IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.343343T.204.2016

Methods

Distinguishing tectonic and induced deformation

Tectonic deformation in drill core includes the structures present in the volume of rock prior to drilling. Here, we use a set of criteria to identify tectonic deformation based on kinematic and geometric evidence following methods practiced on board during Expedition 343. Characteristics of tectonic structures in the split core face include the following:

  • Shear displacement evidenced by measurable offset of piercing points such as marker beds or other predeformational structures, truncation of marker beds or other tectonic structures where displacement may exceed core width, development of slickenlines on fracture surface planes, or significant changes in bedding orientation across a structure, or juxtaposition of different rock types/chemistries/fossil assemblages.

  • Development of fault rock formed by deformation within a brittle shear zone (e.g., scaly fabric defined by the surfaces of phacoid lenses that create an anastomosing network of curviplanar, striated shear surfaces) or as an alternative to scaly fabric, a nonfoliated fault gouge.

  • Presence of tectonic breccia (which may occur without independent evidence for shear displacement) indicated by striated and/or polished surfaces on subangular to angular fragments or by gradational boundaries defined by less fractured rock.

  • Formation of anastomosing or braided networks of structures dark gray to black in color, often with higher X-ray computed tomography (X-CT) numbers (indicating increase in density) than the surrounding rock, with orientations typically oblique to bedding (i.e., phyllosilicate bands [Fossen et al., 2007]).

  • Fracture fill or mineralization evidenced by change in color or mineralogy within structure compared to surrounding rock.

  • Consistent orientation(s) of adjacent structures that define structure sets likely forming under the same stress field.

Induced deformation in core includes structures that form as a result of drilling, coring, or handling operations. Kulander et al. (1990) provide a detailed description of common induced structures observed in drill core, which fall into three categories: drilling-induced structures that propagate in the rock ahead of the drill bit, coring-induced structures that develop anywhere within the core barrel, and handling-induced structures that form during or after removal of the core from the core barrel. Drilling- and coring-induced structures are initiated by stresses related to drilling operations and/or removal of overburden pressure. Handling-induced structures are initiated by stresses related to splitting, bending, and impact after cores are brought on board (Kulander et al., 1990).

Whereas the groupings defined by Kulander et al. (1990) provide a simple overview of induced structure types, complexity arises when core recovery results in open fractures forming along preexisting in situ tectonic structures or sedimentary layering. Therefore, an open fracture in recovered core may be a primary tectonic structure, an induced feature, or an induced fracture superimposed on a primary structure and/or preexisting mechanical anisotropy. Examples of preexisting mechanical anisotropies include natural fractures, faults, microfractures, bedding planes, residual stresses, and solution cleavage (Kulander et al., 1990).

We defined 10 styles of induced deformation from a combination of our observations of the split faces of Hole C0019E cores and previously published descriptions of induced deformation. The definitions of each and the criteria employed to identify them are described below.

Drilling biscuits/discs

Drilling biscuits or discs (“biscuits”) are one of the most commonly recognized examples of induced damage in core. Following observations made by Pendexter and Rohn (1954), Kidd (1978) first cataloged induced damage observations with a focus on “core-discing” in the context of deep-sea drilling after encountering examples in Deep Sea Drilling Project (DSDP) Leg 42A Mediterranean sediment cores. Biscuits appear as separated blocks with convex tops and/or concave bottoms (Fig. F2A), typically separated by drilling mud or a collection of rock fragments in a matrix of drilling mud. Drilling mud can also smear along the side of the core and out from between biscuits, resulting in discrete, isolated biscuits of lithified material (Leggett, 1982). Inclinations of biscuit edges are typically horizontal to subhorizontal in vertical or subvertical cores; however, inclined bedding can influence the orientation (Kulander et al., 1990). Completely developed horizontal coring-induced biscuits cut entirely across the core or abut against earlier-formed natural or coring-induced structures (Kulander et al., 1979). Additionally, circular striations can sometimes be observed on the tops and/or bottoms of biscuits.

Biscuiting is the result of vertical tensile stress acting on the rock during progress of the drill bit followed by blocks spinning with respect to each other in the core barrel (Kidd, 1978; Kulander et al., 1979). Rapid removal of overlying strata permits underlying rocks to expand, contributing to vertical tension and the formation of subhorizontal fractures due to unloading. It common for some biscuits to be held steady in the core barrel while other adjacent biscuits rotate with the drill. Curved edges form as the biscuits spin while staying in contact with each other (e.g., Kidd, 1978). The wear and abrasion results in erosion of biscuit edges and local formation of wear products that may resemble fault gouge or drilling mud.

Induced brecciation

The term “induced breccia” is commonly used in visual core description because of the tendency for drilling operations to recover cores that contain only fragmented angular clasts. For a clear example, see Cores 2R and 3R. Throughout the majority of cores from Hole C0019E, however, induced breccia clasts appear within a single biscuit, as if each fragment is a uniquely fitting piece of a puzzle, or in between biscuits in random orientations. More simply, others have described drilling breccias as angular chips of indurated mudstone in drilling mud (Leggett, 1982). One reliable way to differentiate between natural tectonic brecciation and drilling-induced brecciation is the presence of striations on individual clasts. Induced breccias will sometimes lack slickenlines or polished surfaces because of limited relative motion between clasts. An exception is spiral or helicoidal slickenlines resulting from the biscuiting process (see the “Methods” chapter [Expedition 343/343T Scientists, 2013b]). Most induced brecciation observed in Hole C0019E cores is not in a matrix of drilling mud, such as the induced breccia zone in Figure F2B, where none of the angular clasts exhibit slickenlines or polished surfaces between pieces.

Triangular fracture sets

Triangular fracture sets are common at the edge of the core barrel in the split face of the core. Triangular fracture sets are typically defined by subhorizontal fractures dipping in opposite directions that intersect at the apex of a triangle that points toward the center of the core (Fig. F2C). These structures are interpreted as induced because their orientations locally parallel biscuit edges and/or other induced structure orientations and the triangular geometry is consistent with localized failure during vertical flexing of the core (e.g., Hiraishi, 1984). The triangles can occur at nearly all scales visible within the core. The angle at the apex between the two fractures in a triangular geometry ranges from ~30° (Fig. F2C) to over 100° (Fig. F2D). A complicating factor in determining the origin of fractures that make triangles at the edge of the drill core is the case of reactivation of tectonic fractures. In some cases, fractures with triangular geometries trend directly into tectonic structures, most commonly phyllosilicate bands (Fig. F2E). Triangular fracture sets can sometimes form in between two biscuits and because flexure of the core in the vertical direction ultimately erode away (Fig. F2C).

(Near) right-angle fractures

Continuous, open fractures exhibiting nearly 90° bends are prevalent near the outer edge of Hole C0019E drill cores. The most common occurrence of such fractures is along biscuit corners. Right-angle fractures are observed both opening along preexisting tectonic structures and with no adjacent or parallel tectonic structures (Fig. F2F). In the absence of any evidence of shear or abutting relationships between the differently oriented segments of a right-angle fracture, we infer that they are the result of continued rotation and flexure of the core, resulting in removal of corners from biscuits. As described in Kulander et al. (1990), induced fractures can form when pieces of intact core are chipped off along the edges because of the plucking action of the clockwise-rotating drill bit.

Radiating fractures

Open fractures displaying a variety of orientations that radiate from, or seem to originate at, common points are observed in Hole C0019E cores and are inferred to be induced. This geometry is apparent in the split face and is consistent with formation in response to a load applied at an arbitrary angle and position where mismatched pieces of rock are in contact with intact core. Oftentimes the impacting fragment of rock is preserved, displaying a set of fractures that all radiate from the point of a triangular-shaped fragment (e.g., Fig. F2G). Similar radial arrays of fractures have been noted on previous IODP expeditions as being drilling induced. Arthur et al. (1980) and Dengo (1982) describe the fractures as originating near the core center with a radial symmetry in the downcore direction. Kulander et al. (1990) describe similar curviplanar fractures that originate outside of the core and diverge downcore. The point of origin for some radial fractures can therefore also be absent within the recovered core. However, an observation of the preserved impacting fragment is also evidence of mobile piece(s) of rock during the coring process. Radial fractures that originate off-core are an indication of inception off-core and growth ahead of the advancing bit in rock that is subsequently drilled.

Core axis–parallel fractures

Several forms of induced deformation occur symmetrically about the long axis of the core. The most obvious of these is the presence of a zone of flowage (Dengo, 1982; Lundberg and Moore, 1986), rock fragments in a matrix of dominantly drilling mud between the core and the core liner. Additionally, two types of vertical induced fractures are recognized that vary in scale and orientation. The first are petal-centerline fractures, described in detail by Kulander et al. (1979, 1990) and Li and Schmitt (1998). Characteristic petal-centerline fractures dip between 30° and 75° in a downcore direction near the core edge and curve to dip vertically near the center of the core (Kulander et al., 1990). Petal-centerline fractures form in response to an induced principal tension that rotates downward in a vertical plane from an inclined orientation to horizontal (Kulander et al., 1979), and the strike of the fracture surface is aligned with the direction of the greatest horizontal stress (Li and Schmitt, 1998). We observe no clear petal-centerline fractures in the JFAST cores, similar to other studies in thrust faulting regimes (Li and Schmitt, 1998). However, in drill cores of well-indurated rock, petal-centerline fractures are commonly used to estimate in situ principal stress orientations (e.g., Li and Schmitt, 1998; Davatzes and Hickman, 2010).

The second type of core axis–parallel induced fractures is typically smaller in scale and interpreted as related to both the coring and handling processes. Induced core axis–parallel fractures, sometimes several centimeters in length and open several millimeters, are best observed in Section 20R-2 (Fig. F2H). This section, composed of dominantly stratified pelagic clay with subhorizontal bedding, displays subvertical fractures that appear to have originated near the center of the core and never reached the edge of the core. We interpret these induced fractures as forming under nonuniform deformation of layers under triaxial or uniaxial stress conditions, similar to “axial cracking” described in Dusseault and Van Domselaar (1984) and Kulander et al. (1990), where compression of the core vertically results in the formation of vertical to subvertical Mode I fractures that abut either a subhorizontal biscuit boundary or detachment surface within the core.

Small saw-mark–parallel fractures

Open fractures that are roughly 1 cm or less in length on the core face and aligned parallel to the ridges and grooves produced by the core-splitting saw are common. The fractures have nearly perfectly horizontal rakes along the cut face (Fig. F2I) and do not extend to the edge of the core. Most importantly, small saw-mark–parallel fractures cannot be observed in X-CT imagery collected before core splitting. Small horizontal fractures commonly terminate against other open fractures. Especially where they terminate against reactivated induced structures, the terminating fractures are likely younger in age and not related to tectonic deformation. The orientation of the fractures parallel to the saw blade and relative timing of formation of the fractures shows that they are handling-induced.

Drilling mud injection

Drilling mud may be difficult to distinguish from lithified (or partially lithified) clay and mudstone in cores. Drilling mud is typically lighter in color than the mudstone in Hole C0019E cores and is much softer. However, drilling mud can appear to be very similar to lighter rocks. In X-CT imagery, drilling mud is notably lighter than surrounding rocks because of the low density of the mud. Drilling mud correlates with induced deformation in a number of contexts. It can appear to have been injected through paths between drilling biscuits and is observed to penetrate along open fractures into the interior of the core. Small fragments of rock from cores can be entrained within the drilling mud matrix, giving the material the appearance of a breccia (Fig. F2J); however, such fragments cannot be considered tectonic in origin with any confidence without the presence of slickenlines. In these occurrences, a combination of biscuiting and drilling mud injection results in a drilling breccia.

Fractures crosscutting drilling mud

In rare instances, open fractures can be observed crosscutting drilling mud (Fig. F2K). The crosscutting and relative timing relationship is a clear indication that the fracture postdates the injection of drilling mud and therefore is induced by the drilling or coring processes.

Rubble

Commonly at the tops and bottoms of core sections is a total loss of coherence in the recovered rocks. The common location in the sections is not a coincidence because cores are often sectioned in places where this style of induced deformation occurs. Consequently, it is common for induced damage to be concentrated at the tops and bottoms of core sections and remain fairly evenly distributed throughout the rest of the core. Subrounded fragments with a variety of lithologies and randomly oriented bedding appear as “rubble,” a zone of fragments that are not in situ and likely each have entirely different origins (Fig. F2L). Rubble zones display no associated drilling mud at the top of core sections. Fragments in rubble zones rarely retain a cut face parallel to the orientation of core splitting.

Measurement of structure density

For the purposes of quantifying the density of structures in the examined cores, we categorized each structure based on a confidence index developed as part of this work. The first structural style (Category 1) is deformation that is unequivocally tectonic and in situ—the strongest possible confidence level regarding tectonic origin (see criteria outlined in “Distinguishing tectonic and induced deformation”). The second structural style (Category 2) is deformation that may be tectonic or induced in origin but lacks definitive evidence for either source. The third structural style (Category 3) is deformation that can clearly be demonstrated to have been induced by drilling, coring, and/or handling operations.

The density of core-scale structures was quantified using a linear scanline sampling method performed on core imagery and verified through direct observation of the archive halves. Structure density is defined here as a linear measurement of the number of structures per unit length (meters) (Rohrbaugh et al., 2002). High-resolution imagery of all cores was collected on board during Expedition 343 using a multisensor core logger (MSCL)-I digital imaging system with GEOSCAN IV (Geotek, Ltd.). The instrument uses a line-scan system for split-core imaging, reducing the optical distortion from the lens or variations of lighting and provides 16 bit RGB color TIFF-formatted images with a resolution of 100 pixels/cm. The linear scanline method allows for quick measurement of structure characteristics and is the main method used in borehole image logs and cores (Zeeb et al., 2013). The linear scanline method has been used in many previous studies of damage zone structures at the outcrop scale and has clearly documented elevated damage in the vicinity of many major faults (e.g., Caine and Forster, 1999; Wilson et al., 2003; Mitchell and Faulkner, 2009; Savage and Brodsky, 2011). In order to best avoid induced damage near the edge of the core, the linear scanline was drawn down the center of each core. The depth of every structure intersecting the scanline was recorded, and each structure was assigned a category according to the classification scheme described above.

In intensively deformed intervals of cores (i.e., breccia zones) where the spacing between structures was ~1 cm or less, depths of structures within the zone were approximated by estimating the average size of clasts within the zone. Because clasts within breccias are fragmented rock, the boundaries of clasts are fracture surfaces and the size of the clast is a measure of fracture spacing. We used an image analysis method implemented in Matlab following the methods of Bjørk et al. (2009) to measure clast properties. Because of the irregular sizes and shapes of clasts in breccia zones, the simple linear scanline sampling method could not completely represent the spacing of structures. We therefore calculated the average clast size in the breccia zone and set the structure density to the breccia interval thickness divided by the average clast diameter.