Elsevier Publishing Company, Amsterdam - Printed in The Netherlands
EXPERIMENTAL STRUCTURAL GEOLOGY J. B. CURR1E
Department of Geology, University of Toronto, Toronto, Ont. (Canada)
Experimental work in structural geology comprises principally high-pressure deformation of rock samples and construction of dynamic scale models. During the first half of this century laboratory studies of rock deformation have simulated a wide range of geological conditions in respect of pressure, temperature and strain rate. These studies have increased our understanding of mechanisms by which rock deformation proceeds. Scale models achieve their greatest value when used to illustrate structural processes. Their results aid the appreciation of theoretically derived structural relationships and serve also to relate the stages of structural development that are observed in separate field occurrences.
To understand geological structures one needs to examine the processes by which particular structural types develop. As in other branches of geology, the study of a feature's formative history may cast light upon its genesis. Some geological processes are rapid enough to be immediately observable; others are too slow to be directly measured but, since they occur at the surface, their effects can be interpreted in a sequence of events. Most structural processes share neither of these advantages. A folded rock, now exposed, may have had entirely different properties during deformation. From study of several structures at different stages of growth, a field geologist must infer the course of structural development and the physical condition of rocks at the time they were deformed. To trace such a formative process as a sequence of events, one may construct a model wherein the properties of rocks are simulated by a model material, or one may investigate the manner in which the rocks themselves behave under conditions of elevated pressure and temperature. As a result of this choice, experimental studies in structural geology have developed in two directions (HILLS, 1963). (1) The construction of scale models. (2) High-pressure deformation of rocks.
I (1966) 5 1 - 6 7
EXPERIMENTAL ROCK DEFORMATION
In contrast to model experimentation which first attracted the attention of geologists, laboratory investigation of rocks and rock-forming minerals early drew the interest of physicists, crystallographers and engineers. BRIDCMAN(194%, 1952) studied both non-metallic and metallic materials at pressures up to 100,000 bars. BVERGER(1930) recorded results of early studies on translation gliding in a variety of mineral types. Metallurgical studies have provided concepts that are serving as guides to work on non-metallic minerals and aggregates (GRKiGSet al., 1960a, b). The importance of rock strength and manner of failure to design of underground openings has led to investigation of elastic properties and strength of rocks under short-term loads. Its importance to seismology has prompted extension of this work to high temperature and pressure (HUGHES and MAURETTE, 1956, 1957). Initial interest by geologists in experimental studies sprang from field observation of structures that suggested rock flowage. At the beginning of this century argument existed as to whether such flow occurs through intense granulation of solid rock, whether rocks can become plastic and alter shape while maintaining coherence, or whether shape changes are accomplished by a continuous process of solution and redeposition of minerals. ADAMS and NICOLSON (1901, p.365) advanced the opinion that "it is a matter of great difficulty and in fact in most cases it is quite impossible to decide with certainty upon the relative merits of these conflicting views from a study of the deformed rocks themselves". Adams and Nicolson recognized the problem of delineating structural processes by examination of static structures and, like HEIM (1878), they reasoned that carefully conducted experiments on the deformation of rocks might prove valuable. Since the initial experiments of ADAr~S and NICOLSON(1901), VON KARM~,N (1911). and of ADAMSand BANCROFT(1917) exploratory research has investigated a wide range of geologically interesting conditions. Confining pressures have commonly extended from atmospheric conditions upward to 10 kbars--equivalent to depths of about 40 km; temperature variation has been carried to 800°C--a value that is probably consistent with temperatures attained at such depths (BIRCH, 1955). Strain rates have ranged from values produced by impact loading down to 10 ~ and 1 ~ per year--still some six orders of magnitude above those in geological processes. This work, though varied in character, has been conducted by a relatively small group of workers. Only in the last two decades has the number of laboratories devoted to high-pressure rock deformation expanded. Few studies have been directed toward intensive examination of particular rock types, with the notable exception of the investigation of Yule Marble undertaken by Griggs and his colleagues. Throughout the years of work that followed Adams' investigations, experimental apparatus has been enlarged and refined in order to extend the range of conditions and the accuracy of measurements. However, the objective of rock deEarth-Sci. Rev..
I (1966) 51-67
EXPERIMENTAL STRUCTURAL GEOLOGY
formation has remained unchanged, as evidenced by a recent statement that "laboratory determination of deformational properties of rocks is of foremost importance if we are to ever understand the mechanics of natural rock deformation in the earth" (HEARD, 1963, p.162).
Apparatus Adams' apparatus constitutes one of the first attempts to achieve a confining pressure that would simulate deformation of a small unit of rock at depth. The rock sample, about 2.5 cm in diameter and 3.8 cm long, was fitted snuggly within a thin-walled wrought-iron jacket which had been heated to expand its diameter slightly. When cooled, this jacket firmly supported the sample at every point while the latter was loaded axially by steel pistons. Resistance of the jacket to deformation represented a corresponding resistance by rock adjacent to a small unit deformed at depth in the crust. Later workers replaced the wrought-iron jacket by fluid which surrounds the solid specimen and permits a continuous record to be made of confining pressure. The apparatus of GRIGGS (1936)follows this method and maintains a constant level of confining pressure by an upward movement of the pressure cylinder containing the rock sample, and an accompanying withdrawal of a lower free piston and advance of an upper piston that loads the sample axially. The specimen is encased in a thin jacket which serves only to separate it from the confining fluid. Care must be exercised in selecting this fluid since at high pressure its viscosity may become so great that hydrostatic pressure is no longer transmitted adequately to the sample or throughout the equipment (GRIGGS, 1954). Several workers have employed this type of apparatus, with modifications and improvement, to perform compression tests by increasing axial pressure (era) over confining pressure (a2=a3), and for extension tests by decreasing the axial pressure (az) while maintaining or increasing the confining pressure (al--a2). To conduct shearing tests, BRIDGMAN(1952) devised equipment in which an anvil is rotated horizontally between vertically moving platens which have a small cylindrical boss adjacent to the anvil. Discoid samples, placed between the anvil and bosses, undergo uniform shear while subjected to high normal pressure along their axis. This basic apparatus has been adapted to a variety of experimental requirements and has greatly expanded the flexibility of work at very high pressure (GRIGGS and KENNEDY, 1956; DACHILLE and Roe, 1962). Desire to vary, not only the type of loads but also the rate of loading, has led to experiments at moderate and low rates of strain approaching those of geological processes. Commonly a constant stress or creep apparatus is employed for this purpose. In such equipment the sample is placed under a load of constant magnitude, well below its rupture strength. Recently, in a different type of design, HEARD(1963) developed an effective mechanism for slow uniform advance of the axial-load piston. Earth-Sci. Rev., 1 (1966) 51-67
By adding a heating element, either as a furnace around the high-pressure bomb or as a unit within it, temperature can be raised to levels that are significant to geological interpretation. Many experiments have been carried out in the range of 300-500 °C and work has been conducted at 800 °C (GRIGGSet al., 1960b). Other variations of experimental conditions have also been employed. One important addition is the control of fluid pressure within pores of the rock sample. With other equipment the effects of simple shear can be studied by punching out the central part of a discoid sample (ROBERTSON, 1955; JAEGER, 1962). In further tests, indentation hardness has been correlated with rock behaviour under geological stresses (BRACE, 1960). These varied experiments are subordinate in number to tests of rock strength in compression or tension. Rock samples
Published results of rock-deformation experiments show that the behaviour of representative samples from each of the major rock types--igneous, metamorphic and sedimentary-- has been examined at elevated pressures and temperatures. Of the more ductile rocks, salt and alabaster have most frequently served as samples (ScHMIDT, 1937; HANDIN and HAGER, 1958). Limestone, marble and dolomite represent examples of less ductile material (GR1GGSand MILLER, 1951; HANDIN and FAIRBAIRN,1955; TURNERet al., 1956; ROBERTSON,1960). Indurated sandstones and quartzite comprise brittle rocks that fail to exhibit ductility, even under the highest experimental pressures. For all rock samples a fine grain-size and uniform texture are essential requirements. The anisotropy of shale, gneiss and schist commonly precludes their use in experimental studies, although some workers have specifically examined this variable (DONATH, 1961; PATERSON and WEISS, 1962; WAESHand BRACE, 1964). In other studies the effect of strength anisotropy has been examined in samples consisting of a brittle core within a hollow cylinder of more ductile rock (GRIGGS and HANDIN, 1960), and in specimens composed of alternating layers of brittle and ductile rock slices (LucHITSKIYet al., 1962). Selection of sample shape has developed largely through practical experience. A right circular cylinder, with length-diameter ratio in the range of 1-3 proves most satisfactory because it can be readily prepared, the propagation of shear fractures within it is not hindered by ends of the specimen and there is little tendency toward buckling under axial loads. Diameter of cylinders commonly employed varies from 1 to 5 cm. For studies in which quantitative values of breaking strength were measured, BRACE(1963) employed long cylinders whose central portion comprised a small throat zone, 2.5 cm long, within which failure was concentrated. Under certain experimental conditions the shape of a sample and the manner in which loads are transmitted to it may assume an importance equal to that of its internal physical properties in determining the pattern of deformation, particularly the geometry of rupture. Earth-Sci. Rev.,
EXPERIMENTAL STRUCTURAL GEOLOGY
General pattern of experimental results Study of individual rock types has outlined the main characteristics of deformation at high pressure and temperature. An initial component of elasticity is present in which strain is independent of time. This is commonly followed by a component of flow or creep for which the strain is time-dependent. Continued deformation may lead to rupture. The creep component can be divided into three stages: (1) a primary or transient creep that decreases with time, (2) secondary or steady creep at constant rate, and (3) a tertiary creep whose increasing rate quickly leads to rupture and consequent loss of cohesion. For brittle locks, rupture surfaces lie closely parallel to the direction of compression and display features of tensile failure. Under conditions of greater ductility, failure surfaces subtend larger angles with this direction and display features characteristic of failure in shear. Confining pressure effectively increases the apparent resistance of rocks to deformation and permits samples to undergo a greater amount of strain prior to rupture. Stress-strain curves of Solnhofen Limestone (Fig.lA) illustrate its ductility (HEARD, 1960, fig.3A). At high confining pressure the occurrence of strain hardening necessitates that the load be continually raised if strain is to be increased. Elevated temperature serves to decrease the differential pressure required in deforming a rock sample. Solnhofen Limestone demonstrates (Fig.1 B) the character of this change (HEARD, 1960, fig.4B). The decrease in strain hardening at high temperatures permits an increase of total strain at low values of differential pressure and an associated increase in secondary creep. Elevated temperature promotes two types of recrystallization; one can be brought about by an annealing process subsequent to deformation; the other type accompanies deformation (syntectonic) and is dependent on the rate of strain as well as temperature (GRIGGS et al., 1960a, b). Increase of ductility at high temperature and pressure has led to a suggestion
~'uE7 "~ /
AT 25°C ~
5h t~ / ! / ~ ~ ~. 41"W ~ [~1 ~
--1,5OOl ATM'I ~1,0001 ATM'I
1'0 1'2 14 STRAIN
4 ,.-~300 °C t 400oC
.~ 31| 750 ATM. [ 3 < 2II CONFINING PRESSURE !F R1At C T~DENOTES URE~-I 21 LL
AT 400 ATM. CONFINING
- - -
~ 500oc ~
10 12 14 S T R A I N (°/°)
Fig.1. Stress-strain curves for Solnhofen Limestone at (A) various confining pressures and (B) various temperatures. (After HEARD, 1960, fig.3A.)
Earth-Sci. Rev., 1 (1966) 51-67
that rocks have no fundamental strength capable of supporting even a limited load for an indefinite period without measurable distortion. For example, HANDIN and HAGER (1958) indicate that rock-salt flows at very low differential pressures without strain hardening; hence, under geological conditions, its deformation may be analogous to that of a viscous fluid. However, rate of strain is a critical factor in rock deformation and has been a cause of reticence on the part of geologists to evaluate experimental results. At successively lower rates of deformation yield strength appears to be systematically lowered (HEARD, 1963). It is not certain, however, that over geological spans of time rocks behave as essentially viscous materials (ROBERTSON, 1963). Another important variable is the physical and chemical effect of fluids within rock pore-space. Hydrostatic pressure in pores reduces the differential pressure required to produce rupture and this increase of pore pressure, relative to confining pressure, promotes a return to brittle failure (HEARD, 1960) and a systematic change in porosity (HANDIN et al., 1963). Chemical changes have not generally received the attention that has been given to physical aspects of rock deformation, although alabaster, limestone and marble have each been the subject of study. Systems involving the diagenesis of sandstone have also shown that moving water, together with high pressure and temperature, accelerates compaction and cementation of quartz grains (MAXWELL, 1964). Somewhat specialized procedures have dealt with variables other than those already mentioned. Cyclical loading and unloading of samples at stresses below rupture strength and impact loading of samples to study rupture characteristics uninfluenced by time-dependent deformation, represent examples of two techniques that have been applied. Such studies have supplied useful information on particular aspects of rock deformation. Generalizations on rock failure Results of experimental studies suggest that mechanisms of rock failure are complex and that factors important to the process will vary in importance from one rock type to another. With respect to a particular rock type, changes in conditions of deformation may alter the dominant mechanism by which failure takes place and thereby alter its apparent strength and manner of failure. In the earliest stages of deformation most rocks exhibit an elastic response to loads. In very ductile rocks this portion may be insignificant for geological interpretation, while in quartzite the elastic stage directly precedes rupture in almost all experiments that have been conducted. Between these extremes lies a wide variety of rocks for which some form of time-dependent flow assumes importance and may dominate the pattern of failure. The primary and secondary stages of flow can be expressed as S ~ s(t) + A(t), where S is total strain, s(t) is the primary transient flow and A(t) describes the Earth-Sci. Rev., 1 (1966) 51-67
steady secondary flow. Transient flow can be represented by a logarithmic relation, viz. s(t) ~ B log t, at moderate temperatures (GRIGGS, 1940; LOMNITZ, 1956; ROBERTSON, 1963) and by Andrades' power function, viz. s(t) ~ B g , at higher temperatures. In secondary flow the rate of strain accumulation (St) is dependent on applied load (or) and may be described by the relation Sr = A sinh B~r (GRIGGS, 1940; HEARD, 1963). Three general mechanisms are evident by which primary and secondary flow are accomplished, namely cataclastic flow, gliding flow and recrystallization flow (HANDIN and HAGER, 1957). Cataclastic flow involves intergranular displacements by rupture of grain boundaries. Gliding flow comprises intragranular movement on twin- or translation-gliding planes within the crystal lattice. Recrystallization flow encompasses the rearrangement of material on a molecular scale through solution and redeposition, by local melting or by solid diffusion. The importance of cataclastic flow has been demonstrated over a wide range of conditions by BRIDGMAN (1949), HANDIN and HAGER (1957), and ROBERTSON (1960). Detailed studies of marble and dolomite have shown clearly the significance and mechanisms of gliding flow in these particular rock types (HANDIN and HAGER, 1955; TURNER et al., 1956). Early studies by GRIGGS (1940) indicated recrystallization flow in alabaster by solution and redeposition, while recent study of flow in marble by HEARD (1963) demonstrates the possible importance of a solid diffusion mechanism. Accumulation of these data has led investigators to develop theoretical concepts that generalize the physical relationships. GORANSON(1940) developed an expression for creep of rock based on thermodynamic potential relations. Subsequent applications of thermodynamic theory to consideration of non-hydrostatically stressed solids have been made by VERHOOGEN (1951), MACDONALD (1960) and KAMB (1961). The tertiary stage of flow has not been included in analytical expressions concerning deformation because it is variable, both in rate and in magnitude and in its relation to rupture. Numerous relationships have been advanced to predict the stress at which rupture may be anticipated and the nature of its geometry. Of these the theories of Mohr and of Griflith have been most successfully applied to a variety of rock types and stress conditions. The theory of Otto Mohr holds that failure occurs along planes of shear determined by the magnitude of normal stress across them (NADAI, 1950). While seemingly adequate at normal temperature and pressure, the Mohr theory of strength is not valid for rocks under all experimental stress conditions, nor is there always close agreement between predicted and observed angles of rupture (14ANDIN and I-lAGER, 1957). The Griffith theory is based on the concept that failure by fracturing is due to concentration of stress at the ends of cracks. The theory predicts stress conditions at which such cracks or flaws will enlarge and produce rupture throughout the body. BRACE(1963) has shown that by viewing grain boundaries as potential flaws and assuming that maximum grain diameter Earth-Sei. Rev., 1 (1966)51-67
constitutes a critical crack length, agreement can be achieved in correlation of measured and predicted rock strength in compression. In this analysis a modification of the Griffith theory developed by McCLINTOCK and WALSH (1962) is used to account for the effect of friction across crack surfaces. Prediction of rupture becomes increasingly difficult when deformation takes place under conditions that promote ductile behaviour. In such instances samples may undergo variable amounts of flow by gliding or recrystallization, prior to loss of cohesion (ROBERTSON, 1955; HANDIN and HAGER, 1957; BRACE, 1963). Theories of rupture employed at present are most successfully applied to deformation of brittle rocks whose failure pattern is uncomplicated by ductile flow.
Value of experimental rock deformation Experimental studies are contributing pertinent information to numerous problems in structural geology. For example, experimentally developed rock fabrics indicate that deformation of grains within a rock can be predicted from their orientation relative to external loads, that deformation will occur by a determinable mechanism and that deformation of the grain aggregate will be homogeneous. As a further example, significant progress has been made in resolving the paradox of apparently brittle rock which at depth and over geological time, behaves as a nearly viscous material. However, a quantitative estimate of the fundamental strength of rocks is not yet certain. In addition to exploratory work, detailed and systematic experimental studies of rock and mineral types are warranted if their mechanics of deformation is to be fully understood and if theoretical generalization is to be achieved. Physical properties which are most important in controlling deformation may then be selected and this information is critical to a second kind of experimentation in structural geology--the construction of scale models.
EXPERIMENTAL SCALE MODELS
Just as the variety of response by rocks to geologic forces suggested that laboratory experiments might elucidate processes underlying rock failure, so also has the observed variety in the form of folds and faults suggested that models might assist a study of processes by which these forms develop. Historically, the application of models to study of structural features preceded the first experimental deformation of rock samples. Review of early model experiments has been provided by publications of WILLIS (1892), PAULKE (l 912), LEITH (1923) and SUMMERS(1932). The diverse objectives of model experiments warrant their division into four general groups. The first group embraces experiments that have attempted to imitate, on a miniature scale, the form of exposed geological structures. The second comprises those in which study has been made of a structural process, sometimes Earth-Sei. Rev., 1 (1966)51-67
EXPERIMENTAL STRUCTURAL GEOLOGY
by testing hypotheses already in the mind of the experimenter. Within the third group fall experiments arising from specific application of dimensional analysis to scale models of structure. The fourth group includes experimental studies which have been performed in conjuction with field observation of structure and theoretical analysis of structural development. While many published studies do not lie exclusively in one group, this division does emphasize their main objectives. Experiments of HALL (1815) illustrate studies within the first group. Hall simulated fold shapes seen in sedimentary strata along the coast of Berwickshire by compressing two margins of a pile of cloth sheets. In later experiments he used stiff clay to simulate folded rock. FAVRE (1878) also chose wet clay to imitate deformed rocks but decreased its strength to that of a thick paste. Daubr6e (in: PAULKE, 1912, pp.50-57) carried out experiments on folding and in some of these he deformed prisms consisting of wax mixed with resin or turpentine to produce various degrees of plasticity. CADELL(1888) became interested in the relative competency of rocks and employed plaster-of-Paris layers interbedded with wet sand to construct models of thrust faulting due to lateral compression. In contrast with Hall's study of folds in a small area, Cadell applied his experimental results to interpretation of structural patterns on a broad scale, specifically the northwestern Highlands of Scotland. WILLIS(1892) furthered this objective in his study of Appalachian folding. In the Appalachian Highlands he recognized four separate districts characterized by open folding, close folding, folding and faulting and folding with associated schistosity. Willis attempted to determine whether these simply represented successive stages in a single structural process. or whether the structure of each area was controlled by the character of its competent and incompetent strata. He studied model effects produced by the interaction during folding of thick and thin beds and of layers having markedly different strength. Experiments by PAULKE(1912) on development of Alpine Nappes represent a continuation of interest in regional structure. Some workers have expanded their investigations to features that extend over considerable segments of the crust (R1MBACH, 1913; CHAMBERLAIN, 1925; HAMILTON, 1962). Throughout all these studies, imitation of the geometrical pattern of structures constituted a primary objective, regardless of the size of features being modelled. Experiments of H. CLOOS (1939) illustrate characteristics of the second experimental group in which detailed examination is made of a structural process. Cloos employed weak clay-water mixtures to investigate fractures, fault development and features of folding. KUENENand DE SITTER(1938) demonstrated the occurrence of shear planes parallel to bedding in the limbs of concentric folds. LINK (1928) examined the growth of en echelon folds. NETTLETON(1934) illustrated the fluid mechanics concept of salt-dome development. BUCHER (1956)outlined the possible importance of body forces in formation of recumbent folds. Each experimental study has sought to illustrate the manner in which particular structural features might develop. Earth-Sci. Rev., 1 (1966)51-67
J . B . CURR1E
The work of Bucher also illustrates the third group of experimental studies in which dimensional similitude between the model and a prototype is considered. Although these concepts were discussed by earlier workers (KOENIGSBERGERand MORATH, 1913; MAmLET and BLONDEL, 1934) or were implied by use of weak materials to simulate rock behaviour, the publication of HUBnERT (1937) on theory of scale models gave them fresh emphasis. DOBRIN (1941) re-examined the fluidmechanics hypothesis of salt-dome growth, using asphalt and corn syrup to simulate salt and sediments respectively. Investigating the interdependence of growth rate and relative viscosities, Dobrin concluded that the magnitude of overburden viscosity is more critical to domal growth than is the viscosity of salt, and that the equivalent viscosity of sedimentary overburden in the Gulf Coastal Plain must be about 1021 poises. PARKERand McDOWELL (1955), reporting on further salt-dome model studies in which they used barite-water muds to simulate overburden, were able to suggest an approximate thickness for the salt bed from which domes grow in the Gulf Coast area. They demonstrated development of a rim syncline, the possible existence of secondary domes and the termination of domal uplift by gradual burial of the salt plug during periods of rapid sediment deposition over it. In a more recent study, RAMBERG(1963a) utilized the large increase of body forces that can be attained in a centrifuge to examine a variety of geological processes which may depend on density differences among rock types. His models not only functioned rapidly, but he was also able to use a wider variety of viscoelastic materials for modelling. Each of these authors indicates that his model materials and resulting structures simulate only the more significant properties and form of a representative structural prototype. There cannot be direct correspondence with a particular geologic structure since size, rock strength and rate of deformation cannot be determined with sufficient accuracy. Interpretation of experimental results must take into account the fact that dimensional similitude between the model system and geologic structures is incomplete. The fourth group of experiments has developed from a combination of structural observation in the field and theoretical analysis of structural processes, with scale-model studies. Investigations vary in the degree to which each type of information is utilized but in all there is an interdependence among the three types of study. In describing the mechanics of normal and reverse faulting, ]~[UBBERT(1951) indicated close agreement between the dip of faults in loose sand and dips recorded for faults observed in the field. The similar behaviour of model and prototype materials could be adequately described by a theoretical distribution of stress in granular media. SANFORD (1959) examined theoretically the elastic response of a homogeneous layer to vertical displacement along its lower boundary. Making his examples analogous to a crustal layer 5 km thick and having the properties of sedimentary rock, he subjected the layer to a distributed vertical displacement of sinusoidal form and to a localized step-wise vertical displacement. Theoretical analysis was borne out by results of model experiments on sand and Earth-Sci. Rev., 1 (1966) 51-67
EXPERIMENTAL STRUCTURAL GEOLOGY
sand-clay mixtures. Satisfactory agreement was evident, both in the distribution of displacements and in the location of initial fracturing. In each case a characteristic fault pattern developed. At the crest of the sinusoidal form a complex zone of normal faults appeared which died out at depth. The step-wise displacement caused a series of curved reverse faults which dipped steeply at depth but intersected the upper surface of the model at low angles and had, adjacent to them, a pattern of normal faults which formed at the surface of the uplifted block. CtJRRIE et al. (1962) described folding of sedimentary rocks under tangential loads as a buckling process which is governed initially by elastic instability. They concluded that fold wavelength is controlled by dominant members within less competent material and used photoelastic models to illustrate the theoretical discussion of folds, to test theoretical assumptions and to indicate geometry and strain distribution that might occur in advanced stages of structural development. Results of theory and experiments were applied to study of fold wavelength in field examples, and to evaluation of dominant member thickness in a sedimentary section. BlOT (1961) considered fold development as a process of buckling in viscoelastic members for which, in addition to static instability, the rate and timehistory of formation must be studied. Hence, viscous properties outweighed the importance of elasticity. He found that deformation grows as a function of time, that fold development is controlled by wavelength and that the dominant wavelength undergoes most rapid growth. BlOT et al. (1961) examined theoretical results by experiments on the buckling of elastic and viscous layers in a viscous medium. RAMBERG(1963b) has also examined the viscous buckling of rocks and has employed models to demonstrate geometrical relations suggested in studies of wavelength, ptygmatic structures and drag folds. Apparatus and model materials Most experimental work has been conducted in apressure box which, in its simplest form, provides active motion from one side (Fig.2A). In other designs pressure has been applied by movement of both ends or by movement of all four sides (AvEBURY, 1903). A pressure box, devised by PAULKE(1912), employed not only a movable side but also a segmented base whose square pieces could be individually raised and lowered to simulate a mobile basement zone under a regionally compressed mass of sediments. BUCHER(1956, pl.VI) used the horizontal motion of a rigid block thrust underneath his model materials to force them upward and out in a recumbent fold (Fig.2B). Other investigators have achieved uniform displacement at the lower margin of their model by stretching or relaxing a rubber sheet which, in turn, rests on either a flat or differentially movable base. While pressure boxes provide a variety of loading conditions, the motion of their geometrically regular sides is not directly analogous to that of geological processes. Hence extraneous boundary effects occur within the model that have to Earth-Sci. Rev,, 1 (1966) 51-67
~ ', \,-'~-J LI
Fig.2. Cross-sections of model experiments A. Thrust faults in layers of plaster-of-Paris and sand above a basal zone of clay mud. (After CADELL,1888.) B. Recumbent fold in layers of stitching wax and grease. (After BUCHER,1956.) be considered when interpreting experimental results. This problem is encountered less often in models of domes due to uplift of low-density material at depth because boundaries of the model and geologic systems approach similarity. An asphalt layer beneath dense syrup (DoBRIN, 1941) is analogous to a bed of salt below a sequence of sedimentary rocks. Thus scale models ofdomal features differ less from their geological prototypes than do other commonly depicted model experiments. Continued attention must be given to development of apparatus which simulates accurately the motion of geological boundaries, and to expansion of experimental methods which can be employed in model studies. RAMBERG(1963a) has illustrated the value of centrifuged models in study of certain types of problems. GzovsKl (1959) and BELLand CURRIE (1964) have applied photoelastic experiments to several structural problems. A limited variety of materials has been used in simulating rock behaviour. Clay-water mixtures, employed even in early experiments, have most frequently served as model material. Only recently, however, has attention been given to quantitative study of properties that render it suitable dimensionally (OERTEL, 1962). Other materials commonly utilized are oil-sand mixtures, beeswax, stitching wax, silicone putty and pariffin wax compounds. In few experiments has detailed study been given to the manner in which the model material simulates different aspects of rock deformation (GzOVSKI, 1959; HAMILTON, 1962). The utility of a model material is conditioned by several practical requirements. It must be readily formed to an initial shape and assembled with other members. The material cannot undergo chemical reaction with media around it, nor change its physical properties during an experiment. Also, the model material is usually sectioned for visual inspection of results at the completion of an experiment. Such requirements restrict the selection of a material as much as do the laws of dimensional analysis. Earth-Sci. Rev., 1 (1966) 51-67
EXPERIMENTAL STRUCTURAL GEOLOGY
Value of model experiments There is wide diversity in the kinds of geologic models that have been constructed and an equally wide diversity exists in the apparent value of experimental results. Some have little bearing on the interpretation of structure if judged solely on the basis of complete dimensional similitude, and have indeed been termed misleading. However, model experiments are truly informative only when objectives, materials and procedures are clearly understood. With this background one can appraise their similarity to the structural processes which their authors attempted to examine. Complete dimensional similarity between model and geologic prototype is not likely to be achieved (DANES, 1964). Models frequently lose their full impact because of incomplete communication. Experiments invariably concern a structural process, yet recorded results often comprise a verbal or graphical description of but one or two stages. Little appreciation can be gained of the sequence of structural events. E. CLOOS(1955, p.256) quite naturally suggested that "every geologist should have some experience in experimentation". Benefit accrues not from a casual inspection of results, but from detailed study and duplication of the experiment itself. The value of model studies to structural interpretation is increased by attention to dimensional similitude. Their utility is enhanced further if they become part of a more general study involving field observation and theoretical analysis. In this context, models serve as a connection between the complexity of actual structures and the quantitative statements supplied by theory.
Only two types of laboratory investigation related to structural geology have been discussed here. Other experimental problems are also of interest. The in situ measurement of strain in mine openings or bore holes, the mechanics of sediment compaction and the relation of shearing stress to mineral stability constitute three diverse examples. However, laboratory deformation of rocks and construction of geological scale models serve as illustrations of problems under investigation and the course of further work. Rock-deformation studies are gradually changing their emphasis from an initial exploration of major factors which control failure toward a systemanc study and assembly of observations on physical and chemical behaviour of rock groups deformed under geological conditions. Scale-model experiments are achieving their greatest impact where used to illustrate structural processes. They are effective because they assist an appreciation of theoretically derived relationships and they graphically relate the structural stages which are observed in separate field occurrences. Earth-S¢i. Rev., 1 (1966) 51 67
J. B. CURRIE
Model experiments and rock deformation share a common need of guidance by a gradually expanding body of theoretical analysis. Theory offers a basis for linking observations on deformation over a wide range of rock types and stress conditions. It will assist the planning of studies that are both time-consuming and costly. To scale-model studies, theory provides a means of estimating the physical significance of experimental results. Of equal importance are the information and direction provided by study of natural structures. Effective laboratory investigations gain impetus from constant recourse to field observation.
REFERENCES ADAMS, F. D. and BANCROFT,J. A., 1917. Internal friction during deformation and relativeplasticity of rocks. J. Geol., 25: 597-637. ADAMS, F. D. and NICOLSON, J. T., 1901, An experimental investigation into the flow of marble. Phil. Trans. Roy. Soc. London, Set. ,4, 195: 363-401. ASHGmEI, G. D., 1963. Strukturgeologie. VEB Deutscher Verlag der Wissenschaften, Berlin, 572 PP. AVEBURY,L., 1903. An experiment in mountain-building. Quart. J. GeoL Soc. London, 59: 348-355. BELL, R. T. and CLrRRIE, J. B., 1964. Photoelastic experiments related to structural geology. Proc. Geol. Assoc. Can., 15: 33-51. BHATTACHARJI,S., 1958. Theoretical and experimental investigations on cross-folding. J. Geol 66: 625-667. BIOT, M. A., 1961. Theory of folding of stratified viscoelastic media and its implications in tectonics and orogenesis. Bull. Geol. Soc. Am., 72: 1595-1620. BIOT, M. A., ODE, H. and ROEVER,W. L., 1961. Experimental verification of the theory of folding of stratified viscoelastic media. Bull. Geol. Soc. Am., 72: 1621-1632. BmcH, F., 1955. Physics of the crust. In: A. POLDERVAART(Editor), Crust of the Earth--Geol. Soc. Am., Spec. Papers, 62:101-118. BRACE, W. F., 1960. Behaviour of rock salt, limestone and anhydrite during indentation. J . Geophys. Res., 65: 1773-1788. BRACE, W. F., 1963. Brittle fracture of rocks. Rand Corporation Mere., RM-3583:103 pp. BRIDGMAN, P. W., 1949a. The Physics of High Pressure. Bell, London, 445 pp. BRIDGMAN, P. W., 1949b. Volume changes in the plastic stages of simple compression. J. Appl. Phys., 20: 1241-1251. BRIDGMAN, P. W., 1952. Studies in Large Plastic Flow and Fracture. McGraw-Hill, New York, N.Y., 362 pp. BUCHER, W. H., 1956. Role of gravity in orogenesis. Bull. Geol. Soc. Am., 67: 1295-1318. BUERGER, M. J., 1930. Translation-gliding in crystals. Am. Mineralogist, 15: 45-64, 174-187, 226-238. CADELL, H. M., 1888. Experimental researches in mountain building. Trans. Roy. Soc. Edinburgh, 35: 337-357. CHAMBERLAIN,R. T., 1925. The wedge theory of diastrophism. J. Geol., 33: 755-792. CLOPS, E., 1955. Experimental analysis of fracture patterns. Bull. Geol. Soc. Am., 66: 241-256. CLOPS, H., 1939. Hebung--Spaltung-Vulkanismus. Geol. Rundschau, 30: 401-519. CURRIE, J. B., PATNODE, H. W. and TRUMP, R. P., 1962. Development of folds in sedimentary strata. Bull. Geol. Soc. Am., 73: 655-674. DACHILLE, F. and Roy, R., 1962. Opposed anvil pressure devices. In: R. H. WENTORF (Editor), Modern Very High Pressure Techniques. Butterworths, London, pp.163-180. DANES, Z. F., 1964. Mathematical formulation of salt dome dynamics. Geophysics, 29: 414-424. DoamN, M. B., 1941. Some quantitative experiments on a fluid salt-dome model and their geological implications. Trans. Am. Geophys. Union, 22: 528-542.
Earth-Sci. Rev., 1 (1966) 51-67
EXPERIMENTAL STRUCTURAL GEOLOGY
DONATH, F. A., 1961. Experimental study of shear failure in anisotropic rocks. Bull. Geol. Soc. Am., 72: 985-989. FAVRE, M. A., 1878. Biblioth6que universelle. Arch. Sci. (Geneva), 246 pp. GOGUEL, J., 1948. Introduction ~t l'6tude m6canique des d6formations de l'~corce terrestre, 2ed. MOm. Carte Gdol. France, 530 pp. GORANSON, R. W., 1940. "Flow" in stressed solids: an interpretation. Bull. Geol. Soc. Am., 51: 1023-1034. GRIGGS, D. T., 1936. Deformation of rocks under high confining pressures. J. Geol., 44: 541-577. GRIGGS, D. T., 1940. Experimental flow of rocks under conditions favouring recrystallization. Bull. Geol. Soc. Am., 51: 1001-1034. GRIGGS, D. T., 1954. High pressure phenomena with applications to geophysics. In: L. N. RIDENOUR (Editor), Modern Physics for the Engineer. McGraw-Hill, New York, N.Y., pp. 272-305. GRIGGS, D. T. and HANDIN, J., 1960. Observations on fracture and a hypothesis of earthquakes. In: D. T. GR~GGS and J. HANDIN (Editors), Rock Deformation--Geol. Soc. Am., Mem., 79: 347-364. GRrGGS, D. T. and KENNEDY,G. C., 1956. A simple apparatus for high pressures and temperatures. Am. J. Sci., 254: 722-735. GRIGGS, D. T. and MILLER, W. B., 1951. Deformation of Yule Marble: I. Bull. Geol. Soc. Am., 62: 853-862. GRIGGS, D. T., PATERSON,M. S., HEARD, H. C., and TURNER, F. J., 1960a. Annealing recrystallization in calcite crystals and aggregates. In: D. T. GR1GGS and J. HANDIN (Editors), Rock Deformation--Geol. Soc. Am., Mem., 79: 21-38. GRIGGS, D. T., TURNER, F. J. and HEARD, H. C., 1960b. Deformation of rocks at 500--800°C. In: D. T. GRIGGS and J. HANDIN (Editors), Rock Deformation--Geol. Soc. Am., Mem., 79: 39-104. GzovsKI, M. V., 1959. The use of scale models in tectonophysics. Intern. Geol. Rev., 1: 31~,7. HALL, J., 1815. On the vertical position and convolutions of certain strata and their relation with granite. Trans. Roy. Soc. Edinburgh, 7: 79-108. HAMILTON, W. S., 1962. Structural model of a large part of the earth. Bull. Am. Assoc. Petrol. Geologists, 46: 610-639. HANDIN, J. and FAIRBAIRN, H.W., 1955. Experimental deformation of Hasmark Dolomite. Bull. Geol. Soc. Am., 66: 1257-1274. HANDIN, J. and HAGERJR., R. V., 1957. Experimental deformation of sedimentary rocks under confining pressure: tests at room temperatures on dry samples. Bull. Am. Assoc. Petrol. Geologists, 41 : 1-50. HANDIN, J. and HAGER JR., R. V., 1958. Experimental deformation of sedimentary rocks under confining pressure: tests at high temperatule. Bull. Am. Assoc. Petrol. Geologists, 42: 2892-2934. HANDIN, J., HAGERJR., R. V., FRIEDMAN,M. and FEATHER,J. N., 1963. Experimental deformation of sedimentary rocks under confining pressure: pore pressure tests. Bull. Am. Assoc. Petrol. Geologists, 47: 717-755. HEARD, H. C., 1960. Transition from brittle fracture to ductile flow in Solnhofen Limestone as a function of temperature, confining pressure and interstitial fluid pressure. In: D. T. GRIGGSand J. HANDIN(Editors), Rock Deformation-- Geol. Soc. Am., Mem., 79: 193-226. HEARD, H. C., 1963. Effect of large changes in strain rate in the experimental deformation of Yule marble. J. Geol. 71: 162-195. HEIM, A., 1878. Untersuchungen iiber den Mechanismus der Gebirgsbildung. Basel, 2: 484. HILLS, E. S., 1963. Elements of Structural Geology. Wiley, New York, N.Y., 483 pp. HOBBS, D. W., 1960. The strength and stress-strain characteristics of Oakdale Coal under triaxial compression. Geol. Mag., 97: 422~135. HUBBERT, M. K., 1937. Theory of scale models as applied to the study of geologic structures. Bull. Geol. Soc. Am., 48: 1459-1520. HUBBERT, M. K., 1951. Mechanical basis for certain familiar geologic structures. Bull. Geol. Soc. Am., 62: 355-372. Earth-Sci. Rev., 1 (1966) 51-67
HUGHES,D. S. and MAURETTE,C., 1956. Variation of elastic wave velocities in granites with pressure and temperature. Geophysics, 21: 277-284. HUGHES, D. S. and MAURETTE,C., 1957. Variation of elastic wave velocities in basic igneous rocks with pressure and temperature. Geophysics, 22: 23-31. JAEGER, J. C., 1962. Punching tests on discs of rock under hydrostatic pressure. J. Geophys. Res., 67: 369-373. KAMB, W. B., 1961. The thermodynamic theory of non-hydrostatically stressed solids. J. Geophys. Res., 66: 259-271. KOENIGSBERGER,G. und MORATH, O,, 1913. Theoretische Grundlagen der Experimentellen Tektonik. Z. Deut. Geol. Ges., 65: 65-86. KUENEN, PH. H., 1958. Experiments in Geology. Trans. Geol. Soc. Glasgow, 23: 1-28. KUENEN, PH. H. and DE SITTER, L. U., 1938. Experimental investigation into the mechanism of folding. Leidse Geol. Mededel., 10: 217-240. LEITH, C. K., 1923. Structural Geology. Holt, New York, N.Y., 390 pp. LINK, T. A., 1928. En echelon folds and arcuate mountains. J. Geol., 36: 526-538. LOMNITZ, C., 1956. Creep measurements in igneous rocks. J. Geol., 64: 473-479. LUCHITSKIY,I. V., BELITSKIY,I. A. and GROMIN,V. I., 1962. Deformation of layered models of rocks. Dokl. Akad. Nauk S.S.S.R., Earth Sci. Sect. 144:114-116. MAcDoNALD, G. J. F., 1960. Orientation of anisotropic minerals in a stress field. In: D. T. GRIGGS and J. HANDIN (Editors), Rock Deformation--Geol. Soc. Am., Mere., 79: 1-8. MAILLET,R. et BLONDEL, F., 1934. Sur la similitude en tectonique. Bull. Soc. G(ol. France, 4: 599~502. MAXWELL, J. C., 1964. Influence of depth, temperature and geologic age on porosity of quartzose sandstone. Bull. Am. Assoc. Petrol. Geologists, 48: 697-709. MCCLINTOCK, F. A. and WALSH,J. B., 1962. Friction on Griffith cracks in rocks under pressure. Proc. Natl. Congr. AppL Mech., 4th, Berkeley, 1962, pp.1015-1021. NADAr, A., 1950. Theory of Flow and Fracture of Solids, 2 ed. McGraw-Hill, New York, N.Y., 572 pp. NETTLETON, L. L., 1934. Fluid mechanics of salt domes. Bull. Am. Assoc. PetroL Geologists, 18: 1175-1204.
OERTEL, G., 1962. Stress, strain and fracture in clay models of geologic deformation. Geotimes, 6 (8): 26-31. PARKER, T. J. and McDOWELL, A. N., 1955. Model studies of salt-dome tectonics. Bull. Am. Assoc. Petrol Geologists, 39: 2384-2470. PATERSON, M. S., 1958. Experimental deformation and faulting in wombeyan marble. Bull. Geol. Soc. Am., 69: 465-476. PATERSON, M. S. and WEISS, L. E., 1962. Experimental folding in rocks. Nature, 195: 1046-1048. PAULKE, W., 1912. Das Experiment in der Geologie. Borntraeger, Berlin, 108 pp. RAMBERG, H., 1963a. Experimental study of gravity tectonics by means of centrifuged models. Bull. Geol. lnst. Univ. Upsala, 42: 1-97. RAMBERG, H., 1963b. Fluid dynamics of viscous buckling applicable to folding of layered rocks. Bull. Am. Assoc. Petrol. Geologists, 47: 484-505. RIMBACH, C., 1913. Versuche fiber Gebirgsbildung. Neues Jahrb. Mineral., Geol. Palfiontol., 35: 689-722. ROBERTSON, E. C., 1955. Experimental study of the strength of rocks. Bull. Geol. Soc. Am., 66: 1275-1314. ROSERTSON, E. C., 1960. Creep of Solnhofen Limestone under moderate hydrostatic pressure. In: D. T. GRIGGS and J. HANDIN (Editors), Rock Deformation--Geol. Soc. Am., Mem., 79: 227-244. ROBERTSON, E. C., 1963. Viscoelasticity. Rand Corporation Mere., RM-3583:57 pp. RobINSON JR., L. H., •959. The effect of pore and confining pressure on the failure process in sedimentary rock. Quart. Colo. School Mines, 54 (3): 177-199. SABATIER, G., 1959. Recherches sur la dGformation sous charge ft haute tempGrature de quelques roches 6ruptives. Bull. Soc. Franc. Min&al. Crist., 82:3-11. SANEORD, A. R., 1959. Analytical and experimental study of simple geologic structures. Bull. Geol. Soc. Am., 70: 19-52.
Earth-Sci. Rev., l (1966)51 67
EXPERIMENTAL STRUCTURAL GEOLOGY
SCHM1DT, W., 1937. Festigkeit und Verfestigung von Steinsalz. Z. Angew. Mineral., 1 : 1-29. SUMMERS, H. S., 1932. Experimental tectonic geology. Rept. Australian New Zealand Assoc. Advan. Sci., 21: 49-75. TURNER, F. J., GRIGGS, D. T., HEARD, H. and WEISS, L. W., 1954. Plastic deformation of dolomite rock at 380°C. Am. J. Sci., 252: 477-488. TURNER, F. J., GRIGGS, D. T., CLARK, R.H., and DIXON, R. H., 1956. Deformation of Yule marble. VII. Development of oriented fabrics at 300-500°C. Bull. Geol. Soc. Am., 67: 1259-1294. VERHOOGEN, J., 1951. The chemical potential of a stressed solid. Trans. Am. Geophys. Union, 32: 251-258. VON K/,RM3.N, T., 1911. Festigkeitsversuche unter allseitigem Druck. Z. Ver. Deut. Ing., 55: 1749-1757. WALSH, J. B. and BRACE,W. F., 1964. A fracture criterion for brittle anisotropic rock. J. Geophys. Res., 69: 3449-3456. WILLIS, B., 1892. The mechanics of Appalachian structure. U. S., Geol. Surv., 13th Ann. Rept., 2:211-282. (Received September 22, 1964; revised September 13, 1965)
Earth-Sci. Rev., 1 (1966)51-67