Polyphase deformation structures are widespread in the metamorphic cores of orogens, where the rocks involved may include:
Structures found in these zones include
In contrast to folds in slate belts (where there is usually a clear distinction between competent layers with class 1 geometry and incompetent layers with class 3 geometry) folds in higher-grade metamorphic rocks often show geometries close to class 2 - similar folds.
Example
How do these folds originate?
One obvious mechanism is heterogeneous simple shear
Except in nearly liquid rocks (e.g. migmatites, slump folds in wet sediment) it is difficult to understand how sufficiently variable shear stresses could operate on closely spaced planes to do this. Also, these folds commonly have axial plane fabrics similar (though coarser grained) to those in slate belts, which develop perpendicular to shortening direction, and not parallel to shear planes.
Close examination typically reveals that in fact there are layers that are 'just' class 1C folds, alternating with layers that are 'just' class 3. This suggests a much more likely kinematic scenario. Initally these folds probably result from buckling, similar to that observed in slate belts. Then a pure strain (or other homogenous high strain) is superimposed on the rocks at high temperature, under conditions where competence contrast is lower.
Animation, Examples reviewed
Areas affected by multiple episodes of folding may display refolded folds in a bewildering variety of orientations.
Example
To make sense of these we use a classification (devised, like many other fundamental classifications of folds by J. Ramsay). This classification relates to two generations of folds only - if we have 3 or more generations things get more complicated.
In each case we characterize a first generation of folds by
We examine the effect of the second generation of folds on those two structures.
If later folding does not bend either the axial surfaces or the hinges of early folds, then the early folds are effectively just tightened during the second phase of deformation. The fold pattern is described as 'type 0', and generally it will not be apparent from geometry that there have been two phases of deformation.
If later shortening is roughly parallel to the hinges of the early folds, and the extension direction also lies in the early fold axial surfaces, then later folding bends the hinges of the early folds but does not significantly bend the axial surfaces. This typically produces culminations and depressions on the early fold hinges, and leads to oval outcrop patterns of layer traces on outcrop surfaces. This is called a type 1 fold interference pattern, characterized by domes and basins.
If on the other hand, the extension direction in the second deformation is at a high angle to the early axial surfaces, then those axial surfaces may be folded too. Under these circumstances outcrop patterns are characterized by complex shapes that resemble mushrooms and bananas. This is called a type 2 interference pattern.
There is a third case, which occurs when the shortening direction is at a high angle to the early fold hinges and the extension direction is at a high angle to the early axial surfaces. Under these conditions the early hinges may remain more or less straight but the early axial surfaces are folded. The trace of layering typically follows a multiple zigzag shape described as a type 3 interference pattern.
How to analyse an overprinted fold pattern. The most important first step is to draw axial traces on an image of the outcrop. In general, the latest set of axial traces is likely to be approximately straight. Earlier axial traces may or may not be folded. (In general, type 1 interference patterns, where early axial surfaces are not folded, are the most difficult to analyse; it may not be possible to determine the order of folding.)
Facing directions may also be useful in the analysis of overprinted folds. For example, downward-facing folds (synformal anticlines and antiformal synclines) always indicate that there are at least two generations of folds.
Overprinted fold patterns are difficult to find except where outcrop is continuous. We have to be coincidentally on the axial trace of two generations of folds. It's much easier to unravel the structural history of polyphase deformation by looking for:
Based on these types of relationships we typically number the generations of folds and fabrics in a given area using a numerical scheme. Fold generations in this scheme are numbered F1 F2 F3 etc. The corresponding foliations are S1, S2, S3 etc, where S2 is the foliation that is axial planar to F2 etc.
Corresponding lineations are numbered L1, L2 etc.
(Note that there may be more than one lineation associated with a generation of structures. For interest, if you are working with complex structures, here are some additional notes on handling multiple lineations that are part of a single generation of structures)
Metamorphism is often concurrent with deformation and new mineral growth can often be related to polyphase deformation.
Porphyroblasts of garnet, staurolite etc may be much more rigid than their matrix. Hence we can recognize
Where the porphyroblasts are poikiloblasts, there may be a record of fabric development, if the inclusions are aligned in trails
Porphyroblasts that are contemporary with deformation may show complex intrusion trails
Note: there have been big controversies over whether helicitic structure has been produced by:
Metamorphic petrologists use metamorphic reactions to produce P-T-t paths. By integrating information on fabric development with the timing of growth of metamorphic minerals and isotopic dating, it's sometimes possible to develop quite detailed knowledge of D-P-T-t paths (Deformation - Pressure - Temperature -time).
These paths are described as 'clockwise' or 'counterclockwise' based on their appearance on a graph of P against T. Unfortunately, there are two ways of doing this: with pressure plotted increasing upward, or with pressure increasing downward (as it does in the real Earth). Most researchers in tectonics prefer the upward style, so a path that increases pressure early, and then increases temperature, is typically described as 'clockwise'. (Note that the text by Van der Pluijm and Marshak uses the opposite convention; see the footnote on p. 333.)
Examples of D-P-T-t histories
Blueschists and other high pressure - low temperature metamorphic rocks almost always show clockwise paths in PT space. This is because rapid burial is necessary to produce high P; if cold rocks remain at depth for any time, they start to warm up. To be preserved without completely overprinting the high P low T features, they have to be exhumed rapidly, but the metamorphic history nonetheless shows some of this warming.
In the Alps, for example, within the metamorphic Pennine Nappes, there is an early phase of high pressure metamorphism which produced blue sodic amphiboles, but these are overprinted with green actinolite or hornblende rims formed later in the deformation history.
A series of illustrations in the text by Van der Pluijm and Marshak shows how P-T-D-t paths have been used to construct a model for the development of these nappes. Many of the large recumbent folded structures seen in cross-sections of the Alps seemed to have formed beneath an over-riding plate, part of a continental block that started to collide with mainland Europe around 65 Ma.
Illustrations
These cross-sections don't give the whole picture, because late stages in collision have been accompanied by strike-slip movements along several major faults.
The Himalayas are also a product of large-scale continental collision, between India and Eurasia. This time the northern plate over-rode the southern.
This collision has thickened the crust beneath the Tibetan plateau to approximately double the Himalayan thickness.
There are two end-member models for this thickening. In one model it is accounted for by thickening of the Eurasian margin (an approximately pure-shear model). At the other extreme is a model in which the Indian crust extends beneath the Eurasian crust as far as the north edge of the plateaue. Think of this as a simple shear model.
Deformation in the Himalayas is not a simple matter of convergence. Studies on kinematic indicators in metamorphic rocks and granitoids show that a sheet of high grade metamorphic rocks in the High Himalayan crystalline zone is bounded above by a normal fault, the south Tibetan detachment zone. Deformation and exhumation of these rocks occurred while granites were being intruded.
This leads to the idea that these rocks were weakened by partial melting, causing them to lose the wedge-shape (critical taper) that characterize thrust belts and spread laterally. This type of behaviour has been called channel flow.
As in the Alps, later stages in deformation have been marked by major strike-slip faults, suggesting that blocks of continental crust are escaping laterally from the zone of intense collision. In addition there are north-south graben cutting the Tibetan plateau that may also reflect lateral extrusion of crust.
Normal faults are also prominent in the later history of the North American Cordillera. A large area of the western US shows basin and range geomorphology defined by numerous normal faults. VDPM 16.12
Diagrams showing multiple intersecting normal faults
It's clear that some faults formed in flat orientation, and outline large areas of deep-crustal metamorphic rocks from which the cover has been removed by a combination of tectonism and erosion - core complexes.
VDPM 16.13 cross-section of core complex
The faults surrounding core complexes are typically hybrid brittle-ductile structures, with fault breccia in the hanging wall and mylonite in the footwall.
These features suggest that these extensional zones are dominated by listric faults that flatten into a detachment close to the brittle-ductile transition.
In other parts of the orogen, including the southern Canadian Cordillera, there are core complexes that appear to be older equivalents of the basin and range.
Why core complexes? Several hypotheses have been suggested:
All these factors have affected the Cordillera -the abundance of extensional structures late in orogen history may be due to a combination of factors.