmoothly curving or hooking joint tips arisingfrom the interaction of a joint and an inclusionare not discussed in this paper, although they arepresent in the study area (e.g., McConaughy andEngelder, 1998).

Twist Hackles. In addition to a single kink-ing or curving fringe crack, en echelon fringecracks can also decorate the tip line of a parentjoint in the form of twist hackles (Figs. 1, 2, and5, B and C). Of several hypotheses that consti-tute the mechanical basis for explaining the lo-cal stress field that causes en echelon fringecracks (Engelder et al., 1993), we prefer mixed-mode (I and III) loading induced by a rotationof the remote principal stress direction (Kulanderet al., 1979; Pollard et al., 1982). This is a tem-poral rather than a spatial change in orientationof the stress field. After a rotation of the remotestress field, the joint follows a path that de-pends on several parameters, including the ori-entation and magnitude of the remote stresses,pore pressure, and the elastic modulus of therock in which the joint propagates (Pollard et al., 1982). Stress and pore pressure are some-times combined into one parameter called thestress ratio,R, which is the difference betweenthe internal fluid pressure,Pi, on the joint faceand the remote mean horizontal stress, dividedby the remote horizontal shear stress:

(1)

where σir are the remote principal stresses with

tensile stress taken as positive so that the maxi-mum compressive principal stress is designatedσ3

r. For natural hydraulic fracturing, pore pressureand internal pressure are equal at the initiation ofjoint propagation (Engelder and Lacazette, 1990).As joints grow, the internal pressure may drop be-low the pore pressure outside the joint but it willremain open as long as the internal pressure in thejoint exceeds the minimum compressive stress.The joint will recharge by flow down a pore pres-sure gradient (Engelder and Lacazette, 1990).

Because of the interaction of the remote stressfield with the joint tip, en echelon fringe crackspropagate at an angle relative to the parent joint.The twist angle,β, is a function of the stress ratio,R, the angular change of the remote stress orienta-tion,α, and Poisson’s ratio,ν (Pollard et al., 1982):

. (2)=+( )( )

0 52

2 0 5. tan

sin

cos . ––1

R vβ

Pir> – 1

RPi

r r

r r=

+ +2 1 3

1 3–


YOUNES AND ENGELDER

Wes

t Fal

ls G

roup

Java

fmW

est F

alls

Rhinestreet Shale

Angola ShalePipe Creek Shale

Hanover Shale

Nunda Sandstone and Shale

Gen

esse

Gro

upS

onye

aG

roup

Upp

er D

evon

ian

Middlesex ShaleCashaqua Shale

West River Shale

Ithac

a F

m

Genundewa Limestone

Penn Yan Shale

Geneseo Shale

Ham

ilton

Gro

up

Tully

Mid

dle

Dev

onia

n

Tully LimestoneMoscow Shale

Ludlowville Shale

Skaneateles Shale

Marcellus Shale

westernportion

easternportion

Onondaga Limestone

sandstone interbeddedshale and sandstone

limestone

gray shale black shale

Not

to s

cale

- R

hine

stre

et S

hale

is r

elat

ivel

y th

icke

r th

an s

how

n

Finger Lakes district

Figure 7. A simplified stratigraphic column of major rock units in the eastern and western por-tions of our study area in the Finger Lakes district of New York State (modified from Van Tyne,1983). The facies become progressively coarser to the east as the proximal parts of the Catskilldelta are approached. The unconformity at the base of the Tully Limestone extends up to the baseof the Cashaqua shale in western New York.

For small α (i.e., α < ~ 10°), fringe cracksform perpendicular to the rotated least compres-sive stress only when

. (3)

Otherwise, the twist angle,β, of the fringecrack relative to the parent joint will differ fromthe remote stress rotation angle,α. Here, a dis-tinction should be made between the twist angle,β, and the kink angle,β, with β always equal to αin the latter case. The development of gradualtwist hackles represents a continuous breakdownof the joint tip line, and indicates that α < ~ 10°,and R is relatively large (Cottrell and Rice, 1980;Pollard et al., 1982). Abrupt twist hackles reflect adiscontinuous breakdown of the parent joint andindicate either that α > ~ 10° or that α < ~10°, andR is relatively small (Cottrell and Rice, 1980;Pollard et al., 1982).

GEOLOGICAL CONTEXT

This paper documents the spatial distributionand nature of fringe cracks within the northeast-ern portion of the Appalachian plateau detach-ment sheet (Fig. 6). This portion includes the Fin-ger Lakes district of New York State, whichextends from Broome County to AlleganyCounty, and covers ~ 3000 km2. The detachmentsheet in this area consists of clastic sedimentaryrocks of the Devonian Catskill delta shed fromthe Acadian highlands to the east (Ettensohn,1985). The delta prograded from east to west, at-taining its maximum thickness east of the studyarea. Clastic rocks of the delta complex consist ofpackages grading from black and gray shalesthrough siltstone to sandstone. The clastic groups(Hamilton, Genesee, Sonyea, and West FallsFormations) are separated by three black shaleformations (Genesee, Middlesex, and Rhine-street Shales) that reflect reducing conditions atthe time of deposition (Fig. 7). The Tully Lime-stone is an important marker bed at the base ofthe Genesee Group.

The detachment sheet was deformed into a se-ries of low-amplitude folds mapped in outcrop(Wedel, 1932) and in the subsurface (Bradley andPepper, 1938; Murphy, 1981). Folds of the de-tachment sheet are broad, persistent, and can betraced farther to the southwest in Pennsylvaniaand into West Virginia. In the study area the foldaxes trend east-west in the eastern part and north-east-southwest in the western part. The detach-ment sheet contains layer-parallel shorteningstructures of Alleghanian age above a decolle-ment within the Silurian Salina salt (Engelder,1979; Murphy, 1981; Rodgers, 1970). Much of

Rv

v= +0 5

0 5

.

. –



Figure 8. Gradual twist hackles. A parent joint and a twist hackle are carried within a singlebed of siltstone of the Ithaca Formation at Taughannock Falls State Park, New York. Propaga-tion direction for the twist hackle is upward. The sense of stress field rotation in this example isclockwise. The scale to the lower right is divided into centimeters.

Figure 9. Abrupt twist hackles. This set of fringe cracks propagated downward into a thickshale bed from a thinner siltstone bed hosting the parent joint at Taughannock Falls State Park,New York. These rocks are part of the Ithaca Formation. The sense of stress field rotation in thisexample is clockwise. The scale is a geologic compass with an 8 cm base.

the layer-parallel shortening was accommodatedby pressure solution and the formation of solutioncleavage (Engelder and Geiser, 1979; Geiser andEngelder, 1983). The amount of slip on thedecollement is as much as 22 km to the north-northwest as estimated at the Allegheny front(Engelder and Engelder, 1977). The extent of thedetachment sheet is mapped by using strain mark-ers such as deformed fossils and from subsurfacedata (Engelder and Engelder, 1977; Engelder,1979; Engelder and Geiser, 1980; Beinkafner,1983; Geiser, 1988; Hudak, 1992). The forelandlimit of the detachment sheet and the region offolds in the post-Silurian rocks coincides with thelimits of the Silurian salt (Frey, 1973).

Joints of Alleghanian age strike approxi-mately normal to the axes of the folds within thedetachment sheet (Parker, 1942; Ver Steeg,1942; Nickelsen and Hough, 1967; Engelder andGeiser, 1980; Geiser and Engelder, 1983; Bahat,1991; Lacazette and Engelder, 1992). Thesewere first mapped as dip joints (i.e., parallel tothe dip direction of bedding) by Sheldon (1912),who recognized that outcrops commonly con-tained dip joints in more than one orientation.Multiple sets of dip joints are particularly welldeveloped in interlayered siltstone-shale beds,where the earlier dip joints favored siltstonebeds. It is these dip joints, called cross-fold jointsby Engelder and Geiser (1980), which are occa-sionally decorated with fringe cracks.

Tectonic Problem

In an analysis of the Bear Valley strip mine inthe Appalachian Valley and Ridge, Nickelsen(1979) recognized a group of structures thatwere a manifestation of a clockwise rotation ofmaximum horizontal stress (SH) during the Al-leghanian orogeny. Engelder and Geiser (1980)recognized that jointing on the Appalachianplateau also reflected a clockwise rotation of theAlleghanian stress field. On the basis of rocksthat had two distinct cleavages and outcropscommonly carrying two dip joint sets, Geiserand Engelder (1983) concluded that the Appala-chian plateau was affected by two discrete tec-tonic phases: the Lackawanna and Main phases.However, outcrops in the anthracite coal districtof the Appalachian Valley and Ridge of Pennsyl-vania, including the Bear Valley strip mine, indi-cate that the rotation of the Alleghanian stressfield produced structures carrying a broad rangeof orientations. From these structures, Gray andMitra (1993) concluded that the Alleghanianorogeny was a continuous series of structuralevents reflecting a gradual clockwise rotation inthe Alleghanian stress field, rather than beingpunctuated by two tectonic phases as suggested


YOUNES AND ENGELDER

Figure 10. Kinks. This example of fringe cracks in a siltstone bed of the Ithaca Formation isfound near Whitney Point, New York. The shale above and below the siltstone bed carriesabrupt twist hackles. The sense of stress field rotation in this example is counterclockwise, bothfor the kink and the abrupt twist hackle. The parent joint is found in a bed roughly 20 cm thick.

2

Ithaca

Elmira

Cortland

FillmoreGlen

Stony Brook

Syracuse

NY

PA

N

Areal Distribution of Gradual Twist Hackles

Orientation of parent joint

Amount of rotation

Sense of rotation

3

-6

-3

3

1-7

-3

-5-7

-12

-10

106

155 4 3

5

3

-2-7

1

2

3

3

TaughannockFalls

Letchworth

3

3

4

4

-6

-5

-3

5

4 7

Fi

0 10 km

77˚42˚

Figure 11. Areal distribution of fringe cracks. The amount of rotation shown is the average ofall fringe crack sets at every station. CW—clockwise; CCW—counterclockwise.

A

by Geiser and Engelder (1983). Bahat (1991)and Evans (1994) presented evidence for struc-tures in more than two orientations suggestingthat the northeastern portion of the Appalachianplateau detachment sheet was also deformed bya more continuous clockwise rotation of the Al-leghanian stress field. If so, the tectonic evolu-tion of the Appalachian plateau detachmentsheet is consistent with that found in the Valleyand Ridge.

FRINGE CRACKS OF THE DETACHMENT SHEET

Within the Appalachian plateau detachmentsheet, we identify four types of fringe cracks(Fig. 4). Three of these types are discussed in thispaper: gradual twist hackles, abrupt twist hack-les, and kinks. We refer to the angle between theparent joint and the fringe crack as clockwise orcounterclockwise, depending on the sense of ro-tation going from the parent to the fringe crack inmap view. The fourth type is more difficult to rec-ognize and is not discussed here.

Gradual Twist Hackles

On the Appalachian plateau, gradual twisthackles occur within the same bed and lithologythat hosts the parent joint (Figs. 1 and 8). The sur-faces of a gradual twist hackle occasionally showa plumose structure that indicates an overall ver-tical propagation direction. This surface mor-phology is often continuous with that of the par-ent joint, giving no indication that the tip line ofthe parent joint arrested before the gradual twisthackle propagated. Upon breakdown at the jointtip line, individual fringe cracks twist away fromthe parent as they realign normal to the localmaximum tensile stress. Hence, the hackle is aseries of en echelon fringe cracks that initiatefrom the plane of a parent joint and twist out ofthe plane by propagating normal (i.e., vertically)to the overall lateral propagation direction of theparent joint.

As fringe cracks grow, stress shadows de-velop to suppress the growth of adjacent cracks(Nemat-Nasser and Oranratnachai, 1979; Grosset al., 1995), giving rise to a characteristic pat-tern exhibited by gradual twist hackles (Figs. 1and 8). The frequency of fringe cracks decreaseswith vertical distance away from the parent joint(Helgeson and Aydin, 1991). Gradual twisthackles can propagate downward, upward, or inboth directions within the host bed. With veryfew exceptions, gradual twist hackles within theAppalachian plateau detachment sheet propa-gate toward the top of the bed after the parentjoint passed through the lower portion of the



Ithaca

Elmira

Cortland

FillmoreGlen

Stony Brook

Syracuse

NY

PA

11

Areal Distribution ofCW Abrupt Twist Hackles

Boyd Point

Vector mean

Lateral fringe cracks

Parent joint

0 10 km

N

77˚42˚

Figure 11. (Continued).

B

Ithaca

Elmira

CortlandFillmoreGlen

Stony Brook

Syracuse

NY

PA

Vector mean

Lateral fringe cracks

Parent joint

Areal Distribution ofCCW Abrupt Twist Hackles

Letchworth

TaughannockFalls

0 10 kmN

77˚42˚

C

stratum. Gradual twist hackles occur on bothdip joints and on later systematic joints striking070° within the study area and are equally com-mon in sandstones, siltstones, and shales.

Abrupt Twist Hackles

On the Appalachian plateau, abrupt twist hack-les are commonly found in shale and initiate fromthe tip line of dip joints in siltstone beds that aretypically less than 30 cm thick. Individual en ech-elon cracks that are planar and that emanateabruptly from the edge of a parent joint are knownas abrupt twist hackles (Figs. 2 and 9). Becauseabrupt twist hackles abut the parent joint at apoint, they appear in map view to grow across theedge of the parent joint. In contrast with gradualtwist hackles that grow within the same bed as theparent joint, abrupt twist hackles are separatedfrom parent joints by a bed boundary. These enechelon cracks are equally common propagatingupward, downward, or in both directions fromsiltstone beds (cf. Helgeson and Aydin, 1991).

Fringe cracks rarely exceed one meter inlength (i.e., the dimension parallel to propaga-tion, which is vertical for twist hackles), and havea spacing that correlates well with the cracklength. This spacing is governed by the stressshadows of the largest fringe cracks and formsthe familiar pattern of suppressed growth foundin gradual twist hackles. As a rule of thumb, spac-ing of the largest fringe cracks is about half theirvertical dimension, whereas joint spacing in theparent beds is typically 30% larger than their ver-tical dimension (Loewy, 1995). Within an abrupttwist hackle, all the fringe cracks are usually par-allel. However, in a few examples, smaller crackstend to have a larger twist angle from the parentthan the larger cracks.

Abrupt twist hackles are best developed in theIthaca Formation of the Genesee Group. It is pos-sible that the distinct shale and siltstone interlay-ering of the Ithaca Formation provides a favor-able lithologic contrast for the initiation of theabrupt twist hackles. Field evidence indicates thatsiltstones are “fracturable” at an earlier time andshale at a later time, between which times a largestress rotation occurred (Engelder, 1985). A rota-tion of the remote stress field at some later timewould produce the mixed-mode loading condi-tions necessary for the twist hackles to propagatein the adjacent shale beds.

Kinks

In the Appalachian plateau detachment sheet,kinks occur exclusively in siltstone beds. If twokinks occur with a parent joint, both show thesame sense and magnitude of rotation relative to


YOUNES AND ENGELDER

Ithaca

Elmira

Cortland

FillmoreGlen

Stony Brook

Syracuse

NY

PA

Taughannock Falls

345

000

019

041

020

005

019

355

007

344340

355

334

020

002

350015

350

358

023

356

013

007

344

Orientation of fringe crack

Parent jointLocation of data point

Orientation of parent joint

Fringe crack

Areal Distribution of Kinks

341

001

339

358

Letchworth

0 10 kmN

77˚42˚

Ithaca

Elmira

CortlandFillmore Glen

Stony Brook

Syracuse

NY

PA

7

15

13

4

8

15

15

1

3

912

Areal Distribution ofGradual Twist Hackles

on 070 ° Joints

Parent joint10Amount of rotation

Fringe crack

3 3 7

5

0 10 kmN

77˚42˚

Figure 11. (Continued).

D

E

the parent (Figs. 3 and 10). Some parent jointswith kinks are decorated with abrupt twist hacklesin the adjacent shale beds. These flanking fringecracks show the same sense of angular twist as thetilt of the lateral kinks. No evidence for slip wasobserved on any parent joint carrying kinks withinthe study area. Kinks occasionally show a surface

morphology of plumes that are continuations ofthe plume morphology from the parent joint. Thisindicates that kinks occur as mode I cracks andthe tilt in a crack path is driven by a shear tractionsuperimposed on the parent joint after arrest fol-lowing a finite amount of mode I propagation (Er-dogan and Sih, 1963; Cottrell and Rice, 1980).

AREAL DISTRIBUTION OF FRINGE CRACKS AND PARENT JOINTS

Orientation of Fringe Cracks Relative to Parent Joints

More than 225 sets of fringe cracks weremapped in the northeastern portion of the Appa-lachian plateau detachment sheet within the Fin-ger Lakes district of New York State. Althoughthere are millions of joints throughout the de-tachment sheet, joints decorated with fringecracks are rare. The number of joints decoratedwith fringe cracks within individual outcropsvaries from one to more than several dozen. Fig-ure 11 (A–E) shows the orientation and areal dis-tribution of various types of fringe cracks andtheir parent joints. The maps plot the data froman outcrop by presenting the average strikes of allfringe cracks and their parent joints. Fringecracks are found in rocks of all groups and mostlithologies mentioned in Figure 7.

The general orientation of all fringe crack setsfollows the systematic change in fold trend andstrike of the parent joints from east to westthrough the detachment sheet. In general, de-pending on the type of fringe crack, the magni-tude of the angle between parent joint and fringecracks remains about the same across the area.However, there is a pattern to the sense of fringecrack rotation across the region. To the east ofIthaca, fringe cracks strike counterclockwisefrom the parent joints, to the west they strikeclockwise from their parents, and in the vicinityof Ithaca, both senses of rotation are found. Be-cause the loading conditions and timing of prop-agation are different for each fringe crack type,there are characteristic differences in twist or tiltangle. Kinks are concentrated in the eastern halfof the study area and tend to strike within a fewdegrees of 002°, seemingly independent of orien-tation of the parent (Fig. 11D).

The angle of twist or tilt (β) of a fringe crackset changes according to the host rock, the type ofthe fringe crack, and the sense of rotation. Thetwist angles for clockwise twist hackles are gen-erally smaller than for counterclockwise twisthackles (Fig. 12). Gradual twist hackles havesmall twist angles, with a mode at 3° for clock-wise fringe cracks and 6° for counterclockwisefringe cracks. Abrupt twist hackles have largertwist angles, with a mode at 13° for clockwisefringe cracks and 22° for counterclockwise fringecracks. Kinks have tilt angles with a mode of 16°for clockwise fringe cracks and 14° for counter-clockwise fringe cracks.

The maximum twist angle (β) for gradual twisthackles varies according to the lithology hostingthe parent joint. This is seen in a plot of the strike



0 10 20 30 40 500

5

10Mode=14˚N=14

Fre

quen

cy

Tilt angle (°)0 10 20 30 40 50

0

5

10

CW kinks

Fre

quen

cy

Mode=16˚N=22

Tilt angle (°)

CCW kinks

0 10 20 30 40 500

5

10

Fre

quen

cy

Mode=22˚N=29

0 10 20 30 40 500

5

10

Fre

quen

cy Mode=13˚N=47

CW abrupt twist hackles

0 10 20 30 40 50

60

40

20

0

Fre

quen

cy

Mode=3˚

CW gradual twist hackles

N=67

Twist Angle (°)

CCW gradual twist hackles

0 10 20 30 40 50F

requ

ency

Mode=6˚N=39

60

40

20

0

Twist Angle (°)

CCW abrupt twist hackles

Twist angle (°) Twist angle (°)

Figure 12. Histograms of the magnitude of clockwise (CW) and counterclockwise (CCW) ro-tations for gradual and abrupt twist hackles and kinks. The number of fringe sets and the sta-tistical mode of orientation is shown inside each histogram. The amount of CCW rotation fortwist hackles is larger than the CW rotation. Kinks have a more consistent orientation (mode isalmost the same for CW and CCW rotations).

of fringe cracks versus the strike of parent joints(Fig. 13). In this plot, data for counterclockwisefringe cracks would plot below a line with a slopeof one, whereas data for clockwise fringe cracksare plotted above this line. By drawing delimitingenvelopes parallel to this line, we find the maxi-mum twist angles in coarser beds. For example,twist hackles in shale layers have a maximumtwist angle of 4°, whereas those in sandstone lay-ers reach 10° for fringe cracks with a clockwisetwist angle.

Sequence of Dip Joint Development

Throughout the detachment sheet, evidence forthe timing of joint development includes curvingor kinking of joints as the younger approaches theolder or abrupt termination of younger jointsagainst older joints (Fig. 14). Cases of unambigu-ous abutting are less common than mutual cross-cutting. Nevertheless, abutting between dip (i.e.,parent) joints in the detachment sheet shows aconsistent relationship from west to east wherethe later dip joints, particularly in the western por-tion of the study area, strike a few degrees clock-wise from earlier dip joints (Fig. 15). Younger dipjoint sets are defined by the horizontal clusteringof data in Figure 15. Abutting dip joints cluster atstrikes of 342°, 351°, and 003°, suggesting thatyounger dip joints have a consistent orientationthroughout the region despite predecessors ofwidely varying orientation. The earliest dip jointset, a spectrum of joints in the range of 320°–330°, do not cluster. The same appears to be trueof a dip joint set striking in the range of 006° to021° in the easternmost portion of the study area.These data indicate that joints striking at 342°generally abut joints striking at 320°–330° and351° joints generally abut 342° joints, and soforth. The exception to this general rule for aclockwise sequence of younger joints is found atthe eastern edge of the study area, where 003°joints abut parent joints striking roughly 020°, in-dicating a counterclockwise sense of rotation ofthe stress field in this area with time. Based onthese observations, the sequence of dip joint de-velopment in the detachment sheet is as follows,with parentheses indicating sets of joints clusteredabout that orientation:

.352 003→ °( ) → °( )

320

342

° °( ) →° °( )

°( )

–330–021

East Ithaca

West Ithaca

006


YOUNES AND ENGELDER

310

320

330

340

350

360

315 320 325 330 335 340 345

Siltstone

Parent joint

Maximum twistAngle = 6°

Strike of parent joint (°)

Ave

rage

str

ike

ofC

W g

radu

al tw

ist h

ackl

e (°

)

310

320

330

340

350

360

010

020

300 310 320 330 340 350 360 010 020

Sandstone

Parent joint


Strike of parent joint (°)

Ave

rage

str

ike

ofC

W g

radu

al tw

ist h

ackl

e (°

)

310

320

330

340

350

360

010

020

300 310 320 330 340 350 360 010 020

Shale


Parent joint

Strike of parent joint (°)A

vera

ge s

trik

e of

CW

gra

dual

twis

t hac

kle

(°)

Figure 13. Change in the angle between parent joint and the clockwise (CW) gradual twisthackles according to lithology. The maximum twist angle increases as the clastic grain size be-comes larger.



Other post-Alleghanian joint sets are alsofound throughout the detachment sheet. Onepost-Alleghanian 070° set is best developed inblack shale below the Rhinestreet Formation(Loewy, 1995). Where present, fringe cracks arealways counterclockwise from these parent joints(Fig. 11E). The 070° joints are cut by MiddleJurassic kimberlite dikes (Kay et al., 1983) in the

vicinity of Ithaca, and therefore, are older thanMiddle Jurassic. Another post-Alleghanian 000°set consists of short, curviplanar to planar jointsthat occasionally abut earlier 070° joints. The000° set is the youngest joint set documented inthis paper. These joints strike parallel to theMesozoic kimberlite dikes (Martens, 1924;Parker, 1942).

DISCUSSION

Arguments for a Temporal Change in Stress Orientation

Fringe cracks originate when an advancing jointsurface encounters a shear couple. If the advancingjoint enters a volume of rock subject to a stressfield misaligned relative to that guiding the parentjoint, transverse adjustments in crack propagationdue to mixed-mode loading cause the joint tobreak down into fringe cracks (Lawn, 1993). If themisaligned stress field was present prior to jointpropagation, the misalignment is spatial. If themisalignment took place after propagation and ar-rest of the parent joint, the misalignment is tempo-ral. Evidence suggests that within the Appalachianplateau detachment sheet, the shear couple wasimparted during a temporal misalignment whenthe regional stress field was rotated about a verti-cal stress axis,Sv. The strongest evidence for thisprogressive misalignment of the remote horizontalstresses,SH and Sh, is the regional distribution offringe cracks with a uniform sense of twist or tilt.

Our data show regional trends in fringe crackgeometries; with clockwise twist and tilt anglesfound in the west and counterclockwise twist andtilt angles most common in the east (Fig. 11).This regional distribution of fringe cracks is con-sistent with a temporal change in orientation ofthe remote stress field. The distribution is alsouniform throughout a layered sequence of sand-stones, siltstones, and shales. A nonuniformsense of rotation within a layered sequencewould have indicated a spatial change in localstress orientation rather than a temporal changein orientation of the remote stress field. In fact, itis difficult to imagine a local mechanism causingsuch consistently uniform spatial changes in thestress field from one bed to the next and one out-crop to the next. Even in outcrops such as those atTaughannock Falls, where we find both senses oftwist angle, abutting within the outcrops suggestsa connection to a broader regional remote stresshistory with clockwise fringe cracks developingprior to counterclockwise fringe cracks (Fig. 14).

The apparent lack of structures at the margin ofparent joints and the continuous plumes tracingfrom the parent joints to the gradual twist hacklesare curious. One interpretation is that the fringecracks propagated as a continuous rupture fromtheir parent joint without arrest. If the rupturepropagation from parent to fringe crack was unin-terrupted and continuous, and mixed-mode load-ing was imposed during the joint propagationevent, the front end of the parent joint shouldcurve or hook in response to the same mixed-mode loading. Curving and hooking of parentjoints were not found in conjunction with gradual

Siltstone

Shale 350 parent

340 parent

Pavement

N

350

50 cm

340

340

340350N

Shale

Siltstone

340

350

350

340

50 cm

N

A

B

340 parent 350

Figure 14. A schematic of parent joint interaction in Taughannock Falls, New York. (A) Jointswithin the siltstone layers are straight, except those that propagate in the shale curve or kink asthey approach the older 340° joints in the siltstone. (B) Tip fringe cracks forming as the tip ofthe 350° joint in the shale abruptly kinks as it approaches the 340° joint in the siltstone. Arrowsin the three-dimensional diagram indicate the propagation direction of the 350° joint.

TABLE 1. MECHANICAL DATA

Conditions at a depth of 3–4 km Field data Conditions at gradual-abrupt boundaryLithology Twist Sv SH Sh Pi R α v β Kic R– Internal Driving Joint

hackle (MPa) (MPa) (MPa) (MPa) (°) (°) (MPa= boundary pressure stress at height attype m1/2) at R- R- R-

boundary boundary boundary(MPa) (m)

Shale Gradual –78 –78.8 –77.5 81.5 5.15 10 0.14 4.43Abrupt –78 –78.8 –77.5 78.8 1.00 10 0.14 13.05 1.5 1.77 79.30 1.80 0.22

Sandstone Gradual –78 –80.5 –74.5 81.5 1.33 10 0.06 9.44Abrupt –78 –80.5 –74.5 78.8 0.43 10 0.06 14.76 2 1.27 81.31 6.81 0.03


twist hackles. The parent joints are planar, indi-cating that joint propagation ran through the lowerportion of the bed without leaving its plane andwithout breaking up through to the top bedboundary (Figs. 1 and 8). Consequently, our inter-pretation is that joint propagation along the lowerportion of the bed was arrested within the bed be-fore a temporal change in the remote stress fieldorientation. Only later, as fringe cracks grow in re-sponse to mixed-mode loading, does the jointbreak through to the upper bed boundary. Perhapsthe reason that gradual twist hackles are not morecommon in siltstone beds is that in most cases theinitial rupture of the parent joint broke through tothe upper contact with shale, leaving no opportu-nity for later gradual twist hackle development.

Such temporal changes in remote stress orien-tation have been modeled experimentally (Cookeand Pollard, 1996). During these experiments,propagation of a mode I crack was stopped be-fore mixed-mode loading was imposed. This isour model for gradual twist hackle growth on theAppalachian plateau. However, from the descrip-tion of the experiments, it is not clear whetherreinitiation of crack propagation was continuousor discontinuous. In short, data on twist and tiltangles point to a temporal change in the orienta-tion of remote stress rather than a spatial changein the orientation of the local stress on a bed-by-bed basis.

Twist Angle and the Role of Lithology

Although twist hackles reflect the sense ofthe remote stress field rotation, fringe crackspropagate normal to the local least compressivestress, which is often not in the same orientationas the remote least compressive stress,σ1

r

(Pollard et al., 1982). The type of breakdown(continuous or discontinuous) at the joint tip de-pends largely on the magnitude of the remotestress rotation,α, the stress ratio,R, and theelastic modulus of the rock. The gradual twisthackles in the Appalachian plateau detachmentsheet resemble those described by Pollard et al.(1982) as a continuous or smooth twist zone,

YOUNES AND ENGELDER

West of Ithaca

320

340

360

20

40

300 320 340 360 20 40

300

Older joint orientation (°)

320

340

360

20

40

300 320 340 360

East of Ithaca

N = 73

N = 10

N = 26

N = 28300

20 40

You

nger

join

t orie

ntat

ion

(°)

You

nger

join

t orie

ntat

ion

(°)

Older joint orientation (°)

342˚ Parent joints351˚ Parent joints

003˚Parent joints

003˚ Parent joints

351˚Parent joints

Figure 15. Abutting relationships west and east of Ithaca, New York. The diamond symbolsabove the straight line indicates clockwise abutting relationship, whereas those under the lineare counterclockwise rotations. The abutting of joints west of Ithaca is dominated by clockwiserotations, but mixed rotations are found east of Ithaca.


Zoneof

drag

NewYork

325342 351

020

003

336 006

Anthracitebelt

AppalachianPlateau

Lackawannasyncline

79˚42˚

NJPA

Figure 17. Deformed-state grid given by thedisplacement field determined from finitestrain as indicated by deformed crinoidcolumnals and cleavage (adapted from Geiser,1988). Dashed lines are traces of the fold axes.The samples with numbers indicate the trendsof joint sets from the Appalachian Plateau(this study) and data on the orientation of theAlleghanian stress field in the anthracite beltof Pennsylvania (Gray and Mitra, 1993).

0

5

10

15

20

25

0 5 10 15 20 25

R = 1.5

R =

0

R = 0

.5

R = 2.5

α = βν = .14ν = .06

Remote stress field rotation, α (˚)

Frin

ge c

rack

twis

t ang

le, β

(˚)

β < α

β >

α

Gra

dual twis

t

f ield

Ab

rup

t tw

ist

f iel

dshalesandstone

R = 5.15

(sha

le)

(san

dsto

ne) ∆Pi = -2.7 MPa

R = 1.33

R =

0

R = 0.5

R = 1.5

R = 2.5

Figure 16. Plot of the twist angle of afringe crack relative to the correspondingrotation of the remote stress field (adaptedfrom Pollard et al., 1982). The curves areplotted for different stress ratios (R), and forPoisson’s ratios of 0.06 (representing sand-stone) and 0.14 (representing shale). See textfor explanation. Log data from the detach-ment sheet indicates that νshale = 0.14 andνsandstone= 0.06 (Plumb et al., 1991).

where β < α (i.e., the twist angle of the fringecrack is less than the rotation angle of the re-mote stress). In contrast, abrupt twist hackles inthe detachment sheet are the field analogs of thediscontinuous or rough zone (Pollard et al.,1982), where b > a. We are encouraged that thetheoretical explanation for gradual and abrupttwist hackles offered by Pollard et al. (1982) ap-plies to the detachment sheet because their the-ory predicts that the twist angle of abrupt twisthackles should be larger, as is found in theAppalachian plateau (Fig. 12). The followingdiscussion gives an interpretation of the param-eters that might have controlled the twist angleof twist hackles as a function of lithology withinthe detachment sheet.

Effect of Elastic Properties.The differences

in β as measured on gradual twist hackles withinshale, siltstone, and sandstone beds is striking(Fig. 13). According to equation 2, if both sand-stone and shale beds are subject to the same Rand a regional α, the twist angle will depend onthe value of ν for each bed. Thus, in the detach-ment sheet, joints in beds with lower ν shouldbreak down with smaller twist angles. Geophysi-cal log data indicate that sandstones and silt-stones from the Catskill delta have a lower ν thanthe shales (Plumb et al., 1991). Consequently, thebeds of coarser clastic rocks should contain jointswith a smaller twist angle. Based on our observa-tion that joints in sandstone beds have the largertwist angles (Fig. 13), a dependence on ν can beruled out as the sole explanation for the variationof twist angle with lithology.

Effect of Stress and Pore Pressure. Appar-ently, elastic properties are subordinate to differ-ential stress and pore pressure within the threelithologies in controlling twist angles. If so, wecan estimate the relative magnitude of pore pres-sure using the rotation angle of gradual twisthackles in different lithologies and some reason-able assumptions about principal stresses. We al-ready know some characteristics about in situconditions. First, the surface morphology on par-ent joints is consistent with joints driven by highinternal pressure, where Pi > |Sh| (Lacazette andEngelder, 1992). Second, stress measurementswithin the Appalachian plateau detachment sheetshow that sandstones carry a higher differentialstress than shale interbeds (Evans et al., 1989).This is common in other sandstone-shale se-quences as well (e.g., Warpinski, 1989). Assum-ing that all beds were affected by the same re-gional α under these conditions, equation 2predicts that the twist angle in sandstone layersshould be larger than the twist angle for shale.This is consistent with our data showing thatβshale≤ 4°, and βsandstone≤ 10° (Fig. 13).

To estimate pore-pressure conditions for twisthackle development we assume that Pi is thesame in adjacent beds of sandstone and shaleand that α = 10° in all beds. Estimates of burialhistory suggest that twist hackle propagationtook place at a depth of 3–4 km within the de-tachment sheet (cf. Evans, 1995), so we assumethat σ2

r (Sv = ρgz) ≈ –78 MPa. Based on theanalysis of calcite twinning, the differentialstress in sandstone was estimated to be 6 MPaduring the Alleghanian orogeny (Engelder,1982). This stress difference is about 50% lessthan that measured in sandstone beds but close tothat measured in shale of the detachment sheetusing hydraulic fracture (Evans et al., 1989).Layer-parallel shortening and vertical jointingrequires that SH < Sv < Sh (tensile stress is posi-tive) during development of the gradual twisthackles. Hence, to satisfy this condition we setσ1

r (Sh) = –74.5 MPa and σ3r (SH) = –80.5 MPa

in the sandstone using the more conservative es-timate of stress from calcite twinning. We nowuse equations 1 and 2 to calculate a Pi that willyield β ≈ 10° (see Table 1). For σ1

r – σ3r = 6 MPa

in the sandstone and a crack-driving stress of 7 MPa (i.e.,Pi = 81.5 MPa) the result is R= 1.33and a twist angle of 9.4° (Figs. 13 and 16).

Having established the local pore pressure(i.e.,Pi = 81.5 MPa), we can now calculate thestress in the shale interlayers that will lead togradual twist hackles having β ≈ 4°. A crack-driving stress of 4 MPa (i.e.,Pi = 81.5 MPa) andR= 5.15 gives a twist angle of 4.4° (Figs. 13 and16). Here we find that σ1

r (Sh) = –77.5 MPa andσ3

r (SH) = –78.8 MPa for a differential stress of


YOUNES AND ENGELDER

315353

NY

PAOH

WV

Juniata culmination Albany

Lackawannasyncline

Harrisburg

350330

3

300320

7

310330

6

323300

320 003

2

336 355

1

9

050

km

Extent ofSilurian salt

336

4

006

8

290

320

5

003021

10

340

11

2

322

12

100 200

42°42°

77°

77°

Figure 18. A model for the rotations of the Alleghanian stress field in the Central AppalachianMountains. The heavy black line is the boundary between the Valley and Ridge and AppalachianPlateau. The short straight segments indicate the first and last stages of deformation determinedby the following authors indicated by the corresponding number: 1, Engelder and Geiser (1980);2, this paper; 3, Nickelsen (1979); 4, Gray and Mitra (1993); 5, Nickelsen (1988); 6, Nickelsen andHough (1967); 7, Dean et al. (1984); 8, Dean et al. (1988); 9, Evans (1994); 10, Nickelsen (1996);11, Spiker and Gray (1997); 12, Nickelsen and Engelder (1989). The thin arrows indicate the lo-cal sense of rotation, and the thick arrows indicate the regional sense of rotation proposed in thisstudy. The heavy dashed line marks the location of the Juniata culmination.



1.3 MPa in the shale. For both the sandstone andshale layers, the internal pressure,Pi, must ex-ceed the absolute value of both σ1

r and σ3r in or-

der for R to be consistent with the formation ofgradual twist hackles. The significance of this re-sult is that stress associated with the relative dif-ference between gradual twist angles in sand-stone and shale is consistent with the relativedifference between stress in sandstone and shaleas measured using other techniques such as hy-draulic fracturing. Note also that the higher valueof σ1

r in the sandstone beds leads to early jointdevelopment in these beds rather than in the in-terlayered shales. The earliest jointing is foundwithin the siltstone and sandstone beds of the de-tachment sheet (Engelder, 1985).

Our view of the development of gradual twisthackles requires that the rate of joint propagation

is high relative to the rate of change in orienta-tion of a regional stress field. If so, initial propa-gation must run along the bottom of some bedsand arrest near the central portion of some bedsto await a change in stress orientation. Arrest,which must take place before the entire bed isruptured, occurs upon a decrease in internalpressure (Lacazette and Engelder, 1992). How-ever, cycling of internal pressure to reinitiatepropagation of twist hackles favors the develop-ment of abrupt twist hackles. Following the ar-rest of the parent joint it is assumed that Pi < |σ1

r|,whereby R < –1 (equation 1). As the joint isrecharged and internal pressure increases towardthe propagation condition,Pi > |σ1

r|, the stressratio traverses through the field of abrupt twisthackles as defined by –1 < R < ~1.5 dependingon the lithology (unshaded area of Fig. 16). Ac-

cording to equation 3, the boundary betweenabrupt and gradual twist hackles is R= 1.27 for asandstone. For gradual twist hackles to form insandstone beds, the internal pressure must in-crease enough to drive Rupward above 1.27 andinto the shaded area of Figure 16 without reiniti-ating propagation of the parent joint. When R = 1.27, the joint driving stress is 6.8 MPawithin the sandstone. In a sandstone bed with afracture toughness (Kic) of 2 MPa • m1/2 (Scottet al., 1992), an elliptical joint with a short axisof ~6 cm is required to hold the internal pressurewithout propagation to allow the condition R > 1.27 and the formation of gradual twisthackles. This approaches the vertical dimensionof the parent joints decorated with gradual twisthackles (Fig. 8). For shale with Kic = 1.5 Mpa •m1/2, the development of gradual twist hackles

Ib Ia IIINY 2

Engelder andGeiser (1980)

Parker (1942)

Gray andMitra (1993)

Sheldon (1912)

Geiser andEngelder (1983)

Nickelsen andHough (1967)

Nickelsen (1979)

I III IINY -- --

PA 5

270 9060300330300

Orientation of structures and tectonic events

Author

Dip StrikeNY -- --

Ib IaIII

NY&PA 2

Stages I - V

Non-systematic

AB D EC

SystematicPA 6

PA 6

6

NY 4070¡Dip joints & fringe cracks

Evans (1994)I

WV, PA & OH

5

Dean et al. (1984) II IWV &MD

3

This study

Dean et al. (1988) WV 2

123 5

Figure 19. Selected data on the sense of rotation of the regional stress field in the central Appalachian Mountains during the Alleghanian orogeny.

requires a joint driving stress of 1.8 MPa, whichis too large relative to that required by the di-mension of parent joints in shale of the MoscowFormation of the Hamilton Group (Fig. 7).

When σ1r – σ3

r is low, R is very sensitive tosmall decreases in internal pressure. As jointsgrow, the joint tip stress intensity increases, thusrequiring a smaller crack-driving stress for con-tinued propagation. Hence, the internal pressurerequired to drive joint propagation would drop asa natural consequence of joint growth (Engelderand Lacazette, 1990). The larger twist angles forabrupt twist hackles (Fig. 9) may arise as a conse-quence of this sensitivity to a change in Pi. For ex-ample, using the same values for stress as givenhere, a drop in Pi of 2.7 MPa within the shale bedswill drop R from 5.15 to 1.00. With R≈ 1.0, twisthackles are abrupt and the resultant twist angle is13° in shale (see Fig. 12) when α = 10° (see arrowin Fig. 16). The same drop in internal pressurewithin adjacent sandstone beds has less effect onthe twist angle but nevertheless leads to abrupttwist hackles in sandstone (Figs. 12 and 16). Theformation of both gradual and abrupt twist hack-les within the Appalachian plateau detachmentsheet is consistent with the long-term decrease ininternal pressure as the size of parent joints be-come larger.

Progressive Development of the Regional Stress Field

Pattern of Parent Joint Sets. Data on abuttingof parent joints helps resolve the issue of whetherthere are more than two joint sets extending sev-eral hundred kilometers along strike in a fan-shaped pattern as proposed by Engelder andGeiser (1980). If the early joints in the TullyLimestone at Taughannock Falls correlate withthe 320°–330° set found more than 100 km to thewest, then the earliest 320°–330° set does not fan,but rather is an indication that the same joint setappears in different lithologies in different loca-tions. If the 342° set in the western portion of thestudy area is contemporaneous with the 006° to021° joint set in the east, then we have identified ajoint set that is found in a fan-shaped pattern. Su-perimposed on this fanning pattern are two morejoint sets, each differing in strike about 10°. Indi-vidual sets are best seen in the data for abutting ofparent dip joints (Fig. 15), where later joints clus-ter at 342°, 351°, and 003°. This interpretation ofjoint pattern development is consistent with thatpresented by Nickelsen and Hough (1967), wherejoints of one set cluster about one orientationrather than fanning. However, we cannot rejectthe interpretation (Engelder and Geiser, 1980) thata single joint set fans over several hundred kilo-meters around the Appalachian orocline.

The pattern of dip joints in the AppalachianPlateau detachment sheet developed through fourstages, as indicated by both the abutting of parentjoints and the rotation angle of fringe cracks. Theregional pattern of parent joints is due to the over-lapping pattern of joint sets with different strikes.The pattern developed from west to east as pro-gressively younger joint sets propagated within astress field rotating in a clockwise manner. Dur-ing this progressive development of structures inthe detachment sheet, the Alleghanian stress fieldsweeps through an angle of roughly 40° from320° to north-south. To the southeast in the an-thracite region of the Valley and Ridge, structuresrecord a clockwise rotation of the Alleghanianstress field through an angle of 30° (336° to 006°)(Gray and Mitra, 1993). Although the magnitudeof the stress field differs slightly from the Plateauto the Valley and Ridge, the correlation is closeenough to suggest that both areas recorded thesame tectonic history, which involved a clock-wise rotation of the horizontal stress field.

Counterclockwise Propagation of FringeCracks near the Eastern Boundary. Fringecracks, coupled with the abutting geometries ofparent joints, provide a record of the path fol-lowed by the Alleghanian stress field during thestructural development of the AppalachianPlateau detachment sheet. Not only do thesestructures indicate the orientation of the stressfield when they formed, they also preserve arecord of the temporal sense of rotation of thestress field. The fidelity of the record of stressfield rotation from the fringe cracks and abuttinggeometries can be tested by comparing the rota-tion of the Alleghanian stress field and strainwithin the detachment sheet. An isostrain map forlayer-parallel shortening of the detachment sheetwas constructed from data on the orientation ofboth deformed fossils and cleavage (Geiser,1988). Layer-parallel shortening fans from westto east in a radial pattern (Fig. 17). The east-westlines display a uniform spacing (i.e., strain gradi-ent) except east of Ithaca in the eastern portion ofour study area, where the rectangular grid is dis-torted and the east-west grid lines are moreclosely spaced. This region of the most distortedgrid coincides with the region containing parentjoints decorated with fringe cracks reflecting acounterclockwise rotation on the Alleghanianstress field (Fig. 11,A–D). It is this distribution offringe cracks relative to strain that provides thebasis for inferring a tectonic history of the de-tachment sheet.

In the eastern portion of our study area evi-dence is strong for the counterclockwise sense ofstress field rotation after the formation of an ini-tial set of dip joints striking between 006° and021°. Such tectonics are indicated by both fringe

cracks and abutting geometries. Relative to itsorientation farther west (i.e., 342°), the 20° to30° misalignment of the early stress field to theeast is significant. We note that this misalign-ment occurs in the area of the pinchout of Silu-rian salt, which served as a convenient detach-ment surface. Our interpretation is that themisaligned stress field east of Ithaca is associ-ated with rock drag, produced by the lack of saltdetachment at the eastern edge of the area. Dragat the edge of the detachment sheet set up localremote stress field controlling the orientation ofearly parent joints striking between 006° and021° (see Geiser, 1988). A later stress field pro-ducing both the 352° and 003° sets reversed thesense of shear on the parent joints along the east-ern boundary of the detachment sheet such thatlater fringe cracks are oriented counterclockwisefrom parent joints.

TECTONIC HISTORY OF THE APPALACHIAN PLATEAU DETACHMENT SHEET

A final reconstruction of the tectonic history ofthe Appalachian plateau detachment sheet mustbe consistent with the following observations.

1. Fringe cracks on dip joints reflect a clock-wise stress field rotation in the west, a counter-clockwise stress field rotation in the east, and inan area near Ithaca showing both senses of rota-tion (Fig. 11, A–D).

2. The rotation angle for counterclockwisefringe cracks is generally larger than the angle forclockwise fringe cracks.

3. When developed in the same outcrops, theparent joints in the shale are oriented clockwiserelative to parent joints in the siltstone layers.

4. The relative ages of parent joints indicatedby the abutting relationships show a clockwise ro-tation of the stress orientation from west to east,except along the eastern portion of the study area.

In the following section, we summarize thetectonic events associated with the propagationof parent joints and, subsequently, fringe crackswithin the study area, which is the northeasternportion of the larger Appalachian plateau detach-ment sheet (Figs. 6 and 18).

Structural Sequence in the Northeastern Portion of the Detachment Sheet

320°–330° Parent Joint Set. The earliestparent joints in our study area are limited to thewestern region, the area that is closest to the ex-tension of the Juniata culmination into theplateau (Fig. 17). This set includes joints rang-ing over about 10° of strike (320° to 330°). Thereason for this range of strike is not clear in light


YOUNES AND ENGELDER

of the better-aligned later joint sets. This samejoint set is found west of the Juniata culminationand in the detachment sheet as far south as WestVirginia (Fig. 18). However, areas west andsouthwest of the Juniata culmination are charac-terized by a counterclockwise sequence of jointdevelopment. Zhao and Jacobi (1997) reportedthat, west of our study area, joints striking from322° to 340° predate joints at 312°–320° and asecond group of joints at 280°–305°. Evans(1994) reported that, west of the Juniata culmi-nation, the earliest joint set strikes 350° and isfollowed by the 320° to 330° joints. In centralNew York, the earliest structures include folding(042° fold axes) of the Tully Limestone (Younesand Engelder, 1995). Because the shortening di-rection indicated by the deformation of fossils isat a significant clockwise angle with the earliestjoints (Engelder and Geiser, 1980), these earli-est joints apparently predate significant layer-parallel shortening. The earliest joints in theValley and Ridge also predate appreciable layer-parallel shortening (Nickelsen, 1979). Earlyjoints occur throughout the stratigraphic sectionnear the Juniata culmination, yet farther eastthey are found only below the Tully Limestone.The orthogonal relationship between the foldaxes of the Tully Limestone and the early jointssuggests the synchronous timing of the twostructures. At this stage detachment within thesalt decollement may have initiated but was notwell developed. Farther southeast, early foldaxes of the Lackawanna syncline are parallel tothose found in the Tully Limestone.

342° Parent Joint Set.The propagation of342° parent set marks the initiation of significantdetachment along the Salina salt, as indicated bydeformed fossils. As the stress orientation rotatedclockwise, early stages of layer-parallel shorten-ing were recorded as cleavage in the Tully Lime-stone. As the amount of layer-parallel shorteningincreased, the eastern edge of the detachmentsheet began to drag where the eastern salt pinch-out prevented easy detachment. As a conse-quence of this drag, a shear couple developedwithin the eastern region of the detachment sheetso that the local stress field was rotated consider-ably clockwise relative to that found in the rest ofthe detachment sheet. Joints in the eastern regionpropagate in orientations ranging from 006° to021°. During this tectonic stage, folding contin-ued in the section above the Tully Limestone. Aclockwise stress field rotation, east of the Juniataculmination, is reflected in the initial develop-ment of clockwise fringe cracks in the westernregion of the study area.

351° Parent Joint Set. A third stage is markedby the propagation of 351° joints. By this stage,parent joints had propagated throughout most of

the northern detachment sheet and drag along theeast edge of the detachment sheet was at a maxi-mum. This was the first stage during which coun-terclockwise fringe cracks formed along the east-ern edge of the detachment sheet. Kinks equivalentto the 351° parent joints are also found throughoutthe western portions of the sheet.

003° Parent Joint Set. The continuing clock-wise rotation of the Alleghanian stress field isnext indicated by 003° joints. It is possible thatthe kinks showing a counterclockwise angle haddeveloped during this stage as a result of north-south compression. The general north-south ori-entation of the kinks (Fig. 11D) and their con-stant angles to the parent joint for both clockwiseand counterclockwise sets (Fig. 12) indicate thatthey are related to the same event. This interpre-tation is also supported by the mechanical re-quirement that kinks will only form after arrest oftheir parent joint.

070° Parent Joint Set.This stage of jointing,best developed in black shales (Loewy, 1995), ismost difficult to date. The 070° joints are a man-ifestation of tectonic relaxation before abnormalpressure within shales of the Catskill delta couldleak off (McConaughy and Engelder, 1999). The070° set carries twist hackles that are always ro-tated counterclockwise (Fig. 11E).

000° Parent Joint Set. The propagation of asecond north-south joint set is post-Alleghanian.These late joints formed parallel to the orienta-tion of the kimberlite dikes cutting the detach-ment sheet. The kimberlite dikes were dated asEarly Jurassic and consequently, it is possiblethat this set is also early Mesozoic in age. Thislater 000° set may be a manifestation of de-formation associated with continuing slip onfaults of the Clarendon-Linden zone (R. Jacobi,1998, personal commun.). The 003° and 000°sets are also distinguishable through differencesin length and planarity.

Tectonics on Either Side of the Juniata Culmination

The Juniata culmination correlates with a re-gional lineament known in central Pennsylvaniafrom surface geology and magnetic and gravityanomalies (Gold and Parizek, 1976). The jointpropagation sequence around the Juniata culmi-nation indicates that paleostress trajectories ro-tated away from parallelism with the culminationin opposite directions, leading to a clockwise se-quence of structures to the east and north of it, anda counterclockwise sequence of structures to thewest and south of it (Figs. 18 and 19) (Younes,1996). The trend of the Juniata culmination is par-allel to the oldest joints in the detachment sheet(320°–330°). Nickelsen and Hough (1967)

mapped five sets of dip joints in the Pennsylvaniaportion of the detachment sheet (sets A–E). Set A,oriented 330°, is parallel to the Juniata culmina-tion and perpendicular to the Lackawanna syn-cline, and is flanked by joint sets D and E to thenortheast, and joint sets B and C to the southwest.Based on the similarities between the joint orien-tations in central New York (this study) and thosein northeastern Pennsylvania, and also on the sim-ilarities between those in southeastern Pennsylva-nia to those in West Virginia (Dean et al., 1984,1988), we interpret the joint sets of Nickelsen andHough (1967) as a manifestation of clockwisestress field rotation in the northeast (set D) andcounterclockwise rotation in the southwest (sets Band C). This interpretation is consistent withEvans (1994) analysis west of the culmination.Zhao and Jacobi (1997) reported a clockwise su-perposition of joints to the east of the Juniata cul-mination and a counterclockwise superposition ofjoints west of the feature.

Exposures of clockwise and counterclockwisefold rotations in the Antes Shale of the SinkingValley fault zone, Pennsylvania, show a trans-port direction that is compatible with transportparallel to the Juniata culmination (Nickelsenand Engelder, 1989). The direction of transportwas estimated as 322°, similar to the oldesttrends in our study area. Deformation at the Al-legheny front near Williamsport, Pennsylvania,shows a single transport direction at 340°(Spiker and Gray, 1997). Nickelsen (1988) notedthat the Jacks Mountain fault, farther south, hasbeen intersected by a series of small wrenchfaults from which he was able to determine thecompression direction. These directions fall intoorientation about 322°, then 335°, and then 312°,which indicates both senses of stress field rota-tion that might be expected for locations near theculmination. On a larger scale, the Juniata cul-mination divides both the Appalachian Plateauand Valley and Ridge into an area of a clockwisepaleostress rotation to the east and an area ofcounterclockwise paleostress rotation to the west(Fig. 18).

CONCLUSIONS

The character of fringe crack development isdependent on lithology. Gradual twist hackles arefound in clastic rocks regardless of grain size,whereas abrupt twist hackles are most commonin shale with their parent joints found in coarsersiltstone and sandstone beds. Gradual twist hack-les in sandstone develop with larger twist anglesthan those found in shale. In contrast, the abrupttwist hackles in shale display the largest twist an-gles. Theory indicates that this is behavior largelycontrolled by the difference in stress conditions




in the shale and coarser grained clastic rocksrather than by rock properties. The geometry ofthe gradual twist hackles is consistent with lowstress differences (< 6 MPa) in the clastic sectionwhere the joints were driven by a relatively high(~81 MPa) internal fluid pressure.

Fringe cracks are, indeed, systematic across ageological province to the extent that they are use-ful as a tool for unraveling the tectonic history ofthe region. In the northeastern portion of theAppalachian Plateau detachment sheet, fringecracks reflect the clockwise sense of Alleghanianstress field rotation seen to the southeast in the an-thracite district of the Pennsylvania Valley andRidge. Both fringe cracks and abutting of parentjoints suggest that the clockwise rotation of theAlleghanian stress field is more akin to a continu-ous series of tectonic events rather than two punc-tuated phases, as originally proposed by Geiserand Engelder (1983). However, the eastern edgeof the detachment sheet was dragged at the saltpinchout to contribute to a locally perturbed re-mote stress field, as reflected in the counterclock-wise sense of fringe cracks in this area.

Joints coaxial with the Juniata culmination arethe earliest joint sets in the Appalachian Plateaudetachment sheet and in the Valley and Ridgeboth east and west of the Culmination. With laterjoint development, the Juniata culmination di-vides the Alleghanian stress rotation into a clock-wise domain to the east and a counterclockwisedomain to the west.

ACKNOWLEDGMENTS

This work was supported by funds from PennState’s Seal Evaluation Consortium and the NewYork State Geological Survey. The Fault Dynam-ics Research Group at the Royal Holloway Uni-versity of London provided facilities during the fi-nal preparation of the manuscript. We thank HansAvé Lallemant, Michelle Cooke, Mary BethGray, Byron Kulander, David McConaughy, JonOlson, Dave Pollard, and Kelly Rust for criticalreviews of early versions of this paper and for pro-viding helpful comments about the concepts pre-sented within the paper.

REFERENCES CITED

Bahat, D., 1991, Tectonofractography: Amsterdam, ElsevierPress, 250 p.

Bankwitz, P., and Bankwitz, E., 1984, Die symmetrie von kluft-flache und eine systematik ihrer strukturen: Geologie, v. 15,p. 896–941.

Beinkafner, K. J., 1983, Terminal expression of decollement inChautauqua County, New York: Northeastern Geology,v. 5, p. 160–171.

Bradley, W. H. and Pepper, J. F., 1938, Structure and gas possi-bilities of the Oriskany Sandstone in Steuben, Yates, andparts of the adjacent counties: New York Geological Sur-vey Bulletin 899-A, 68 p.

Broek, D., 1991, Elementary Engineering Fracture Mechanics:Dordrecht, Kluwer Academic Publisher, 516 p.

Cooke, M. L., and Pollard, D. D., 1996, Fracture propagationpaths under mixed mode loading within rectangularblocks of polymethyl methacrylate: Journal of Geophysi-cal Research, v. 101, p. 3387–3400.

Cottrell, B., and Rice, J. R., 1980, Slightly curved or kinkedcracks: International Journal of Fracture, v. 16, p. 155–169.

Cruikshank, K. M., Zhao, G., and Johnson,A. M., 1991,Analy-sis of minor fractures associated with joints and faultedjoints: Journal of Structural Geology, v. 13, p. 865–886.

Dean S., Baranoski, M., Bertoli, L., Kribbs, G., Stephens, T.,Kulander, B., Sochman, D., and Mumpower, D., 1984,Regional fracture analysis in western Valley and Ridgeand adjoining plateau, West Virginia and Maryland:American Association of Petroleum Geologists Bulletin,v. 67, p. 448.

Dean, S. L., Kulander, B. R., and Skinner, J. M., 1988, Structuralevolution of the Alleghanian orogeny in southeastern WestVirginia: Geological Society of America Bulletin, v. 100,p. 299–310.

Engelder, T., 1979, Mechanisms for strain within the UpperDevonian clastic sequence of the Appalachian Plateau,western New York: American Journal of Science, v. 279,p. 527–542.

Engelder, T., 1982, A natural example of simultaneous opera-tion of free-face dissolution and pressure solution:Geochimica et Cosmochimica Acta, v. 46, p. 69–74.

Engelder, T., 1985, Loading paths to joint propagation during atectonic cycle:An example from the Appalachian Plateau,U.S.A.: Journal of Structural Geology, v. 7, p. 459–476.

Engelder, T., and Engelder, R., 1977, Fossil distortion anddecollement tectonics of the Appalachian Plateau: Geol-ogy, v. 5, p. 457–460.

Engelder, T., and Fischer, M. P., and Gross, M., 1993, Geolog-ical aspects of fracture mechanics: Geological Society ofAmerica Short Course Notes: Boston Meeting, Boulder,Colorado, Geological Society of America, 281 p.

Engelder, T., and Geiser, P., 1979, The relationship betweenpencil cleavage and lateral shortening within the Devo-nian section of the Appalachian Plateau, New York: Geol-ogy, v. 7, p. 460–464.

Engelder, T., and Geiser, P. A., 1980, On the use of regionaljoint sets as trajectories of paleostress fields during the de-velopment of the Appalachian Plateau, New York: Journalof Geophysical Research, v. 85, p. 6319–6341.

Engelder, T., and Lacazette, A., 1990, Natural hydraulic frac-turing, in Barton, N., and Stephansson, O., eds., Rockjoints: A. A. Balkema, Rotterdam, p. 35–44.

Erdogan, F., and Sih, G. C., 1963, On the crack extension inplates under plane loading and transverse shear: Journalof Basic Engineering, v. 85, p. 519–527.

Ettensohn, F. R., 1985, The Catskill Delta complex and theAcadian orogeny: A model,in Woodrow, D. L., andSevon, W. D., eds., The Catskill Delta: Geological Soci-ety of America Special Paper 201, p. 39–49.

Evans, K., Engelder, T., and Plumb, R. A., 1989, Appalachianstress study 1:A detailed description of in situ stress vari-ations in Devonian shales of the Appalachian Plateau:Journal of Geophysical Research, v. 94, p. 1729–1754.

Evans, M. A., 1994, Joints and decollement zones in MiddleDevonian shales: Evidence for multiple deformationevents in the central Appalachian Plateau: Geological So-ciety of America Bulletin, v. 106, p. 447–460.

Evans, M. A., 1995, Fluid inclusions in veins from the MiddleDevonian Shales:A record of deformation conditions andfluid evolution in the Appalachian Plateau: GeologicalSociety of America Bulletin, v. 107, p. 327–339.

Fischer, M. P., Gross, M. R., and Engelder, T., 1991,A re-evalu-ation of the origin of fracture twist hackles: Geological So-ciety of America Abstracts with Programs, v. 23, p. 469.

Frey, M. G., 1973, Influence of Salina Salt on structure in theNew York–Pennsylvania part of the Appalachian Plateau:American Association of Petroleum Geologists Bulletin,v. 57, p. 1027–1037.

Geiser, P. A., 1988, The role of kinematics in the constructionand analysis of geological cross sections in deformed ter-ranes,in Mitra, G., and Wojtal, S., eds., Geometries andmechanisms of thrusting: Geological Society of AmericaSpecial Paper 222, p. 47–76.

Geiser, P., and Engelder, T., 1983, The distribution of layer par-

allel shortening fabrics in the Appalachian foreland ofNew York and Pennsylvania: Evidence for two noncoax-ial phases of the Alleghanian orogeny,in Hatcher, R. D.,Jr., et al., eds., Contributions to the tectonics and geo-physics of mountain chains: Geological Society of Amer-ica Memoir 158, p. 161–175.

Gold, D. P., and Parizek, R. R., 1976, Field guide to lineamentsand fractures in central Pennsylvania: International Con-ference on New Basement Tectonics, 2nd: Newark, Uni-versity of Delaware, 27 p.

Gray, M., and Mitra, G., 1993, Migration of deformation frontsduring progressive deformation: Evidence from detailedstructural studies in the Pennsylvania anthracite region,USA: Journal of Structural Geology, v. 15, p. 435–450.

Gross, M. R, Fischer, M. P., Engelder, T., and Greenfield, R. J.,1995, Factors controlling joint spacing in interbeddedsedimentary rocks: Integrating numerical models withfield observations from the Monterey Formation, USA,inAmeen, M. S., ed., Fractography: Fracture topography asa tool in fracture mechanics and stress analysis: Geologi-cal Society [London] Special Publication 92, p. 215–233.

He, M-Y., and Hutchinson, J. W., 1989, Kinking of a crack outof an interface: American Society of Mechanical Engi-neers Transactions, v. 56, p. 270–278.

Helgeson, D., and Aydin, A., 1991, Characteristics of jointpropagation across layer interfaces in sedimentary rocks:Journal of Structural Geology, v. 13, p. 897–911.

Hodgson, R. A., 1961a, Classification of structures on joint sur-faces: American Journal of Science, v. 259, p. 493–502.

Hodgson, R. A., 1961b, Regional study of jointing in CombRidge-Navajo mountain area, Arizona and Utah: Ameri-can Association of Petroleum Geologists Bulletin, v. 45,p. 1–38.

Hudak, P., 1992, Terminal decollement in the AppalachianPlateau of northwestern PA: Northeastern Geology, v. 14,p. 108–112.

Kay, S. M., Snedden, W. T., Foster, B. P., and Kay, R. W., 1983,Upper mantle and crustal fragments in the Ithaca kimber-lites: Journal of Geology, v. 91, p. 277–290.

Kulander, B. R., Barton, C. C., and Dean, S. L., 1979, The ap-plication of fractography to core and outcrop fracture in-vestigations: U.S. Department of Energy Morgantown En-ergy Technology Center Report, METC/SP-79/3, 174 pp.

Kulander, B. R., and Dean, S. L., 1985, Hackle plume geome-try and joint propagation dynamics,in Stephansson, O.,ed., Fundamentals of rock joints: Sweden, Centek Pub-lishers, p. 85–94.

Kulander, B. R., and Dean, S. L., 1995, Observations on frac-tography with laboratory experiments for geologists, inAmeen, M. S., ed., Fractography: Fracture topography asa tool in fracture mechanics and stress analysis: Geologi-cal Society [London] Special Publication 92, p. 59–82.

Kulander, B. R., Dean, S. L., and Ward, B. J., 1990, Fracturedcore analysis: Interpretation, logging, and use of naturaland induced fractures in core: American Association ofPetroleum Geologists, Methods in Exploration Seriesno. 8, 88 p.

Lacazette,A., and Engelder, T., 1992, Fluid-driven cyclic prop-agation of a joint in the Ithaca siltstone, AppalachianBasin, NY,in Evans, B., and Wong, T-F., eds., Fault me-chanics and transport properties of rocks: London, Acad-emic Press Ltd., p. 297–324.

Lawn, B., 1993, Fracture of Brittle Solids: Cambridge, Cam-bridge University Press, 378 p.

Loewy, S., 1995, The post-Alleghanian tectonic history of theAppalachian Basin based on joint patterns in Devonianblack shales [Master’s thesis]: University Park, Pennsyl-vania, Pennsylvania State University, 179 p.

Martens, J. H., 1924, Igneous rocks of Ithaca, New York andvicinity: Geological Society of America Bulletin, v. 35,p. 305–320.

McConaughy, D. T., and Engelder, T., 1999, Joint interactionwith embedded concretions: Joint loading configurationsinferred from propagation paths: Journal of StructuralGeology (in press).

Muller, O. H., and Pollard, D. D., 1977, The stress state nearSpanish Peaks, Colorado, determined from a dike pattern:Pure and Applied Geophysics, v. 115, p. 69–86.

Murphy, P. J., 1981, Detachment structures in south central NY:Northeastern Geology, v. 3, p. 105–116.

Nemat-Nasser, S., and Oranratnachai, A., 1979, Minimum

YOUNES AND ENGELDER

spacing of thermally induced cracks in brittle solids:Journal of Energy Resources Technology (Trans.ASME), v. 101, p. 34–40.

Nickelsen, R. P., 1979, Sequence of structural stages of the Al-legheny orogeny at the Bear Valley strip mine, Shamokin,Pennsylvania: American Journal of Science, v. 279,p. 225–271.

Nickelsen, R. P., 1988, Structural evolution of folded thrustsand duplexes on a first-order anticlinorium in the Valleyand Ridge Province of Pennsylvania,in Mitra, G., andWojtal, S., eds., Geometries and mechanics of thrustingwith special reference to the Appalachians: GeologicalSociety of America Special Paper 222, p. 89–106.

Nickelsen, R. P., 1996, Alleghanian sequential deformation onthe SW limb of the Pennsylvania Salient in Fulton andFranklin Counties, south-central Pennsylvania: Pennsyl-vania Geologic Survey, Annual Field Conference ofPennsylvania Geologists, 61st, Guidebook, p. 1–9.

Nickelsen, R. P., and Engelder, T., 1989, Fold-thrust geometriesof the Juniata culmination, Central Appalachians of Penn-sylvania,in Engelder, T. E., ed., Structures of the Appala-chian foreland fold-thrust belt: International GeologicalCongress, 28th, Field Trip Guidebook T166.

Nickelsen, R. P., and Hough, V. D., 1967, Jointing in the Appa-lachian Plateau of Pennsylvania: Geological Society ofAmerica Bulletin, v. 78, p. 609–630.

Olson, J. E., and Pollard, D. D., 1989, Inferring paleostressfrom natural fracture patterns: A new method: Geology,v. 17, p. 345–348.

Parker, J. M., 1942, Regional systematic jointing in slightly de-

formed sedimentary rocks: Geological Society of Amer-ica Bulletin, v. 53, p. 381–408.

Plumb, R. A., Evans, K. F., and Engelder, T., 1991, Geophysicallog responses and their correlation with the bed to bedstress contrasts in Paleozoic rocks of the AppalachianPlateau, New York: Journal of Geophysical Research, v. 96,p. 14509–14528.

Pollard, D. D., Segall, P., and Delaney, P., 1982, Formation andinterpretation of dilatant echelon cracks: Geological So-ciety of America Bulletin, v. 93, p. 1291–1303.

Purslow, D., 1986, Matrix fractography of fiber-reinforcedepoxy composites: Composites, v. 17, p. 289–303.

Rodgers, J., 1970, The tectonics of the Appalachians: NewYork, Wiley Interscience, 271 p.

Scott, P. A., Engelder, T., and Mecholsky, J. J., 1992, The cor-relation between fracture toughness anisotropy and sur-face morphology of the siltstones in the Ithaca Formation,Appalachian Basin,in Evans, B., and Wong, T-F., eds.,Fault mechanics and transport properties of rocks: Lon-don, Academic Press Ltd., p. 341–370.

Segall, P., and Pollard, D., 1983, Nucleation and growth ofstrike slip faults in granite: Journal of Geophysical Re-search, v. 88, p. 555–568.

Sheldon, P., 1912, Some observations and experiments on jointplanes: Journal of Geology, v. 20, p. 53–70.

Spiker, E. C., and Gray, M., 1997, A study of progressive de-formation at the Allegheny Front, Lycoming County,Pennsylvania: Geological Society of America Abstractswith Programs, v. 29, no. 1.

Van Tyne, A., 1983, Natural gas potential of the Devonian

black shales of New York: Northeastern Geology, v. 5,p. 209–216.

Ver Steeg, K., 1942, Jointing in the coal beds of Ohio: Eco-nomic Geology, v. 37, p. 503–509.

Warpinski, N. R., 1989, Determining the minimum in situstress from hydraulic fracturing through perforations: In-ternational Journal of Rock Mechanics and Mining Sci-ence, v. 26, p. 523–532.

Wedel, A. A., 1932, Geologic structure of the Devonian strataof south-central New York: New York State Museum Bul-letin 294, 74 p.

Woodworth, J. E., 1896, On the fracture system of joints, withremarks on certain great fractures: Boston Society of Nat-ural History Proceedings, v. 27, p. 63–183.

Younes, A. I., 1996, Tectonic implications from joints and foldhistory in the Appalachian Plateau, central New York: Ge-ological Society of America Abstracts with Programs,v. 28, no. 3, p. A-111.

Younes,A. I., and Engelder, T., 1995, Paleostress rotation as in-dicated by joints within the Appalachian Plateau: Clock-wise, counterclockwise, or both?: Geological Society ofAmerica Abstracts with Programs, v. 27, no. 6, p. A-383.

Zhao, M., and Jacobi, R. D., 1997, Formation of regional cross-fold joints in the northern Appalachian Plateau: Journal ofStructural Geology, v. 19, p. 817–834.

MANUSCRIPTRECEIVED BY THESOCIETY MAY 2, 1997REVISEDMANUSCRIPTRECEIVEDFEBRUARY 2, 1998MANUSCRIPTACCEPTEDAPRIL 6, 1998



Printed in U.S.A.

fringe cracks: key structures for the interpretation of the …jte2/references/link124.pdf ·...

Documents