The crystallographic preferred orientations (textures) of three samples of
Black Hills Quartzite (BHQ) deformed experimentally in the dislocation creep
regimes 1, 2 and 3
Quartz textures (crystallographic preferred orientation), usually presented
in the form of pole figures, are frequently used for the analysis of deformed
rocks. Optical, X-ray or electron backscatter diffraction (EBSD) data are widely considered to make
interpretations about deformation kinematics such as shear senses
In experimentally deformed quartzites, regimes of different mechanical
behavior and distinct microstructures have been recognized as three distinct
dislocation creep regimes
In this contribution we will focus on the following questions: which factors
influence the texture geometry (shape of pole figure skeletons)? Is the
texture controlled by deformation temperature, geometry and kinematics, or
recrystallization processes? How reliably can certain texture components be
used to infer the activity of a specific slip system? To this end, EBSD data
obtained from Black Hills Quartzite
Experimental conditions and sample strain.
Shear stresses (
Definitions and reference frames.
The samples analyzed are experimentally deformed BHQ
of
Explanation of the mean grain kernel average misorientation (gKAM):
crops of EBSD maps of 0.25 and 1
EBSD maps were collected on a Zeiss Merlin field emission gun scanning electron microscope,
equipped with an Oxford EBSD camera in low vacuum mode, using 2
Crystal directions
Contoured pole figures and inverse pole figures (IPFs) are calculated from the
orientation distribution functions (ODFs). The ODF was calculated either from
all measurements (area weighted) or from the grain modal orientations (one
orientation per grain). In the case of ODF calculations, a de la Vallée-Poussin
kernel was used. The kernel width was either fixed or estimated using the
Kullback–Leibler cross-validation implementation in MTEX. Estimated kernel
half widths are between 7 and 14
Specific types of textures and certain texture components have been given
genetic or descriptive terms in the literature. In particular, grains with a
specific
Orientation maps for samples w1092, w946 and w935.
The grain kernel average misorientation (gKAM) is calculated from the kernel
average misorientation (KAM) from noise-reduced EBSD data (Fig.
In regime 1, the pole figures show a broad, asymmetric peripheral
distribution of
On inverse pole figures of sample strain and other reference directions
(Fig.
Orientation maps are color-coded for crystal directions parallel to specific
reference directions (Fig.
Figure
Pole figures are calculated for grains grouped with respect to aspect ratio (
For all regimes,
For
The position of the peripheral part of the kinked
Pole figures showing average grain size (
Pole figures of average grain properties are obtained by calculating
spherical interpolations for grain size, aspect ratio and axial ratio (the
inverse of the aspect ratio), based on the
For regimes 1 and 2 the average grain size is largest at the periphery and for
regime 3 it is in the center of the
Pole figures are calculated for all orientations of grains with a gKAM below
and above the median gKAM and below the 20 and above the 80 % quantile
values (Fig.
Assuming that texture strength, i.e., texture index, is an expression of the
contribution of texture-forming processes to deformation, a quantitative
comparison of texture strength as a function of gKAM,
In all regimes, the texture index increases with increasing aspect ratio and
for higher aspect ratios also for increasing grain size. Also, the classes
containing the largest (recrystallized) grains and highest aspect ratio
possess the highest texture strength (Fig.
The volume of the B fiber is largest for regime 1 and smallest in regime 3,
while the volume of the Y fiber is smallest in regime 1 and largest in
regime 2, reflecting what can be seen in the pole figures (Fig.
Dependencies of
Quantitative comparison of texture index and volumes of texture
components. Separate color maps for texture index, B-fiber and Y-fiber
volume as a function of aspect ratio
Misorientation axes and texture domains.
Misorientation axes and texture domains. Data are identical to Fig.
Misorientation axes have been determined in specimen and crystal coordinates
for misorientation angles of 2–9
In all samples and for all types of grains, the axis distribution in specimen
coordinates shows a maximum parallel to the structural
In specimen coordinates, the density distributions of misorientation axes of
some B, R and
The average, generalized Schmid factor shown for different slip systems
(or combinations of slip systems) as a function of the trend of
The generalized Schmid factor
The mean generalized Schmid factor is plotted as a function of the trend of
the maximum principal stress direction of the stress tensor
(Fig.
The
Textures of the high-strain samples of all regimes share the alignment of
Pole figure skeletons similar to the ones shown here have also been reported
form naturally deformed rocks
In the high-strain experiments analyzed here, the temperature range was
considerably small (850–915
On the sample scale, a dependence of the texture transition on strain is
evident from the comparison of the regime 3 low- and high-strain samples.
However, the dependence of the texture transition
on strain can also be verified along strain gradients
with the high-strain samples of the three regimes. The
The more highly deformed the grains in each regime, the further away from the
periphery of the pole figure the
However, while the strain-dependent texture development can be used to
explain the differences between the low- and high-strain regime 3 samples as
well as variations inside each sample, it cannot account for the complete
texture transition between regime 1 and regime 3 experiments, all deformed roughly to the
same strain. The high volume of B domains in regime 1 but also in the low-strain regime 3 sample must have developed from the uniformly textured BHQ.
Because all strain-related effects on the texture seem to drive
In order to better understand the processes involved during deformation of
the samples, the deformation and recrystallization processes involved are
discussed. The strength of the texture in a deformed rock is a function of
the contribution of different processes and the strain these processes
accommodated. Therefore, the texture strengths produced by dislocation glide
as a texture-forming process and grain boundary sliding as a texture-weakening process have to be compared quantitatively. The concurrent
operation of dislocation creep and grain boundary sliding is well known from
metals
The texture strength increases from regime 1 to regime 3 as well as with
increasing grain-scale deformation (Fig.
In regime 1, bulging recrystallization (BLG) is thought to be the predominant
recrystallization mechanism
In regime 2 and regime 3 (high-strain samples), surviving large grains
usually show systematic substructures characterized by discrete orientation
domains of a size comparable to small recrystallized grains. In addition,
orientation domains that are approximately the size of
original BHQ grains
From these observations, the contribution of dislocation creep is interpreted to be highest in regime 3 and the contribution of grain boundary sliding is highest in regime 1. Since all mechanisms operate concurrently, observed changes in texture strength are a result of a different contribution of each individual process (regime 1 to regime 3) and bulk strain (regime 3 low strain to high strain). Regime 2 is interpreted as transitional – more dislocation creep than in regime 1 and more grain boundary sliding than in regime 3.
In the regime 1 sample analyzed, a large fraction of recrystallized grains
attain an orientation unrelated to the original orientation of the BHQ
grains. While grain boundary sliding may contribute to the dispersion of
orientations of newly recrystallized grains away from the host orientation,
it is not a process that initially develops grains with
Small, equiaxed grains are systematically found at two conjugate, peripheral
Quartz
The misorientation axes of the B domain show a distribution close to uniform.
In the case of
The interpretation above corroborates with the Schmid factor analysis
(Fig.
Conceptual model of the development of a crystallographic preferred orientation during general shear.
As outlined above, a deformation-dependent rotation of
The principal difference in that suite of experiments is the sample strength,
mostly controlled by the amount of water added to the sample assembly, a
technique well known to modify the strength of experimentally deformed
quartzites
We suggest a model in which the texture transition can be explained (a) by
the motion of
In regime 3, SGR is the most active and fastest recrystallization process,
and glide-induced rotation of
We do not have direct evidence of a fracture, nucleation and/or growth origin of the
newly formed grains in regime 1; however, in most experimental studies
B domains seem to be the first to form at high-stress conditions. Our model
also explains the texture transition observed by
Textures with peripheral
The analysis presented yields several implications with respect to the
applications of quartz textures for the analyses of deformed rocks, in
particular with respect to the interpretation of deformation temperatures. In
nature, microstructural transitions
in which different recrystallization processes developed over a metamorphic
gradient have been documented
In addition to the formation of new grains by nucleation and growth (dominant
in regime 1) and by SGR (dominant in regime 3), contributions of grain
boundary sliding (possible in all regimes) may challenge our understanding of
the grain size–stress relation with respect to piezometric applications.
Together with the grain size–gKAM relation documented in the companion
paper
To study the textures (crystallographic preferred orientations) of Black Hills Quartzite, deformed in general shear in regimes 1, 2 and 3, EBSD data were analyzed using a number of new methods for combined texture and microstructure analysis.
The textures of all regimes are dominated by the alignment of
Recrystallization in regime 3 happens by subgrain rotation recrystallization, and deformation occurs
through dislocation creep with a minor contribution from grain boundary sliding. Porphyroclasts in all regimes
deform by dislocation glide and climb. The texture varies as a function of grain size, grain lengthening,
long-axis alignment and gKAM. Differences in experimental temperatures and strain rates are negligible, while the water-moderated
stress level varies between 100 and The finite texture balances the contributions of dislocation glide on several The hypothesis that That the peripheral
The MTEX function to calculate the gKAM can be found here:
Grain properties were interpolated in
The estimated texture strength of a given set of
orientations depends on a proper kernel width and since kernel width
estimators should always be conservative, the texture strength, when properly
calculated, will not be overestimated. Conversely, this makes it
difficult to compare textures based on very different numbers of
orientations. In those cases, an estimation based on the same number of
orientations with a fixed kernel width is more suitable for a quantitative
comparison. In order to compare texture strength of the grain property
classes (e.g., grain size–aspect ratio classes), subsets of orientations
with identical sizes were estimated in a bootstrapping approach. Within a
property class, 100
randomly chosen subsets of the size of the smallest of any population
(
Renée Heilbronner is a member of the editorial board of the journal.
This article is part of the special issue “Analysis of deformation microstructures and mechanisms on all scales”. It is a result of the EGU General Assembly 2016, Vienna, Austria, 17–22 April 2016.
We wish to thank Jan Tullis for letting us use her thin sections again. We are also grateful to Tom Eilertsen, Kai Neufeld and Michel Bestman for support and advice during EBSD image acquisition at Tromsø University. The comprehensive reviews by Luiz Morales and Dave Prior are greatly appreciated; they helped to improve the paper significantly. Support from the National Science Foundation of Switzerland, grant no. NF 200021-138216, is gratefully acknowledged. Edited by: Florian Fusseis Reviewed by: David Prior and Luiz Morales