pzz326 - trl.co.uk fileabo~ ~l the transport research laboratory is an executive agency of the...

Iby Andrew S Smith and Ansel C Dunham(University of Leicester)

I—h exed~ WV ofJiTHE DEPARTMENT

OF TRANSPORT---

Contractor Report 291

ABO~ ~L

The Transport Research Laboratory is an executive agency of the Department ofTransport. It provides technical help and advice based on research to enable theGovernment to set standards for highway and vehicle design, to formulatepolicies on road safety, transport and the environment, and to encourage goodtraffic engineering practice.

TRL also sells its services, acting as contractor, consultant or providing facilitiesandstaff ona fee-paying basis for customers in the private sector.

TRUs expertise ranges from the construction of highways, bridges andtunnels, toall forms ofroad safety, traffic control and driver behaviour.

For instance, highways and structures research at TRL develops improvedmaterials and methods which enable earthworks, roads and bridges to bedesigned, built and maintained more cost-effectively. New ways of reinforcingearth can cut construction costs, while bridges can be modified to reducecorrosion. Road surfaces developed at TRL can reduce noise and cut wet-weatherroad spray from lorries by 90 per cent.

Safety research varies from monitoring the incidence of drinking and driving anddevising ways of reducing it, to improving junction designs and cooperating withEuropean partners on new standards for improved impact protection in vehicles.

Traffic research seeks to make the most of existing roads by, for instance,improving traffic signal coordination and devising systems which help driversavoid congestion. Other research looks at the effectiveness of parking controlsand improved crossings for pedestrians.

TRL research informs Government transport policy by studying, for example, theeffects of bus deregulation and how land use interacts with the road and railtransport system.

TRL employs around 600 scientists, engineers and support staff and is headed by aChief Executive. Its headquarters are at Crowthorne, Berkshire where its facilitiesinclude a 3.8km test track, a separate self-contained road network, a structures hallwhere bridge structures can be stressed to breaking point, a facility for carrying outaccelerated tests on road structures and advanced computer systems which areused to develop sophisticated traffic control programs. A Scottish section inLivingston provides advice to highway authorities on local problems such as rockengineering, bridges, road construction and maintenance.

A large proportion of the research is sub-contracted to industry, consultants anduniversities. The Laboratory also collaborates with local authorities and otherorganisations within Europe and elsewhere. In addition, TRL expertise isprovided to developing countries as part of Britain’s overseas aid programme.

For more information: TRL Public Relations, 0344770587

,,

TRANSPORTRESEARCH LABORATORY

Department of Transpofi+Zak

Contratior Report 291

ALWLI-SILICA-REACTION IN CONCRETE: A SURVEY OF UNDULATORYEXTINCTION OF QUAR~ IN GRANITES AND SANDSTONES

by Andrew S Smith and Ansel C Dunham(Industrial Mineralogy Unit, Depatiment of Geology, University of Leicester)

Copyright tintroller HMSO 1992. The views expressed in this pubihtion are not naessari~thoseof theDepartment of Transport. ~racts from the text maybe reprodwed, exapt for mmmemial pupses,provided the soum is achowledged. The wok de=ribed was =rried out undera mntract placed on theUniversity of Lekester by TRL.

The wok de~ribed in this repo~ forms part of a Se~m Reseamh Proj~ (DDRW) funded reseamhprogramme mndtied by the Transport Reseamh Laborato~.

Highways Resource Centre

Transport Research Laborato~

Old Wokingham Road

CroMhorne, BerkshireRG11 6AU

1992

Undulato~ Extinction of Quati in Granites and Sandstones 1

CONTENTS

Page

3

3

33

44588

13

16161618181818

1818192020202121

21212123

23242428292929

31313333

39

Sedion

1,0.

2.0.

3.0.

4.0.

5.0.

6.0.

7.0,

8.0.

9.0.

ABSTRACT

INTRODUCTION

THE ALWLI-SILICA REACTION; the role of strained quati3.1. The Reaction

UNDUUTORY ~lNCTION IN QUAR~4.1, Introduction4.2. Previous Work4.3. Undulatory Extinction Measurements: Flat Stage or Univer~ Stage?

4,3,1. Theoretical Evaluation

THE MECHANICS OF QUAR~ DEFORMATION

SAMPLE COLLECTION6.1. S.W.England6.2, Wales6.3. hke District6.4. Pennines6,5, Scotland6,6. Sampling Technique

METHODS FOR MEASURING UNDUUTORY EXTINCTION IN QUAR~7.1. Introduction

7.1.1. Universal Stage Method7.1.2. DeHills/Cowalhn Method7.1.3. Dolar-Mantuani Method7.1.4. AU~ Extinction point - extinction point on the universal stage7.1.5. ROTA Extinction point - extinction point on the flat stage7.1.6. Test of reproducibility of the measurement of the apparentundulato~ extinction angles, and the Dolar-Mantuani methodof measurement of strain in quati grains

7.1.6.1. Introduction7.1.6.2. Data7.1,6.3. Discussion7.1.6.4. Relation beWeen the apparent undulatoryextinction angle and the Dolar-Mantuaniextinction angle7.1.6.5. Universal stage measurements7.1.6.6. Conclusions

7.2. Results7.2.1. Summary of findings

7.3. Other methods of investigation7,3.1. Point counting7.3.2. X-ray diffraction broadening

QUAR~ GRAIN TEXTURAL CHARACTERISTICS8.1. Introduction8.2. Quati grain textures of milected samples8.3. Results

DISCUSSION

Undulato~ Mnction of Quartz in Granites and Sandstones 2

3941

43

44

44

474749495151

52

58

61

63

9.1. Conclusions of the survey9.2. Suitability of the Concrete Society guide-lines

10.0 AUTHORS’ RECOMMENDAmONS

ACKNOWLEDGEMENTS

REFERENCES

Appendix Al .0. Undulatory extindion measurement promduresAl. 1. Universal stage methodAl .2. AUEA methodAl .3. DeHills/Cowal~n methodAl .4, Dolar-Mantuani methodAl .5. ROTA method

Appendix A2.0, Point count data and undulatory extintiion measurement statisti~

Appendix A3,0. Crystallite size determination from x-ray line broadening

Appendix A4,0. Samples

Appendix A5,0, Terminology

Undulatory Extinction of Quartz in Granites and Sandstones 3

1.0. ABSTRACT

Undulatory extinction in quartz is a wide spread feature found in many different rock types. Thedegree of undulosi~ exhibited is a function of the processes of rock formation and deformation. Asundulosity suggests a level of instability of the qu-, the degree of undulosity has been used, for sometime, to indicate the potential reactivity of the quartz in the alkali-silica reaction. The method of routinemeasurement of the degree of undulosity, on investigation, has been found to be inaccurate. A newmethod for this routine measurement is proposed, utilising the universal stage. assembly fixed to ametrological microscope, Instability of quartz grains can also be gauged by the assessment of graintextures, and grain and crystallite sizes, Textures and grain size are classified by the visual featuresexhibited when the grains are viewed using a metrological microscope, whilst c~stallite size is measuredby the use of X-ray diffraction line broadening techniques. hdysis of a number of quartz-bearing rocktypes; mainly granites, gneisses, schists, porphyrites, and sandstones, from different parts of GreatBritain has been carried out. The results indicate that if atl these features are investigated and quantifiedthen it is possible to classify the stability of the quartz, eg. a mean true undulato~ extinction value of >5degrees would be classed as highly strained, and therefore assess the potential for reaction in thealkali-silica reaction.

2.0. lNTRODU~lON

Undulatory extinction in quartz is believed to result from strain within individual grains, causingdifferent parts of the original crystal to have slightly different orientations. The boundaries betweendifferent areas are sometimes diffuse or can be relatively sharp; both are probable areas of highdislocation densities. This structural mismatch within original grains means that there is more internalenergy present, tiich means that the quartz is more chemically reactive. Hence, it is possible that suchquadz could react with the alkalies in concrete to form the deleterious gel responsible for the expansionfound in concrete structures. The reaction is known as the alkali-silica reaction,

Strained quartz is known from a number of different types of geological environment; granitesand other quartz bearing igneous rocks, sediments derived from such rocks and metaquatizites formedby metamorphism.

The purpose of this project is to investigate the distribution of strained quartz-bearing rocks inGreat Britain, and the method of measurement, with particular reference to those that are or might beused for concrete aggregates. The investigation is conducted in the light of suggestions made in theConcrete Society’s Technical Repod No.

3.0. THE ALWU-SIUCA REACTION;

The atkali-silica reaction (ASR),

30:--

the role of strained quati

also known as ‘concrete cancer’, has, since the 1930s been aproblem in the construction industry. Since its identification in the late 1930s by Stanton (1940), ASRresearch has provided indust~ with a lot of possible methods of identi~lng, and even preventing thereaction taking place in concrete structures.

3.1. The Reaction

The reaction in question is one of the three ‘Nkali-Aggregate Reactions’, that take place inconcrete. These are:-

Undulato~ Mnction of Quati in Granites and Sandstones 4

1. Afkali-Silica Reaction (ASR)2. Nkali-Silicate Reaction3. Nkali-Carbonate Reaction

me ASR is a reaction between the hydroxyl ions in the pore water in concrete and certain formsof silica, ~vian, 1951). Normally calcium, sodium and potassium ions are present in the concrete. Thesurface of any silica particle, when immersed in water, displays a weak acid characteristic whichincreases with increasing surface area or increasing disorder. An acid-alkali reaction takes place at theaccessible surfaces of the silica forming a hydrous silimte. Hydroxyl ions are imbibed into the silicaparticle and some of the silicon-oxygen linkages are attacked, weakening the bond locally. Sodium andpotassium cations then diffuse to maintain the electrical neutrality and attract water to form a gelatinousalkali-metal-ion hydrous silicate, (Hobbs, 1988). The gel swells with the addition of water which exertsgreat pressure within the concrete leading to the breakdown of the concrete body,

As mentioned previously the surface characteristics of the silica, its surface area or structuraldisorder, are regarded as playing an impotint role in the reaction. Sitica mmes in many different forms,relating to its atomic structure or ordering of the Si-O tetrahedral, At its most disordered the silica formis known as opal, Opal contains a relatively large percentage of water (6 to 10YO)within its structure,possibly due to the random orientation of the Si-O tetrahedral. Opal-bearing rocks have been found tohave a very high potential for the ASR reaction. At the other end of the scale is the most commonvariety of silica, quartz, Quartz is well ordered, being made up of a regular configuration of” the Si-Otetrahedral,

In the Akali-Silica Reaction quartz has been regarded as a low potential reactant due to itsmore stable atomic configuration. A problem with this view is that quartz is often found (using ametrological microscope) to exhibit strain in the form of deformation shadows (undulatory extinction).Quartz showing a high degree of this deformation is regarded as being potentially reactive, (ConcreteSociety Technical Report No 30, 1987). This assumption is based on the idea that quartz showingundulatory extinction has a greater sudace area due to an increase in the dislocation density,dislocations being micro cracks in the atomic structure of the quartz, The higher the degree of strainthat the quartz exhibits then the greater the surface area available for the reaction.

Whether or not strained quartz is a reactive component in the ASR has for some time been indispute. In the last 10 years evidence has come into the literature that indicates that it does plays a rolein the reaction, (Mullick, et a/, 1985 and 1986, Buck, 1986, Grattan-Bellew, 1986, Adersen, et a/. 1989,and Rae, et a/. 1989), This has upheld the belief that Mielenz, (1954 and 1958), Brown, (1955), andDolar-Mantuani, (1981) suspected, that strained quartz was potentially reactive under the rightconditions. With this information in mind further research into the reactivity of strained quartz is required.

4.0. UNDUMTORY E~lNCTION IN QUAR~

4.1. Introduction

The term ‘undulatory extinction’ refers to an optical characteristic of quartz, when seen in thinsection (of 30pm thickness) and viewed in cross polarized light using a metrological microscope. Underthese conditions, when the microscope stage is rotated the quartz grain is seen to go into extinction(becomes isotropic, i.e., black) in a non-uniform manner, i.e., the whole grain does not reach the samedegree of extinction at the same time, zones of partial extinction occuring alongside zones of totalextinction. This non-uniformity of extinction can, in certain cases, be identified by the sweeping motionof the zone of extinction across the grain upon rotation of the stage. In other circumstances undulato~extinction appears as segmented extinction zones within a single grain boundary. The term “strainedquartz’ is often used when describing quartz showing undulato~ extinction. This is because undulatoryextinction is an optical representation of the degree of strain within the crystal lattice of the quati grain.The internal strain causes a displacement of the c-axes of the quartz thus inducing undulatoryextinction.


4.2. Previous Work

In the past a large amount of work has been carried out on quartz , especially on the origin(Bailey, et al., 1958), mechanics of deformation and recovery (Bailey, et a/., 1958; White, 1973), themeasurement and application of quartz showing undulato~ extinction for the provenancing of sediments(Blatt and Christie, 1963; Conolly, 1965; Nakashiro, 1965 ; Blatt, 1967; Basu, et a/., 1975; Young, 1976),and the possible relationship between the degree of undulosity and the age of the rock type (DeHills andCorvalan, 1964).

Figure 4.1.1. Undulatory extinction as seen in a quartz grain viewed under cross polarized light.An example of a quartz grain from the Arran Granite.

Undulatory extinction has been used as a means of identifying quartz under the microscopesince petrographical analysis began. This would imply that this characteristic of quartz is commonlyfound. Historically it was thought that undulato~ extinction indicated a metamorphic origin. This is nowthought not to be the case. Blatt and Christie, 1963, state “Tuttle (1952) and Gilbert (1954) have bothstated explicitly that undulato~ extinction in quartz is not restricted to metamorphic rocks. A review ofthe literature on the petrography of plutonic igneous rocks would show that the existence of undulatoryquartz in granitic rocks is frequent enough to be considered commonplace.” They went on to state “It isquite clear that plutonic igneous rocks, schists, or gneisses containing large percentages of quartz withnon-undulato~ extinction are uncommon. In the majority (70.3%) of these rocks, less than 10 percentof the total quartz is of the non-undulato~ type.” A corresponding study of extrusive igneous rocksshows that non-undulatory quartz is the norm, making up 91YO of the total quartz.

Blatt and Christie (1963) have presented a historical review of the development of theclassification of undulato~ extinction and polycrystallinity in quartz. These methods of classificationwere developed to aid in the provenancing of sedimentary rocks. Van Hise (1890) was the first personto identify undulato~ extinction, though this was attributed to dynamic metamorphism in sedimenta~rocks. Rosenbusch (1893) appears to be the first to have documented the existence of undulosity inquartz from a granitic origin. Grout (1932) distinguished igneous and metamorphic produced undulosityby the fact that igneous quartz showed “wavy extinction” and that the undulatory extinction (straineffects) in gneisses “were more pronounced and in more nearly parallel bands. ” Krynine (1940) definedseven categories of quartz based on the polycrystallinity, degree of undulatory extinction or lack of it,


and the shape of quartz grains in sandstones, This classification was as follows (taken from Blatt andChristie, 1963);

1, “Normal igneous quartz”, possessing non-undulatoryextinction,

2. “End phase quartz igneous quartz formed under slightpressure during the rest magma stage, ....Strain shadowsmay be weakly developed, ”

3. “Hydrothermal quartz”.4. “Modified igneous quartz: The strain shadows are ve~

strong,,,.,. but borders are still smooth.5, “Coarse-grained fragments or single grains of

subsequent metamorphic quartz derived from coarsequartzites and elongated (Iensoid) grains from coarse-grained schists and probably gneisses. Characterizedby undulato~ extinction, crenelated borders, and insomeplaces, inclusions of kyanite and sillimanite .....”

6. “Quartzitic quartz, made up of joined grains of allsizes, ....”

7. “Schistose quartz, distinguished from quartzitic quartzby the elongated character of the grains,,,,..”

Follow up work by Krynine (1946) made a modification of the previous work and produced threetypes of “igneous quartz”:

1, “Plutonic or granitic”. An amalgamation of categories1, 2, and 4.

2. “Hydrothermal”. Characterized by “comb-structure” andabundant bubble inclusions.

3. “Volcanic quartz”. Idiomorphic shape and water-clearappearance.

and from “metamorphic rocks”, two types:

1. “Pressure quartz”. Elongated, polycrystalline quadzaggregates with crenelated borders and strongundulatory extinction.

2. “Injected quartz”. Elongated single crystal unitswith, in general, non-undulatory extinction.

Folk (1961) recognised that quartz of different origins could exhibit the same characteristics,resulting in a six tier classification: (taken from Blatt and Christie 1963)

1, Quartz with non-undulatory extinction.2. Quartz with slight undulato~ extinction, The

extinction shadow sweeps across the grain on rotationof the stage of less that 5°.

3. Quartz with strong undulato~ extinction, in which thegrain requires more than 5° of stage rotation for thestrongest part of the extinction band to sweep acrossthe grain,

4. Polycrystalline quartz in which the grain is composedof elongate crystals with very close opticalorientation. Extinction generally non-undulatory,Frequently charged with water bubbles.

Undulatory ~nction of Quartz in Granites and Sandstones 7

5. Polyc~stalline quartz in which each c~stal unit hasnon-undulato~ extinction but tidely differingextinction position from its neighbors in the grain.

6, Polyc~stalline quartz in which each crystal unit hasstrongly undulatoy extinction and commonly hascrenelated borders.

Bailey, et a/,. (1958) suggested that undulatory extinction is the mmmonest form ofmicrostructure in quartz, and that this microstructure resulted in complex undulato~ extinction patterns,depending on the development of cataclastic features, such as granulation and fracturing. In generalthe more deformed the rock type the more pronounced the development of well defined zones ofundulatory extinction, oflen reported as being parallel or near parallel to the c-axes.

The distinction between non-undulatory and undulato~ extinction in quartz has had significantdiscussion in the literature. Bailey, et al., (1958) proposed a three tier (1, 2a, 2b) classification foridentifying the level of deformation of the quartz. This classification was based on the appearance of thequartz and the general texture of the rock.

1.

2a.

2b.

Undeformed; included grains that showed uniformextinction, though some had weak and broad undulatoryextinction (strain shadows) with a variation of up to4°. These grains were seen in rocks such ashydrothermal quartz veins, mica schists, quartzites,and most dykes, sills, and rhyolites in addition tosome reportedly magmatic graniticrocks,

Cataclastic; strain shadow dominant; included rockswith developed undulatory extinction (strain shadows)and tended to be of a magmatic granitic nature. Thedisplacement of the c-axes varies between 10 to 7° evenwithin the same thin section. Most of the extinctionbands were of a broad and straight nature tith novisible fractures between them.

Cataclastic; shattering dominant rocks showed a moreintense deformation. l~general rocks of a moremassive granitic texture. The undulato~ quatiz, insome cases, develops into sinuous bands of undulatoryextinction, c-axes variations from 3° up to 25°. Thezones of extinction were observed to be separated byfractures, in addition to some quartz grains beingbroken along their borders and rec~stallizing intosmaller, less strained, fragments.

The most significant problem is how to distinguish between slightly undulato~ and stronglyundulatory quartz. Hubed (1960) suggested the this distinction should be at 25° rotation of the flatstage microscope, whereas Folk (1961) subsequently put forward 5° of rotation of the stage, andAndresen (1961) proposed an angle of 30° rotation. In Blatt and Christie’s (1963) investigation thedistinction point was put at 10 of rotation of the flat stage, It should be noted that these measurementswere an attempt to ‘classi~ quartz grains into either an igneous or metamorphic origin, for studies intothe provenancing of sediment, using the idea that undulatory extinction is a characteristic ofmetamorphic derived material as opposed to an igneous derived material.

Blatt and Christie (1963) were the first to oint out that measurements of undulosity on atflat stage could produce any angle between O and 90° (actually 160°, see Se~on 4.3.1.)

depending on the orientation of the of the c-axes. As these axes are present in 3 dimensions, theorientation of both c-axes is unlikely to be in the plane of the flat stage. Therefore measurementsare far from a true representation of the angular displacement of the two c-axes. They concluded

Unduiato~ Extinction of Quati in Granites and Sandstones 8

that the only reproducible metiod of measuring undulatory extinction in qua- was to use auniversal stage.

4.3. Undulato~ Extinction Measurements; flat Stage or Universal Stage?

Much has been written on the use of measurements of the extinction angles tithin strainedqum. both from the point of view of the provenance of the qu- in geologi- studies, and as a meansof assessing possible reactivity in aggregates. However, dl these studies depend on being able to m~eaccurate, reproducible measurements which genuinely reflect the angle between Me c-axes of thedifferent strained portions of a grain. me inter-mid angles can be measured with great accuracy with auniversal stage mounted on the normal microscope stage. ~is argument was described in detail byBlatt and Christie (1963), and is presented in the following ~eoreticai Evaluation” section.

4.3.1. ~eoretical Evaluation

me problem of measuring extinction angles tithout a universal stage lies in the fact that the c-taxes of the strained parts of a quati grain may. lie in any orientation. %e relationship between trueand apparent angles of undulosity in a quati grain is a function of (1) the angle between the planecontaining the the c-axes and the plane of the thin section, and (2) the pitch of the axes themselveswithin the optic plane (Pautitsch and Ambs, 1963)’, in Basu, et a/. (1975). Blatt and Christie (1963) havepresented an elegant description of the problem. ~eir fig. 1 is reproduced here as figure 4.3.1,

0c--------------

------ ~-

--.. <’----------<-:,--)

Y,

~

‘. 1’ *.,8.. . . ,’‘., ,1

c’,\30. \& - .>--------

,,,,

\

d

figure 4.3.1. Effects of c~stallographic orientation on the degree of undulatory extinction visiblein a randomly cut thin section. (After Blatt, et d. 1963); a is the thin section and b, c, and d arestereographic projections.

figure 4.3. 1(a) shows a strained quati grain, in tiich dl the c-axes are lying in the plane of thepaper (or microscope thin section). me projection of the poles of c and ~ we shown in figure 4.3.1 (b).~is is the on~ orientation in which the conct inter-aial angle can be measumd on the flatstage.

Undulato~ Extinction of Quartz in Granites and Sandstones 9

If the same grain is rotated so that the c-axes are dl in a verti~ plane, then the extinction anglemeasured on a flat stage will be zero. ~is situation is represented in figure 4.3.1(c).

A more general case is noted in figure 4.3.1 (d), where the c-axes c and c’ ~win a plane steeplyinclined to the plane of the thin section. me red angle of 30° is represented by a measured angle of68° on the flat stage.

An real angle between the c-=es can give rise to an apparent angle which can vary from20° to 180 , depending on the inclination of the plane containing the c-axes to the plane of the thin

section, and the plunge of the axes w.thin the plane.

me two limiting cases are illustrated in figures 4.3.2. and 4.3.3. In both cases shown in figures4.3.2(a) and 4.3.2(b) the true angle between the c-axes is 20‘. In Figure 4.3:2(a) one of the axes is in

the plane of the thin section. me other lies in a plane which varies from horizontal to vertical in 10°steps. it the axial plane is vertical, then the difference between the extinction positions is zero. As theplane is inclined away from the verticaf, so the apparent difference in extinction angle increases up to20°, when the plane is parallel to the thin section (honzontaf).

figure 4.3,2. Stereographic projections showing two examples of the possible variations of theorientation of the c--es in a c~std.

In figure 4.3.2(b) the two poles are symmetrically arranged about a vertical E-W plane acrossthe stereogram. When the plane is vertid, the extinction difference is 180‘. As the plane is rotatedtowards the horizontal, the extinction angle difference reduces until the plane is horizontal, when the trueangle is measured. Small changes in the attitude of the axial plane tien it is steeply inclined (close tovertical) produce large variations in the apparent extinction angle, whereas changes when the plane isneady horizontal produce much smaller variations:

me variations in apparent extinction angle with inclination of the axial plane are shown ingraphid form, for a true angle of 20°, in figure 4.3.3. As these two situations are the timiting cases,then any other orientations of the c-axes will produce apparent extinction angle differences between thetwo cuwes. Similar diagrams could be made for other real extinction angles; the form of each diagramwill be similar, but with the true extinction angle varying on the right hand verti~ axis.

Dolar-Mantuani (1983) and others -e the important point that only grains tith the highestpossible birefnngence should be measured; this indicates that the c-axes of the c~std are near to the

Undulatory ~tinction of Quartz in Granites and Sandstones 10

plane of the thin section. However, the birefringence can on~ be a matter of judgement, not ofmeasurement, unless the thickness of the thin section is ~ known. Inaccuracies can occur in themeasurements if only one of the ~es falls in the plane of the stage, as the angle being measured here issmaller than the true angle if both axes were in the plane of the stage.

We can see no merit in Dolar-Mantuani’s (op.&) suggestion of measuring the first and lasthazy extinctions (A detailed explanation of the method is described in appendix Al .4). me size of thisangle is probably a function of the angle of inclination of the optic-axial plane, and has no relevance tothe strain within the grain. me conclusion that can be drawn from this is:

me only reliable optical method for obtiining tie true angle of undulosity etindon is byuse of a universal stige.

I9I

\

\

I

\

\

\

\

\

\

t

\

\

\

\

\

figure 4.3.3. Variations in apparent undulato~ extinction angle with inclination of the axial planein graphi~ form, for a true angie of 20°

A counter argument to Blatt and Christie’s has been put fomd by Basu, et d. (197~, in ~ichthey argue that the use of the flat stage method of measuring the degree of undulosity is still accurateenough if a ●inimum number of constraints are employed.. ~ o-ves of these investigations wereto ~ify the quartz type into @utonic or metamorph~ derived Wtment grains. ku et af. (197~ sfsomnsidered the problem of measurement of dnction _ m ~.ned q- ~ey measured tiehe angles of more than 900 g~ns from granites and metamorphic rocks, using a unive~ stage. ~ey


showed that quartz in granites has a mean angle of undulosity of 3.6° (standard deviation = 3.5°) andmetamorphic quartz has a mean undulosity angle of 7.9° (standard deviation = 5.00). Their plot isreproduced in figure 4.3.4. Note that the bulk of measurements tie below 15° (he angle of undulosity).

Basu, et d. (1975) proposed that the distinction between the two categories should beat the 5°level as measured on the universal stage. Measurements were made of 355 pi@onic qum grains and571 low grade metamorphic quanz grains tith the result that 78% of the plutonic qum gtins exhibiteda tme undulosity value of S5° whereas 66% of low grade metamorphic quartz exhibited a true undulosityvalue of >5°. In addition only 7% of plutonic qu- had a he undulosity vatue of >10° in comparisonwith 23% of low grade metamorphic qua~

me investigation that Basu, et al. (1975) undeflook was to see what percentage of grains, whenmeasured on a flat stage, whose apparent undulosity was >@ had corresponding tme undulatoryextinction angles (universal stage measurements) of s+, ie. error vatues. The discussion of therelationship between true and apparent undulatory extinction angles presented by Basu. et d. (1975)pages 875-876, are similar to those in Blatt and Christie (1963) which is explained in detail in thefollowing section. me difference in the interpretation of the relationship is the factor which differsbetween the authors. figure 3. (page 876) in Basu, et d. (1975) shows the findings of the theoreticalrelationship behveen true and apparent extinction angles, and is show in figure 4.3.5.

.

,,

Iulonic quarrz

(n= 355)

..<“.....,,i

. “...Metamorphic quartz... ... (n=j71): .... ‘,..: .

:

.. ... ... “..-.-r.1 ........ -. . - .................. -...*___,..-”.

True Afiqle of Undulosity

figure 4.3.4. Histogram showing the distribution of true angles of undulosity in detriti qua~ ofplutonic and low grade metamorphic parentage. Values based on universal stagemeasurements. (Afier Basu, eta/. 1975)

me curved tine is a product of values of true extinction angles, 2°, 4°, 5°, 10°, 15°, 20°, and30° from which the mrresponding apparent extinction angles have been determined. me area (a)below the curve represents the percentage of apparent angles >5° with corresponding true angles ofs#. ~is value of 7% can be regarded as the error value, which implies that 93% (area @ of the @ure)

&of grains showing >@ apparent unduiosity have conespondingtrue extindon angles of > afso.

Men looking at the relationship between grains of true undulatory etinti.on angles of <@,56% of the grains will exhibit corresponding apparent undulatory extinction angles of “s5° (area (c)of the

Undulato~ ~tinction of Qu- in Granites and Sandstones 12

figure). This corresponds to an error value of 44% for this relationship (area (d) of the figure). Toconclude; there is a ve~ high probability that a true undulatory extinction @ue of >5° will have acorresponding apparent undulato~ extinction angle value of >5° but, the same can not be said for trueundulato~ extinction angles of C5‘, for which there is only a 56% to44Y0 chance of the apparentundulato~ extinction angle being dso <5°.

Around 600 grains were measured by Basu, et al. (1975) for both apparent undulatory extinctionangles and true undulatory extinction angles. Their results are shown here as figure 4.3.6., their figure4.

Figure 4.3.6. shows a tendency for points to fie above the fine of equtity. This would imply thatthe apparent angle of undulosity is more likely to be less than the true angle of undulosity.

100

o

1A1’n/,,,,,7 .!:,(

93 “1.

b

Figure 4.3.5. Cume showing the percent apparent angles of undulosity >5° and <Socomesponding to specific true angles as determined with a stereographic net. (After Basu, et al.1975)

A fimiting factor in theory of Basu, et d. is that the spread around the line of equality is too greatfor this argument to be used when the actual true degree of undulosity is required. If a fixed angle ofundulosity is to be employed, as in the tincrete Society Technical Report No 30 (198~ which proposeda mean undulato~ extinction angle of 25° to distinguish highly strained quati from strained qu~ thenthe absolute angle of undulatory extinction is required.

As the statement above says ‘the apparent angle of undulosity is tikely to be less than the trueangle of undulosity.’ This being the case, under-estimates of the true degree of undulosity would be

_*t ~ a large ~ge of possible tie vafues, eg from Figure 4.3.6. if the apparent angle ofundul~ obtained was # then the correspondinghe angle of undulositymuld fall anywhere afongthe vertid fine, corresponding to a ~.ble range of between 1‘-l 1‘. For accurate reproduciblemeasurements this inaccuracy is too great therefore the on~ reproducibly accurate method of


measuring the degree of unduiosity is that employing the use of the unive- *ge. Unfofiunately, suchdevices are becoming rarer, and the number of petrographers trained in their use is even smaller.

- result many studies have relied on measurements made on the flat microscope stage. Anexample is the method proposed by Dolar-Mantuani (1983) for the assessment of strain in quartz inaggregate for making concrete.

<n A

\\\] /. “/. I. . . .. ./”. ●. “/. ”

. . /“. . . . “/ . .-

.../. ./...”.. . . .. .. . .. /. .

: :.:;2:”;0.. ... . . .. ...:::X : . . .kl10 ‘ \.\.

:,.\,\\: .,>;:.7 . . . . . . .

. . . . . .

.

. . .

I

i s !0 Is a ~s 30Apoarenr anqle of unaulasify

figure 4.3.6. Relationship between true and apparent angles of undulosity in actual grains asdetermined using the universal stage. Dots of variable size represent points where more tianone grain had the same value of true and apparent undulosity. The dashed line has a slope of45° All points on this line have true angles equal to apparent angles. (Afier Basu, et d. 1975)

5.0. THE MECHANICS OF QUAR~ DEFORMATION.

Strain occurs in c~stals as a result of the action of a differential stress system within the plastic,viscous or “creep’ field of the straidtime graph, Figure 5.1a,b. The conditions generally required forductile deformation are relative~ high homologous temperature (actual tempflemp of melting), low strainrate, relatively high hydrostatic pressure and long periods of deformation. Experiments have shown thatqu~ crystats deform in a ductile manner tith a ptilctable sequence of changes in microscopic andsub-microscopic characteristics Noung, 1976). ~ can be seen in figure 5.1a below, the result of thep-tic deformation or creep is the permanent strain of the material involved.

Undulato~ ~tinction of Quati in Granites and Sandstones 14

me degree to which a quti grain shows stored strain characteristics, in the form of unduiato~e~inction, is not necessarily an indicatorof the level of defamation that the grain has undergone. ~isis due to processes of “recovev” that can take place at the same time, thus having the effect ofreducing the stored strain, which atlows the processes of defamation to continue past the point wherecataclasis (brittle failure) would take place.

lntrac~stalline deformation of quartz takes place on the microstructure scale due to thepresence of minor imperfections in the atomic lattice. ~ese imperfections can be grouped into threetypes:-

i) Point Defectsii) tine Defectsiii) Planar Defects

.Scousfl ----- I

-. . . . . . . ------- - ------ ------- ------- ------- ------

/I.. . . . . . . . . . . . . . . I total elastic componentI

visco-elastic !

/

.----------------- yield’point

elastic II

permanent strain1

L I 4

stress applied stress removed nME

b

failure

).

accelerated-time

7-”--”-“-“ ~V...............visco-elastic..............elastic

figure 5.1a,b. Graphs showing the conditions leading to plastic deformation.

Point defects are zero dimension imperfections in the atomic bttice. ~ese can take the formof missing atoms, ie. vacancies which weaken the ovedl bonding of the atoms, or can be tierep~cement of certain atoms by a different type, Mich dso produces a change in the bond stren@and thereby a potential weak point M ~mple of such a substiM.on is that of the replacement of theSi-O-Si bond in qu~ by the much weaker Si-OH:HO-Si bond. ~i -pie dso fll~tes the


phenomena of ‘water weakening’, so named because the presence of water in the c~stal structure (inthe form of hydroxyl) renders the material weaker.

tine defects, which are referred to as dislocations, separate slipped from unslipped areas of theatomic lattice nd are probably the most important agents in deformation. On a microscale, deformationis the formation and movement of line defects. Slip occurs in the atomic lattice due to the increasedmobility of the defects, Most slip starts off at a focal point and moves outwards, but stays within one slipplane.

A dislocation is made up of two components, an Edge mmponent and a Screw component,depending from which direction the dislocation is being viewed. The edge component operatesperpendicular to the direction of propagation of the dislocation line (Burgets Vector) whilst the screwcomponent acts parallel to the Burgets Vector,

Planar defects are the amalgamation of multiple line defects in one region of a grain. In quartzsuch defects occur as the dislocations work through the grain and then tangle, halting the movement ofeach other; work hardening. This process is prevalent when the homologous temperature is low, Asthe dislocations are the areas where the residual strain is being carried such a tangling together of thesedislocations localises the strain into specific areas or bands. men the disorientation of the bands isslight (less than 1‘) undulatory extinction is seen as a continuous sweeping motion throughout thegrain.

As more and more dislocations tangle up, it becomes harder for further dislocations to move.This process of ‘work hardening’ or ‘strain hardening’ reduces the ductile nature of the grain and thusleads to cataclasis of the grain,

With medium to high homologous temperatures the process of deformation is active alongsidethe process of recovery. Recovery has the effect of reducing the stored energy in the crystal so that thepoint of cataclasis is not reached. Where the degree of deformation is greater grains progressivelydevelop a greater angular mismatch (failure in the cVstal structure). In addition the grain is seen to bemade up of bands or segments of va~ing positions of extinction divided by well defined boundaries.This process is known as ‘polygonization’ which results in the formation of banded or segmentedundulosity.

With increased temperature the dislocations migrate and further development of the extinctionbands, also known as ‘deformation bands’ mite, 1973). As polygonization and deformation banddevelopment progresses the crystals bemme more elongated and the grain boundaries becomesutured. Suturing of the crystal boundaries reflects gross discrepancies in the strain energy either sideof the boundary formed as a response to the local build up of high densities of dislocations.

At medium to high homologous temperatures the process of recovery is found in addition to thedeformation processes which distods the identification of the actual amount of strain the grain has beensubjected to. During periods of raised temperature the cataclasis phase of deformation can be avoidedand a reduction in stored strain achieved by these recove~ processes.

Subgrain development can be regarded as one of the recovery processes as it follows on fromband deformation. In its simplest form recove~ can take place without any further deformation, ieannealing, when the increase in temperature allows the dislocations to have a greater degree of freedomand therefore enables them to move out of the original plane into others (three dimensions).

The further stage of recove~ is that of ‘prima~ recrystallization’. This is the next step on fromthe development of deformation banding. The difference between primary recrystallization anddeformation bands is that the individual recrystallized grains are themselves small (s50 #m), strain free,showing normal, even extinction with a slight mismatch beWeen the grain and its neighbour. ~esenew crystals form along the margin of the host grain or atong deformation bands (Hobbs, 1968). Thenew crystals develop either by continued polygonization so that all the crystallite become new crystafs(syntectonic recrystallization) according to Hobbs, 1968) or by nucleation at the borders of crystallite inpolygonized crystals where the strain energy is great enough (Phillips, 1965)” (as in Young, 1976).


Another process of remvery that can produce new grains is that of ‘grain bounda~ migration’,This is activated by the difference in stored strain at the boundary between the strained grain and thestrain free grain, The grain boundaries migrate throughout the aggregate, strain free grains growing atthe expense of the strained grains, by atom to atom movement.

After initial recovery the dislocation density is reduced but the aggregate still contains storedenergy at the grain boundaries, During the period of the increased temperature regime further recoverycan take place by ‘normal grain growth’. This process acts by the removal of the complex grainboundaries, produced by primary rec~staliization, by further grain boundary migration, ending with aregular polygonal grain structure, showing triple junctions of roughly 120°.

Further recove~ can then take place by ‘seconda~ recrystallization’ which involves the growthof regular shaped grains at the expense of grains which have complex grain boundaries, This reducesthe surface area of the grains, therefore lowering the surface stored energy. A feature of seconda~recrystallization is the development of ‘core and mantle’ textures where a single large crystal issurrounded by smaller crystals,

One complication with grain deformation is that the process of deformation and recovery can actat the same time or staggered, with phases of deformation followed by recovery, so that the point where“work hardening”or ‘strain hardening leads to cataclasis is not reached. This allows deformationprocesses under the ductile state to continue over long periods of geological time.

It is possible, from the processes of deformation and recove~ described previously, to regardthe deformation and recove~ of a material, in this case quartz, as a cycle. The processes progressivelyincrease the stored strain in the crystal to a point tiere the c~stal either breaks under brittle failure(cataclasis) in a static regime (relatively instantaneous removal of stored strain) or more progressiveremoval of stored strain energy through the processes of recove~ in the dynamic regime.

6.0. SAMPLE COLLECTION

For the purpose of sampling mainland Britain was divided up into 5 sampling areas, S.W,England, Wales, Lake District, Pennines, and Scotland. These areas were selected as they containknown sources of aggregates used in the instruction industry. The aim was to sample as many of theknown, and potential sources that contained quartz as a major mineral in the rock type. The rock typesincluded granites, greywackes, meta-siltstones, vein quart~quartzite, gneisses, and sandstones. Ageneralised map of the sample localities is shown in Figure 6.0,1.

6.1. S.W. England

This sampling area covers the two counties of Devon and Cornwall. The majority of aggregateproduced in this area is from the local Hercynian Granite bodies. Sampling included the large bodies ofthe Dartmoor, Bodmin Moor, St Austell, Carnmenellis, and Land’s End Granites, in addition to the smallerbodies, Hingston Down, Godolphin, and Stilly Isles. As well as the granite bodies samples werecollected from the middle Devonian slates, the Bude Formation Sandstone, and the Budleigh SaltertonPebble Beds. In total 18 sites were sampled and initially examined.

6.2. Wales

Four samples were mllected from this area, one from S.W. Wales and three from N. Wales. Therock types selected include a quartz-diorite, part of the Johnston Complex, from Pembrokeshire,microgranite from the Ueyn Peninsula, and two samples of the @edana Granite, from the Mona~mplex on Anglesey.

Undulato~ Hndion of Qu- in Granit- and %dSonm

*IIIV

?nrl

A VT’”’””’”

17

Hl@ln Cramtta

.,

Fi~re 6.0.1. Generalised map of sample localities inGreat Britain. Granitic bodies are named and shorn tmeto size, other samples are named and located.

Undulatory Extinction of Quati in Granites and Sandstones 18

6.3. hke District

This Sampling area covers the county of Cumbria and includes sites within the Lake DistrictNational Park. Six rock types were collected; Threlkeld microgranite, Shap granite and the associatedhornfels, Silurian meta-siltstone, and Permoflrias sandstones from the St Bees Head Sandstone andfrom the Penrith Sandstone. Seven samples were collected for the study.

6.4. Pennines

This sampling area covers the county of North Yorkshire and the eastern edges of Cumbria. Althe samples collected in this area were sandstones of the Namurian period of the Carboniferous. Sevensamples were mllected for initial investigation.

6.5. Scotland

The final and by far the largest area sampled is that of Scotland. The area includes the islandsof Mull, Skye, Harris, and Lewis, in addition to the mainland in general. A large variety of different rocktypes were sampled ranging from sediments, through igneous, to metamorphic rocks. A large numberof granites were sampled including, Broadlaw, Cove, ~avie, Criffel, Creetown, Hill of Fare, Kemnay,Strontian, and the Furnace microgranite. Three quati-dolerites and a felsite were sampled from theMidland Valley region in addition to two examples of quati-bearing porphyrites (microtonalite) from theSilurian rocks in Dumfries and Gatloway. In the north and west of Scotland a number of metamorphicrocks, of differing grades, were sampled. These include Dalradian slates and quartzose-mica-schists,Moine quartz-feldspar-granulite rocks, and Letisian gneiss from the Outer Isles (Harris and Lewis).The sedimenta~ rocks collected include Torridonian, Devonian and Permian sandstones and siltstones,these totalled nine samples,

6.6. Sampling Technique

The majority of the samples collected were obtained with the co-operation of the quarries in theareas. Sampling, of approximately 4 kg of crushed rock, took place from the stock pile (+40mm) in orderto obtain a sample that was representative of the present production. Where quarry access was notpossible, samples were obtained from outcrops or from old quarry workings. The samples were thencoded, a grid reference recorded, and returned to Leicester for thin sectioning.

7.0. METHODS FOR MEASURING UNDULATORY ~NCTtON IN QUAR~.

7.1. Introduction

From the literature it is clear that different workers have differing views of the best method for themeasurement of undulato~ extinction angles. As stated in Section 4.0. the only accurate, andreproducible method of measuring the angular difference in the c-axes, is that of the universal stagemethod. This method has been taken as the primary method in this investigation, In addition to theuniversal stage method, four further methods were also used, two methods from the literature and twoderived by the authors during the development of the project. The methods used areas follows:

1. Universal Stage Method. ~UEA)2. DeHills/Corvalti Method. (DEHIUS)3. Dolar-Mantuani Method. (DOUR)4. Extinction Point-Extinction Point on Universal Stage. (AUEA)5. Extinction Point-Extinction Point on Hat Stage. (ROTA)

In light of the mmments presented in ‘Notes on Table: (2) Appendix 3, Page 30, of theConcrete Society Technical Report No 30. Hawkins, et d. 1987, 20 measurements of the undulosity of

Undulato~ ~tinction of Quartz in Granites and Sandstones 19

quanz, for each sample, using each method were obtained, where possible. Measurements of theundulosity were all made using crossed polars (analyser in and polariser in). The data obtained ispresented later in this report.

Each method is described in the following section. For details of the procedures refer toappendix Al .0.

7.1.1. Universal Stage Method.

This method of measuring the angular variation between adjacent c-axes uses the universalstage attachment to a norrnd metrological microscope. We used a Leitz S-axis universal stage mountedon a Swift microscope, figure 7.1.1. (For operating procedure see appendix Al.1.) When using theoptical hemispheres (Refractive Index 1.554) the thin section (3@m) -S kept in optical continuity byusing clove oil (Refractive Index 1.54) at contacts of glass-glass.

Two methods employing the use of the universal stage are given in the literature. Nakashiro(1965) and Gay (1982) describe a method that, with the aid of the universal stage, orientates the w c-taxes, in question, so that they lie in the plane of the microscope’s flat stage. ie. that the c-axes weperpendicular to the optical axis of the microscope. In this position, Nakashiro states that the first sign ofextinction, ie. the beginning, and end of the extinction sweep can be measured four times with a 360°rotation of the microscope stage. A possible dratiack of using this method is the problem ofreorientating the thin section so that both the c-axes are in the plane of the stage. If this method isused then the only way that alignment of the c-axes can be checked, is by the subjective eye of thepetrologist making the measurements, who has to estimate that the two zones exhibit the samebirefringence value.

The complicating factor is the accuracy that can be attained by using this method. A furtherproblem is that the birefringence value is a function of the thickness of the thin section. Values arenonaily quoted for thicknesses of 30pm so a variation of only a few microns can alter the vatuesignificantly, and therefore if the thickness of the thin section is not exactly 30p then differentbirefringence values have to be identified for the actual thickness. The thickness of the thin section isdependent upon the skill of the sectioning technician. If the section is of equal thickness throughoutthen the variation of birefringence can be accounted for, but the problems arise when the thin section iswedge shaped, ie. thinner at on end than the other. This wedge shape causes the birefringence to va~greatly across the section making accurate alignment of the two c-axes very difficult.

.

--

AI;S of microscope

Am

I A,

;5

AZ

Figure 7.1.1. The Leitz 5-~is Universal Stage assembly. (After Ernmons, 1959)


As a result of this implicating factor a semnd method of measuring the c-axes variance wasinvestigated, suggested in Gay (1982 pages 173-177). This method reorientate the c-axes so thatthey lie in the plane of the optic axis of the microscope. The procedure is given in appendix Al. 1. Thesemeasurements give the position of the c-axes as an inclination to the horizontal (plane of the flat stage)and as rotation of the stage, The data collected for each grain measured must then be plotted on astereonet to enable the true angular difference between the two to be calculated. A limiting factor forthis procedure lies in the fact that it is only possible to measure inclinations up to =60° due to thearrangement of the optics needed for the universal stage. When this is the case the previous methodcan be used, (Gay, 1982. page 177), which orientates the c-axes into the plane of the microsmpe stage.

7.1.2. DeHills/Coma14n Method.

DeHills and Corva14n (1964), in their paper investigating the relationship between the degree ofundulosity and the age of the rock type used a different method. This investigation was carried out on anumber of Chilean granites. They measured the flat stage rotation needed to rotate a given quartz grainfrom the point where hhe first clear evidence of undulato~ extinction appearedw through the point ofm~imum extinction to the point where the undulato~ extinction disappeared from the grain. Thisrotation of the flat stage was recorded as the degree of undulosity. See appendix Al .3. for the detailedprocedure,

The method includes the instruction to select grains showing high birefringence in an attempt tomake measurements when the c-axes lie as close to the plane of the flat stage as possible. Asexplained in the Universal Stage Method previously, the recognition of maximum birefringence is ve~subjective. If both c-axes are not in the plane of the flat stage then the values recorded are inaccurate.

7.1.3. Dolar-Mantuani Method.

This method of measurement of undulato~ extinction of quadz was first described in Dolar-Mantuani (1975) though a more detailed description can be found in Dolar-Mantuani (1983), Themethod is a development of the DeHills/CowalAn Method described previously. Dolar-Mantuaniproposed that a more accurate way of measuring the sweep of the undulatory extinction was tomeasure the angular variation between the first appearance of extinction in the grain to the grain’s lastfull extinction and that with the angular variation between the grain’s first full extinction and the point ofthe extinction’s last disappearance from the grain. (For detailed description of method refer to appendixAl .4.) Once again this method uses the fiat stage of a standard metrological microscope. The possibleadvantage of this method over that of DeHills and Cowalan is that a check of the two ranges can bemade against each other. In theory the two measurements should be within one or two degrees of eachother. This being the case the mean of the two measurements is taken as representing the undulosity ofthe grain.

Nthough not explicitly stated, it is the Dolar-Mantuani method that is refered to in Appendix 3 ofthe Concrete Society Technical Report No. 30 (1987).

7.1.4. AUW. ~nction point - extinction point on the universal stage.

This method of measuring the angular difference between pairs of c-axes is a by-product of theuniversal stage method. As part of the orientation procedure that is carried out in the universal stagemethod (see appendix Al .2.), it is possible to record the apparent extinction positions of the grain as if itwere on a flat stage. The actual measurement is the rotation of the stage between the two points ofmaximum extinction. This method differs from the univeti stage method in the fact that there is noallowance for the variation of the inclination of the c-axes to each other and to the stage, ameasurement that is recorded in the universal stage method and accounted for in the calculation of thetrue angle of c-axes variation. These measurements of AU= (apparent undulato~ extinction angle)measured on the universal stage are made on grains that show low birefringence, (when the c-axes arehighly inclined, as near as possible to the optic axis of the microscope), one of the criteria for theuniversal stage method.


7.1.5. ROTA Efiinction point - etiinction point on the flat stage.

Again this measurement is a by-product, this time of the Dolar-Mantuani method (see appendixAl .5.). As in the AUEA method, the ROTA method is a measurement of the rotational angle between theWo points of m=imum extinction in the grain on the flat stage, It differs from the AUEA method in thefact that as part of the criteria for the Dolar-Mantuani method grains of high birefringence are selectedfor measurement, ie. grains in which the c-axes are as close as possible to the microscope’s flat stageaxis. Once again no account is taken for the variation in the relative inclination of the two c-axes to thestage and themselves. Such measurements, Men either none or only one of the axes lies in the planeof the flat stage, will always be less than the angle measured when both axes lie in the plane of the flatstage,

7.1.6. Test of reproducibili~ of the measurement of the apparent undulato~ etinction angles,and the Dolar-Mantuani method of measurement of strain in quartz grains.

7.1.6.1. Introduction

The purpose of this investigation was to test the reproducibility of measurements of theundulatory extinction angle made by different operators on the same sample. Measurements were atsomade of the true undulato~ exthction angle using a universal stage. The sample was a thin section ofDevonian sandstone, from North Glen Sannox, Arran, chosen because of the well developed undulatoryextinction in many of the quartz grains. The operators were six members of the MSC Class in IndustrialMineralogy in the University of Leicester, and the first author. Al have roughly the same background inpetrography from undergraduate and postgraduate classes.

7.1.6.2. Data

Each operator made 20 measurements on the same thin section, recording the first extinctionshadow, first extinction position, last extinction position and last extinction shadow, following the methodproposed by Dolar-Mantuani (1983). In addition, the extinction positions of the same grain weremeasured by the six MSCstudents. The results are presented in Table 7.1.6.2.2.

Sample: Dalradian Grit

Operator E N s M R

AUEA 17 20 6 11 15DOUR 33 34 17 31 39

J

1936

Table 7.1.6.2.2.

finally, the true and apparent undulato~ extinction angles of nineteen grains were measuredusing a universal stage.

7.1.6.3. Discussion

The mean, standard deviation and range for each operators’ measurements, and the overallstatistics, are presented in Table 7.1.6.3,1. There are surprisingly large variations between operators.

.

Undulato~ Mnction of Quati in Granites and Sandstones 22

Apparent UE Angle Dolar-Mantuani

Oper. Mean S.D. Min. Max. Oper Mean S.D. Min. Max.

E 16.5 7.8 6.0 31.0 E 35.8 9.8 24.5 53.0N 18.7 12.6 3.0 45.0 N 28.9 15.6 9.5 60.8s 13.3 15.7 2.0 76.0 s 26.5 15.3 13.5 83.0A 7.9 6.2 3.0 30.0 A 28.6 8.3 19.5 54.5M 17.6 17.4 4.0 84.0 M 32.2 23.4 10 107.5R 9.5 4.6 5.0 20.0 R 26.0 6.6 19.0 41.5J 17.2 6.5 6.0 28.0 J 43.1 8.3 28.0 54.5

TOTAL 14.3 11.9 2.0 84.0 TOTAL 31.5 14.6 9.5 107.5

Table 7.1.6.3.1.

The overall range of the extinction angle difference measurements (2-84°)mvers much of therange of values predicted inthe theoretical study. This range would have been smaller ifmore care hadbeen taken to select only grains with a high birefringence; operators A and R give results that aresignificantly lower than those by other operators. me overall distribution has the same form as thedistribution produced by Basu. et d, (1975), except for a rather higher elongate plateau above 15°, fromuniversal stage measurements of the true undulato~ extinction angle in granites and metamorphicrocks.

The significance of the differences between operators can be assessed using Student’s t test, tocompare the means and standard deviations. The matrix of vafues for Student’s t for the apparentundulatory extinction angle is presented in Table 7.1.6.3.3. Those values which exceed the 1YO

probability level ~.e. there is less than one chance in a hundred that the difference in the means aroseby chance) are shown in bold type,

Operator A (the most experienced petrographer in the group) has three Student’s t values aboveor close to the 0.1 VOlevel; his determinations are different, and give lower extinction angles. His choicesof grain were better, in the sense of being closer to those which might be obtained using a UniversalStage. Operator R has one value above the 0,1% level and two above the 1% level. The results of R andA can be said to have come from the same population; all the other measurements can be saidstatistically to have come from the same population, but possibly a different one from operators A and R.

As allresults were measured on the same slide the two “populations” must reflect the mode ofselection of the grains to measure. ~perience obviously matters !

The Student’s t matrix comparing the mean and standard deviations of individual sets ofmeasurements, is shown in Table 7.1.6.3,4,

Operator N JE 0.62 0.~6 3.:4 0.;3 3.:1 0.29N 1.17 3.34 0.22 2.98 0.46s 1.39 0.77 1.01 1.00A 2.18 0.90 4.50M 1.86 0.09R 4.21

Significance levels for 38 degrees of freedom:5% - 2.01% - 2.60.1%- 3.6

Table 7.1.6.3,3. Student’s t Matrix for Apparent Undulatory Winction Angle


Operator N sE 1.57 2.14 2.;8 0.!9 3.:4 2.:2N 0.48 0.07 0.51 0.75 3.51s 0.53 0.89 0.13 4.16A 0.63 1.07 5.39M 1.11 1.92R 7.03

Table 7.1.6.3.4. Student’s t Matrix for Apparent Undulato~ Extinction Angle, folloting the Dolar-Mantuani method,

Operator J has four values above the 0,1% level; these are significantly different from themeasurements of the other operator in the pair being tested. R has one value above the 1% level. J itwould appear was not measuring the same thing as the other operators.

Note that operators A and R, whose apparent undulato~ extinction angle measurements weresimilar to each other and closer to the likely universal stage measurements, measured sets of anglesindistinguishable statistically from the other members of the group, except for J.

7.1.6.4. Relation BeWeen the Apparent Undulatory Mntion Angle and the Dolar-MantuaniMnction Angle.

It would appear that there is a statistid relationship between the apparent undulatory extinctionangle and the Dolar-Mantuani extinction angle, A strong linear correlation is shown (Correlationcoefficient = 0.89 ). This relation suggests that the Dolar-Mantuani measurement does not contributeany new information, especially in view of the impossibility of predicting the actual extinction angle froma measurement on a flat stage, unless great care is taken to ensure that only grains with both c-axeslying in or very close to the plane of the section are measured.

7.1.6.5. Universal Stage Measurements

Nineteen measurements of the true undulatory extinction angle have been made (by ASS),following the procedure outiined in an appendix to this report.

Measurements, both real and apparent, are compared tith the multi-operator results using a flatstage. The grains were chosen to have relatively low birefringence; the c-axes were not parallel with thethin section. The mean, standard deviation, minimum and maximum values are given in Table 7.1.6.5.2.

Mean St.Devn Minimum MaximumApparent UE Angle 8.63 4.23 2.00 18.00True UE Angle 7.05 3.22 2.00 14.00Average Inclination 43.45 10.96 23.00 61.00

No.of measurements -19

Table 7.1.6.5.2. Statistics on Universal Stage Measurements of the True Undulatory ExtinctionAngle

me differences between the universal stage measurements and the flat stage measurementsmade at the same time are not enormous, but there are large, statistically significant differences betweenthe measurements made by atl but one of the MSCclass and the universal stage measurements.

Undulato~ ~nction of Quartz in Granites and Sandstones 24

At the present time we can find no agreement in the literature of what constitutes a ‘bad’average extinction angle, from the reactivity point of view. If accurate values are required, if for examplethe difference between an average angle of 7 and 10 was significant, then the universal stage must beused. However, if the cut-off is not so small then it may be possible to measure the extinction angles ona flat stage, taking great care to measure only those grains with the maximum birefringence. It is worthnoting that most of the time spent on measuring is spent looking for suitable grains, whether themeasurements are to be on a flat stage or a universal stage. ~th a little practice, the universal stagemeasurement takes only a Iitie longer than the flat stage measurement. Hence, we recommend the useof the universal stage wherever possible; the results are not open to question,

7.1.6.6. Conclusions

1. Differences between operators can be statistically significant, even for reasonablytrained operators, Specific training, in selecting suitable gr~ns, is cleady required.

2, There is a linear relation between the apparent undulatory extinction angle, and theDolar-Mantuani measurement, As the apparent extinction positions require only twomeasurements, we can see no value in making the Dolar-Mantuani measurement.

3. The only urtsin way of measuring undulatory efiintiion angles is to use auniversal stage.

7.2. Results

In total 49 of the original samples were selected for further investigation, The selection of thesesamples was made after thin sections of the samples had been prepared and initially inspected optically,The main criteria for the rejection of certain samples was two fold, firstly on quafiz actually being presentin the sample, and secondly, on that of grain size. For the purposes of optical investigations the thinsections are required to be of a thickness of 30pm. Any sample exhibiting a grain size of 30#m or lesswas declared unsuitable for further detailed optical investigation as no grain measured could beguaranteed not to be exhibiting interference patterns from a grain lying below. This criteria meant thatfine grained sandstones, siltstones, microgranite were not measured optically.

In addition to these criteria the results obtained for the sedimentary rocks are based on themeasurement of 20 quartz grains which showed the highest degree of undulato~ extinction, andtherefore represent the mean of the maximum undulosity present.

The results from the optical measurements made are presented here in graph form. The datasets from which the graphs have been produced can be found in appendix M.O.

The acquired data was used firstly to see if the data collected for this project matched thefindings of previous investigations of the relation between true undulatory extinction angles ~~)(universal stage) and apparent undulatory extinction angles (ROTA) (flat stage). Figure 7.2.1. show thedistribution of data points from the relationship between TU~ and ROTA The TU~ values representuniversal stage measurements, and are taken as being the definitive measurement, and the ROTAvalues represent the flat stage, high birefringence, measurement. A direct mmparison between Figure7.2.1. and Figure 4,3.6., from Basu, et a/. 1975., can be made. In Figure 4.3,6. the majority of the pointsfall in the ‘True angle of unduiosity’ half of the graph. This relationship is mirrored in the data setcollected for this project, Figure 7.2.1., where, as expected, the majority of points fall in the TU~ hdf ofthe graph. This would indicate that the ROTA method of measuring undulatory extinction anglesproduces vales that are generally lower that the true undulatory extinction angles.

The distribution of points shown in Figure 7.2.1. indicates that the relationship is not welldeveloped and the possible range of ROTA values for a specific TU~ value is large, eg, for a TUWvalue of 4° the range of ROTA values is from about 1.7° to

Undulato~ &tintiion of Quati in Granites and Sandstones

00 2 4 6

figure 7.2.1. Relation beween TU= and ROTA for dl data

7G

60

50

40

o

o%&-00°

3015 20 30 35 40

25

figure 7.22 Refation be-n DOW and OeHIU for all*

Undulato~ Mndion of Qua@ in Granites and Sandstones

4[]

35

30

25

Q[)

o

0

J

figure 7.2.3. Relation beween WU and DOUR for all data

7J

60

50

40

00

300 2 4 e 8

I

figure 7.2.4. Relation hewn W-and DeHIM for ~1 data

0

00 2 4 6 6

MeanTUEP)V*S (@ees)

figure 7.2.5. Relation be~een TU~ and AU~ for dl data

10

8

6

4

2

Undtilato~ =tinction of Qum in Granites and Smdstones 27

0

0

0 5 10 ti

figure 7.26. Relation bewen AU* and ROTA for dl data

Undulatory Extinction of Quati in Granites and Sandstones 28

about 4.8°. This means that the degree of scatter is too great for a simple mathematical conversionbetween the two methods of measuring undulatory extinction.

Figure 7.2.2. shows a ve~ strong relationship between the DOUR and the DEHIM methods ofundulatory extinction measurement. This graph was produced to check tiether the modifications madeby Dolar-Mantuani on the method described by DeHills and tirval~n (1964) gave any greater accuracyin the measurement of the undulato~ extinction angle. The results show that the two methods producea linear relation. From this it would appear that there is no advantage in using the Dolar-Mantuanimethod in preference to that of DeHills and Grvaldn.

In an attempt to evaluate the different methods of measuring undulatory extinction proposed thefollowing figures show the relationships between the true undulato~ extinction angles measured on theuniversal stage and the other methods described in section 7.1.

Figure 7.2.3, shows the distribution of points for TUEA values against DOUR vafues. As theDO~R method is accepted as the method used tien evaluating the suitabihty of strained quatibearing aggregate by the Concrete Society then the possible relationship between the two methods isimportant. From the graph it is possible to see that the scatter of the points do not indicate a welldefined trend. The points appear to have a random scatter pattern. This implies that the DOmRmethod of measuring undulatory extinction does not accurately reflect the true angles of undulatoryextinction, This supports the theoretical evaluation presented in section 4.3.1.

Figure 7.2.4. shows the relationship between TUEA and DEHIM vafues. As shown earlier inthis section the DEHIUS values reflect similar characteristics to that of the DOUR values. As would beexpected there is no apparent linear relationship between the two methods. Once again the scatterpattern is that of a random distribution. The DeHills/Cowa14n method can therefore be classed as notrepresentative of the true undulatory extinction angle values.

Figure 7.2.5. shows the relationship between TUEA and AUEA (low birefringence) values. Fromthis graph all the data points fall above the line of equality (X=Y), in the AUEA half of the graph. Thisindicates that when measuring the undulatory extinction when the c-axes are inclined towards the opticaxis of the microscope then a slight deviation from the optic axis produces a greater value for theundulato~ extinction angle than the true undulatory extinction angle. This demonstrates part of thetheo~ presented in section 4.3.1. in accordance with Figure 4.3.2b. that the actual angle of inclination ofthe plane containing the two c-axes being measured is the major factor influencing the measured valueof undulato~ extinction. (It should be noted that it is not possible for the AUEA value to be less than theTU~ value.)

Figure 7.2.6. shows the relationship between the AUEA (low birefringence) values and the ROTA(high birefringence) values. The possible relationship between AUEA and ROTA was investigated to seeif the angle of inclination of the plane containing the c-axes could account for a possible linearrelationship, As can be seen from the graph, once again the points are scattered, though a broad butweak linear trend is possibly present.

7.2.1. Summary of findings

From this section it is apparent that the use of flat stage methods of measuring undulatoryetiinction do not produce values of undulatory etiinction that match the true values found bymeasurement on the universal stage. Therefore the only accurate method of measuring theundulatory etiinction of strained quati is that of the universal stage method.

Undulato~ ~tinction of Quartz in Granites and Sandstones 29

7.3. Other methods of investigation

7.3.1. Point counting

Part of the present method for evaluating the potential reactivity of strained quartz in the alkali-silica reaction is the calculation of the amount of highly strained quartz in the rock type. ~is can beobtained by using the method known as point counting, This method is widely used in the field ofpetrography. The procedure for point counting involves the use of a graduated automatic thin sectionholder mounted on a metrological microscope stage. The normal method involves the identification ofthe mineral lying at the point of the cross hairs in a 1000 point network over the thin section. Thenumber of points needed to obtain a representative reflection of the sample depends on the grain size ofthe sample and the thin section size. The 1000 points are located by the use of the graduated automaticthin section holder. This piece of equipment moves the thin section a specified distance along apredetermined track. For the purposes of this project the 1000 points were divided up into 20 tracks of50 points each,

In addition to calculating the quartz content in the thin sections the method was also used togive an indication of the mineralogy of certain selected samples. Seven specific minerals were lookedfoc

plagioclase feldspar (undifferentiated)potassium feldspars (undifferentiated)biotite micacalcitehornblendemuscovite micapyroxenes (undifferentiated)

in addition to these minerals, 3 further categories were identified;

ore minerals (opaque minerals, metalliferous)other (any other crystalline minerals)void (porosity or holes in the thin section)

The results of the point count operation are presented in appendix W.O. The point count datagives a general impression of the major mineral phases present. In addition to quartz it is possible thatother minerals may have an effect on the reaction in concrete. The feldspars may contribute to the freealkalies in the system, as they contain Na, Ca, and K ions. The possible role of such minerals isdiscussed later in the report.

7.3.2. X-ray diffraction line broadening.

X-ray diffraction (XRD) has been used in this study in an attempt to measure the disruption inthe crystal lattice, due to strain, of ce~in selected quartz bearing samples. The phenomenon of linebroadening is a result of variations in the crystallite size of the ‘building blocks’ that make up the quartzlattice, Strain disrupts the size of these ‘building blocks’ which is reflected in the XRD trace by the widthof the peaks. The greater the disruption, the wider the peak,

Five samples were chosen, representing a range of measured true undulatory extinction angles.Four of the samples were from the material mllected during this project, namely, knds End Granite(knds), Shap Granite (Shap), Strontian Granite (Stron), and a sample of vein quartz from the BodminMoor Granite ~or). The fifth sample was that of man-made hydrothermal quartz (Hydro), selected dueto it being, in theory, strain free. ~his was confirmed after optical investigation). Measurements weretaken of the line broadening and corrected for machine broadening effects by use of the standard(Philips Standard, Akansas Stone.)

Undulatory &tinction of Qua* in Granites and Sandstones 30

After. obtaining a concentration of quati grains from the samples, these were powdered andandysed using X-ray diffraction methods. (See appendk A3.O. for detailed procedure). me linebroadening measurements (peak width) were taken using the first quati peak using CUW incidentradiation. The resulting peaks are shown in figure 7.3.2.1. Measurements taken from the X-raydiffractogram were the used to calculate the crystallite size using the Scherrer equation.

1

‘18.6 18.8 19.0

Nc-S 2HT,A,

figure 7.3.2.1. Graphical representations of peak positions and peak widths for quartz from

selected samples.

Results

Table 7.3.2.2. shows the resultant crystallite sizes of the samples measured:-

Peak Position (02e) C~statlite Size (A)Std 18.91 NIAHydro 18.85 3175Shap 18.87 454Tor 18.80 212Stron 18.86 847hds 18.85 1587

Table 7.3.2.2. Calculated crystallite sizes for selected samples with v~ing degrees ofundulatory etinction.

If these resuhs are compared ~ the @ues of undufat~ ti.nction measured optidly therelationship between the -Iite size and the true @ue of undu~i can be studied. figure 7.3.23.shorn tiis relationship.

Undulato~ &tinction of Qua~ in Granites and Sandstones 31

It is assumed that the Hydro sample has no undulato~ extinction due to internal strain, and isplotted with a value of zero. Even though there are only five actually measured points on the graph it ispossible to see a possible linear relationship. The only exception in the group is that of the Shap samplewhich plots awayfrom the main trend.

6

-

i4

8_.

o

0 Ta

o Srm

o .%p o Ltis

Hytio

Figure 7.3.2.3. Relationship between crystallite size and true undulatory extinction angle.

From this prelimina~ study it would appear that this line of investigation would be worth

following up, as it may offer another method for evaluating the degree of strain in the quati grains.

8.0. QUAR~ GRAIN TWRAL CHARACWRISnCS.

al. Introduction

In addition to the optid methods of measuring the degree of stabitity of quati grains, thegeneral texture of the qu~ grains in a sample is another method in assessing the stability of the grains.In section 4.2. some of the methods for evacuating quartz grain stability have been presented. In thissection the actual stabitity of the quartz grains in the mliected samples are assessed on the basis of thetextures displayed. Features indicating stabifity or instability range from, type of grain boundties, gr~nsize, grain shape, and in addition to the actual value of the undulatory extinction angle measuredprevious~ a general destiption of the appearance of the undtito~ etinction pattern.

Terns used in this section are in part dran from standard geo~id terminology, but in placeshave been modtied to d-be a specific de~.1. A gened outiine of me terns used and their m-ingin the mntext of WISrepoti are presented here:-

.


Grain shape

Subhedral In igneous or metamorphic rocks this refersAnhedral to the degree of development of the crystal

faces of a grain. Subhedral grains have anoverall general shape relatingto thecrystallographicfaces of quartz but withsome irregularfaces.Anhedral grainsshow none of thecharacteristic crystal faces and the shapehas no regularity.

Angular For sedimentary quati grains the terminologySubangular is slightly different. A scale from verySurrounded angular through to very rounded is used toRounded describe the shape of the quartz grains.

Grain size

Grain sizes are quoted in mm for all the types of rocks examined. Forsedimenta~ rocks only standard descriptive grain size terms are usedalso, eg. fine grained.

Grain boundaries

BoundariesSuturing

StraightCurvedLobate

For the purpose of this report two terms areused to relate the actual grain to grainboundaries, When this is a quartz-othermineral bounda~ then bounda~ or boundariesis used to identify this. When the boundayis a quartz-quartz then suture or suturing isused to define this. In both cases the samedescriptive terms are used to describe thecomplexity of the boundary.

The complexity of the bounda~ refers to thestraightness or the irregularity of the grainboundary. Straight and curved are selfexplanatory. Lobate indicates that thebounda~ is an interdigitation of the twograins meeting at the boundary. Open andtight refer to the complexity of thisinterdigitation, open being less complex,less intricate. Tight is more complex andmore intricate.

The extinction pattern displayed within thegrains. Banded extinction refers torelatively regular areas of extinction,stretching across the grain to give a moreor less even width band showing the samedegree of extinction. Zonal extinctionrefers to more irregular shaped areas ofextinction and even extinction refers tograins that do not show undulatory extinction.

—

Undulato~ Ntnction of Quartz in Granites and Sandstones 33

Grain distribution. .

Single This section refers to the distribution ofSmall groups the quartz grains within the sample. TheseMultiple are descriptive terms which define whether

the grains are found singly, in small groupsor as large areas made up of many grains.

8.2. Qua* grain tetiures of collected samples.

Figure 8.2.1. shows three examples of rock textures that indicate that the rock is unstable. Theexamples are those stated in Figure 8.3,1.

Ml the figures are produced using crossed polars and at the same scale so direct comparisonscan be made,

8.3. Results

IThe textures found in quati show a wide range of variation in the samples examined. Using a

flow diagram first presented by Young, (1976), and with some modification, it is possible to infer thestability or instability of quartz grains by their textures. Figure 8.3.1. shows the modified flow diagram,

IYoung presented a system for predicting the stability of detrital polycrystalline quartz grains.

The figure presented here is essentially the same but with modification to the terminology used. Usingthe modified version it is possible to infer the stability of the quartz grains for the igneous andmetamorphic samples. In this study sedimentary samples are not treated using this method due to theuncedain origin of the quartz in the sediment.

ITable 8,3.2, shows the results of applying this textural classification to the samples collected.

The figure also includes the relevant true undulato~ extinction angle ~U~) for the sample, wherepossible, For the samples which are classified as being unstable, the failure point is specified. This isthe point on the flow diagram, Figure 8.3,1., where the sample is defined as being unstable due to thatfeature, either A. Crystal-crystal contact, B. ~tinction, or C. Grain size.

IUsing the textural characteristics results and comparing them with the TU~ values for the

samples, Table 8,3.2., it is possible to compare the relationship between the optical measurements andthe physical appearance of the quartz grains. Histograms allow the relationship to be shown. Figures8.3.3.a. and b. are the resultant histograms for the samples that are classed as stable and those that areclassed as unstable.

Figure 8.3.3a. shows the distribution of quartz grains that have a stable texture, There wouldappear to be a normal distribution about the 4° value for TU~, with a maximum value of 5°. Figure8.3.3b. shows the distribution for the unstable quartz grains. The distribution for these are less welldefined, tith a large range of observed TU~ values. From this evidence, it is only possible to predictthat grains that have a TUW value greater than 5° have textural characteristics that show then to beunstable. TU~ values less than 5° though can not be inferred to be stable, as from this evidence thereare significant numbers of samples with TU~ values less then 5° that have unstable texturalcharacteristics.

I me significance of the textural characteristics to potential alkali-silica reactivity is based on theassumptions that

1. Grain boundaries (of dl ~pes) act a potential routes for the movement of alkaline fluids,

Undulato~ ~nction of Quartz in Granites and Sandstones 34

a. SW~DN/1 Vein quartz form the Bodmin Moor Granite.

This sample shows anhedral quartz grains which show tightly serrated suturing. The grain sizeranges from 0.5mm to 1.5mm, also showing slight elongation of the grains. Zonal extinctiondominates this sample,

b. SW~H/Ql Quartzite pebble from the Budleigh Saltefion Pebble Beds.

This example shows anhedral quartz grains with open Iobate suturing, Both zonai and bandedextinction patterns are present. The grains size ranges from 0.2mm to 3.Omm.

c. SW/DUG/l Bodmin Moor Granite

This sample exhibits anhedral quartz grains of two dominant sizes, both having open Iobatesuturing. The smaller grains, of 0.2mm in size, show slight elongation. The larger grains of0.5mm to 1.Omm show zonal extinction. me quartz is found as distinct multiple grainconcentrations throughout the sample.

figure 8.2.1. Photographs of typical examples collected, viewed under crossedpolarized light. a. shows an example of quartz that would be classified unstable due toits crystal-c~stal contacts. b. shows an example in which the quadz would be classedunstable due to the extinction pattern. c. shows an example of quati grains that wouldbe classed unstable due to the variation in grain size. The classification of the grains arebased on the information presented in figure 8,3,1.

Unstablo “.. - .Q”. -

t ?.

Unstablo

tight Iobata boundaries cmplax zonal and banded extinction/

largo ●longato grain. with

or ●uturi~

t “/

S=011 nOnUndUIOOO g;ein=

s~mple xona 1 andA =YST~L ~YS7AL~NYA~

\ /“~xNc’xoN’~

‘\

banded ●xtimtion k 1

o-n lobato straight mrved

boundarloc or .uturing~~ ‘ ‘“’\-”’t

●v.n axtlnatlon Unlcom quidlmansi-1 graino with

\\

120 dagrao trlplo junotlons

.\.

StabloStablo Stablo

.

\

Figure 8,3.1. Flow diagram for predicting stability or instability of quati grains by textures. aa

(Modified from Young, (1976))

Undulato~ Mndion of Quati in Granites and Sandstones

SampleCode

SW.England

BOGCLGCYGGGGKDGCHGDLGHGGTDGTDVMEGMGTHGSIGLEGHDGHHQQ

Wales

CGGG-GN-MGBHQD

Lake Dstrid

SPGTKMG

Stable or UnstableSu

susussuuuusssssuu

uusu

ss

BBGBDGBNGCFGCTGCWGKMGTFGGSGBRQDOADLWLALWBNMNBUMNCGSHBLPY

suuussusssuuuuss

--

BB

A,C----

A,B------

cB,CAB----

Table 8.3.2. Stability results inferred from tetiurd charatieristi~.

Failure PointAB, C

--

A--

A----

cAAA----------

A,CB

BB,C--

B

----

37

TUW

4443--

43456344--

454

68--

5

4--

--

35444445--

675634

Undulatory &tintion of Quartz in Granites and Sandstones 38

2. Grain boundaries are potential sites for reaction. me more complex the bound~ thegreater the surface area available for the reaction to take place upon.

3. Complex extinction may act in a similar fashion as grain bounties, aa they are anopti- representation of the dislocation density tithin a grain. Dislocations are the equivalent ofmicro boundaries, therefore the same reasoning can be used as in point 2., above.

4. Grain size once again affects the sutiace area available for the reaction to take place.me smaller the grains the greater surface area available, though against this, smaller grainstend to be more stable.

5. In addition to the previous points, grains that appear unstable and do not show anyrecove~ features, as described in section 5.0., in theory have a greater potential stored energythan stable grains. ~is stored energy may indicate a higher potent~ for reactivity in the dkali-silica reaction.

11.4 5 6 7 8

a

6

L

4

2

,.

2 3

—

TUEA

b

r

5 6 7 8

the textural

2 3 4

TUEA

figure 8.3.3a,b. Histograms showing the distribution of ~~ values relating toclassification of stability of the quartz grains. a) stable textures, b) unstable textures.

Undulatory &tinction of Qua@ in Granites and Sandstones 39

9.0.

9.1.

DISCUSSION.

Conclusions of the Sumey

At present the suitability of an aggregate for concrete, with relation to the possibility of ASR as aresult of strained qu~, is governed by the criteria set out in the Concrete Society Technical Report No30 (1987); the guide-lines presented in the report, relating to strained quam are to be found inAppendix 3. The repofi states that if the mean degree of undulatory extinction, measured using ametrological microscope, measured from at least 20 grains in the thin section is greater that 25° then thequati should be classified as highly strained. In addition, if the rock contains more than 30% high~strained qu- then the rock should be classified as being poten~ly reactive.

Using these guide-lines and the data mllected during this project it is possible to identifysamples that would fdl into the ‘potentially reactive’ category. figure 9.0.1. shows the results of aselected number of the samples measured. The points which tit in the ‘Potentid~ Reactive Zone’ arethose that would fail the criteria in the Concrete Society report.

<m

75

50

v

o

00 0

0 I

I

o0

00 000 &

figure 9.0.1. Graphical representation of the guide-lines for the identification of potentiallyreactive quati bearing aggregate.

Table 9.0.2. shows the data and the results of the implication of the guide-lines on the samplesfor this project. Of the 49 samples, 18 would be ~ as being potentially alkali-silica reactive. Thesuitability of these aggregates for concrete would therefore be in question.

Undulato~ Extindion of Quati in Granites and Sandstones

Sample Code

BOGMEGCLGTHGLEGGGGDLGTDGHGGCGGG-GSPGCTGBNGGSGTFGKMGBHGCFGRMGHDGUGCYGCHGBDGCWGKMG2SHGBUMNBNMNADLWtiLWCGSHBLPYHHQQTDVBHQDDDGWBHSSBTSSBKSSCsssHGSSPGSSSsssScssPLSSLBssWFSS

Rock Type

GRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITEGRANITE

.

QTZ.FELD.GRANQTZ.FELD.GMNGNEISSGNEISSSCHISTPORPHYRITEQUARTZITEVEIN QUARTZQUARTZ-DIORITEGREYWACKESANDSTONESANDSTONESANDSTONESANDSTONESANDSTONESANDSTONESANDSTONESANDSTONESANDSTONESANDSTONESANDSTONE

MeanUEAngledegrees30242823242626252525321926282828272827262321202424242024293029372122233228222726213023252121222228

Qti Content%

28402129223934383437403934353039503318354030253535255030328145303510999535409599999999999999999995

40

PotentialReadion

NONONONONOYESYESNONONO

YESNOYESYESNOYESYESYESNOYESNONONONONONONONOYESYESYESNONONONOYESYESNOYESYESNOYESNONONONONONOYES

Table 9.O.2.. fieinfomation presented isinaurdance tothe~uidelines setoutin ~~ndix3 of theConcrete Smiety Techni~ Report N“30 (1987). In light ofth~ information presented in this projectreport the potential for reaction should not be regarded as definitive, (QTZ.FELD GRAN = quatifeldspar granulite, MOINE)

Undulato~ ~tinction of Quati in Granites and Sandstones 41

9.2 Suitability of the Concrete Socie~ guide-lines

The suitability of these guide-lines though is dso in question. *is shon in this report, the method fortie measurement of the degree of undulosity in the quati does not reflect the true degree of undulosity.% no accurate direct comparison between

40

35

30

25

~o

figure 9.0.3. Graph showing the variation possible in the universal stage measurements whencompared with values of undulatory extinction when measured using the method prescribed bythe Concrete Society guide-lines.

the Dolar-Mantuani method measurements and the universal stage measurements can be made, thereis no way of stating the equivalent TU- value for the 25° Doiar-Mantuani method measurement. Thiscan be clea~ illustrated by inspection of the TUWDOUR graph shown in figure 9.0.3.

It is possible to see that values of DOUR that lie tithin *I 0 of the 25° line have a ve~ tiderange of corresponding TU~ values. Four points have been highlighted in figure 9.0.3. to show this.Points (a) and (b) lie above the 25° tine, therefore the samples from which they came are classified ascontaining highly strained quati men their equivalent TUW values is measured they show a markedcontrast in values, (a) having a value of about 3.5° where as (b) has a value of about 5°. The same isevident if DOUR values less than 25° are taken, (c) and (d). The variations possible mean that onty theTU- values have any accurate reproducibility.

me textural variation of the quati gm.ns obsewed in the samples give an extra line of evidenceas to the stabi~i of the samples, and therefore a possibledue to their potential for reaction in MR. meresults of investigation, presented in section 8.3., onl go as far as cotirming the instability of qu-

$gtins in a sample * a ~- ~ue of greater that of undulosity.

Undulatory Extinction of Qua* in Granites and Sandstones 42

me results of the preliminary test of using X-ray dtiraction tine broadening techniques toindicate strain.in quati grains, again show some useful findings, though tihout extensive testing thismethod could not be used solely as an indicator of potential reactivii.

During the course of this project it has come to tight that ream i~olving stifled qu~ andphysical testing in prism tests is in progress. me results after 8 months show that excessive expansionhas occurred in prisms made from Coedana Granite (Sibbick, Ph.D.researchinprogress, 1991, pem.comm.). me results are presented here in Figure 9.0.4. with the kind permission of the author.

alo

11am -

aw -

am.

O.m-

Akdi content is quoted as

N~O equivalent (as K@rn:

+ 4.5

- .s.04 53- 6.0+ 7.0

Figure 9.0.4. Expansion data for the ~edana granite for different cement alkali contents. Inefifine aggregate Cheddar Iimestome, reactive coarse aggregate (5-20mm) Coedana granite.AtkAi content is quoted as Na20 equivalent (as k@m3). Recommended expansion level after 6months is 4,05% according to the Building Research Establishment criteria for reacttilty. (AfterSibbick, 1991, pen. comrn.)

From Figure 9.0.4. it is possible to show that 2 of the 5 samples shows excessive expansion dueto reactivity, these samples having the highest alkati content in the cement. me inefi fine aggregate hasno histo~ of any previous reaction, which impties that the expansion is due to the reactivity of the-dana granite. [f we compare these results with the degree of strain that the same sample showsopticat&, tien we see that the reacdvity is matched by a 32° undulat~ extinction measurement.~is wuld have classified the sample as being potentiat~ reactive, and therefore unsuitable, H the@nmte Society guid-fines were employed. ~ corresponding tie undu~o~ extincdon angle forthe @dana granite is that of 8° the highest recorded in the proj- From the work presented insection 8.2 the observed ,texture of&e ~ granite has a number of unstable features. me rock in

Undulatory Wtnction of Quartz in Granites and Sandstones 43

general is highly altered and the quartz shows signs of a potential for reactivity in the alkali-silicareaction. Al the indicators show similar results, that the Coedana granite contains reactive quartz.

10.0. AUTHORS’ RECOMMENDATIONS.

It is suggested that the reactivity of quartz with undulato~ extinction is a function of a number offeatures of the quartz grains, namely:

1. Undulatory Mnction Angle2. Texture3. Grain size

Al three features are interconnected and are the result of the processes of rock formation anddeformation. They depict a visual representation of the overall stability of the quartz. Instability leads tohigh dislocation densities and microcracking, a process ultimately leading to stabilisation (recove~), thatprovides a greater surface area, with a higher degree of surface energy, for potential reactivity.

As a result of the findings of this project and recent published works it is suggested that thefollowing criteria be assessed before a quaflz-bearirrg igneous or metamorphic aggregate berecommended for use in concrete:

1. The undulato~ extinction angle vatue for quati be measured using the universal stagemethod described in appendix Al .0. Only aggregates that have a undulatory extinction anglevalue of =5.00° should be classified as being highly undulatory. Any value c5.00° but showingundulosity should be classified as undulatory, and any sample showing no undulosi~ at allshould be classified as being non-undulatory, A value of =5,00°, alone, does not imply that thequartz is alkali-silica reactive,

2. Using the criteria set out in section 8,3., quati grains should be assessed, by theirtexture, for instability. Any rock type exhibiting unstable quartz textures should be classified asunstable. Any rock type exhibiting stable textures should be classified as stable.

3. As more and more cases of ASR resulting from reaction with undulato~ quartz arefound, previous case histories of the aggregate should be examined (Mere possible), either inexisting structures or in laboratory test prisms, to ascertain any past reactivity. Any rock typeshowing past reactivity should be investigated with a view to limiting or excluding its use instructures where there are potentially reactive environments,

A rock type which is classified as highly undulatory, has unstable textures and has, or issuspected to have been, involved in deleterious reactivity should not be used for concrete structureswithout further detailed, and extensive testing for its suitability. This testing would be to ascertain itspessimum levels and type of reaction. This type of rock should be avoided and classed as potentiallyalkali-silica reactive if these tests can not be undertaken. If afler testing the rock is found to be suitablethen it should be used with caution.

Any rock type that is highly undulatory and shows unstable textures but has no previous histo~of reactivity should be used Mth great caution and the find structure monitored, If possible prism testingshould be undedaken and further assessment made following the results findings.

Rock types which only exhibit one of the deleterious categories above, should be used withcaution. Again monitoring should be undertaken of the soundness of the structure throughout its lifetime.

UndulatoryExtinction of Quartz in Granites and Sandstones 44

Any rock type that does not exhibit any of the above deleterious features should be used until

any new information regarding its suitability, from testing or in situ obsewations, comes to light. At thispoint the rock type should be investigated further,

Rocks of a sedimenta~ source can not be classified using this scheme as the stability of thequartz is a product of the nature of the source or sources of the quartz. As the origin of quartz grains insandstones can be very varied and deleterious features could be found across the range of possibilities,such samples can only be assessed from testing or in situ obsewations,

ACKNOWLEDGEMENTS.

This project was carried out in the Department of Geology at the University of Leicester. Theauthors would like to thank the academic and technical staff of the Geology Department for the helpgiven during this project. We would also like to thank Mr R.G. Sibbick for very helpful discussion andadvice relating to the testing of concrete prisms for reactivity. Finally thanks go to Dr. Graham West forhelpful comments and guidance throughout the project.

REFERENCES.

Andresen, M.J. 1961 Geology and petrology of the Tfioli Sandstone in the Illinois Basin. ///inois StateGeological Survey Circular. 316 31 p

Andresen, K.T. and Thaulow, N, 1989 The application of undulatory extinction angles (UEA) as anindicator of alkali-silica reactivity of concrete aggregates. 8th /ntemationa/aggregate reaction. Kyoto. p 489-494

Bailey, S.W., Bell, R.A., Peng, C.J. 1958 Plastic deformation of quartz inGeological Society of America. 69 p 1443-1466

&nference on a/kdi-

nature. Bulletin of the

Basu, A., Young, S.W., Suttner, LJ,, James, W.C., and Mack, G.H, 1975 Re-evaluation of the undulatoryextinction and polycrystallinity in detrital quartz for provenance interpretation. Journa/ of Sediment-Petrology. 45 p 873-882

Blatt, H., and Christie, J.M. 1963 U,E. in quartz of igneous and metamorphic rocks and its significancein provenance studies in sedimentary rocks, Joumd of Sedimentary Petro/ogy. 33 p 559-579

Blatt, H. 1967 Original characteristics of elastic quartz grains. Journal of Sedimentary Petro/ogy. 37 p401-424

Brown, LS. 1955 Some observations on the mechanics of alkali-aggregate reactions. ASTM Bu//. 205p 40-56

Buck, A.D, 1986 Petrographic criteria for remgnition of alkali-reactive strained quati. 7th /nternationa/&nferenca. &ncrete alkali-aggregate reactions. Ottawa, p 419-423

Conolly, J.R. 1965 The occurence of polyc~stallinity and undulato~ extinction in quartz in sandstones.Journal of Sedimentary Petro/ogy. 35 p 116-135

Concrete Society. 1987 Akali-silica reaction: minimizing the risk of damage to concrete, ConcreteSociety technical re~d. No 30 34p

,


Dehills, S.M., and Corvaldn, J. 1964 Undulato~ extinction in quartz grains of some Chilean graniticrocks of different ages, Geological Society of America Bulletin. 75 p 363-366

Dolar-Mantuani, L.M.M. 1981 Undulato~ extinction in quatiz used for identifying potential alkali-reactive rocks, 5th International Conference on Ak-Agg Reaction. Cape Town. S25~6

Dolar-Mantuani, L.M, M, 1983 Method of Undulatory extinction (UE) angles. Handbook of Concreteaggregates. a petrographic& technologi~ evaluation. 345pEmmons, R.C. 1943 The universal stage. Gee/. Sot. Am. Mere. 8 p 1-205

Folk, R.L. 1961 Petro/ogy of Sediment~ Rocks. Hemphills Bookstore, Austin, 154 p

Gay, P. 1982 An introduction to crystal optics. Longman.

Gilbert, C.M. 1954 in Williams, Howell, Turner, F.J. and Gilbert, C.M. Petrography: San Francisco. W.H.Freeman and Co. p 406

Grattan-Bellew, P.E, 1986 Is high undulatory extinction in quartz indicative of alkati-expansivitygranite aggregates? 7th International Conference. Concrete alk~i-aggregate reactions. Ottawa.434-439

Grout, F.F. 1932 Petrography and Petro/ogy. McGraw-Hill Book Company Inc. New York.

Henry, N.F.M., Lipson, H., Wooster, W.A. 1961 The interpretation of X-ray dfiraction photographs,ed. Macmillan and Co Ltd.

Hobbs, B.E. 1968 Recrystallization of single c~stals of quartz. Tectonophysics. 6 p 353-401

Hobbs, D.W. 1988 Nka/i-si/ica reaction in concrete. Thomas Telford, London, 198p

ofPP

2nd

Hubert, J.F. 1960 Petrology of the Fountain and Lyons Formation, Front Range, Colorado, Quafier/yJournal of the Colorado School of Mines. 55 242p

Klug, H,P. and Nexander, L.E. 1974 X-ray Dfiction Procedures 2nd ad. Wiley and Sons, New York,

Krynine, P,D. 1940 Petrology and Genesis of the Third Bradford Sand. Bu//etin Pennsylvania StateCollege. 134 p

Krynine, P.D. 1946 Microscopic morphology of quartz types, Proc 2nd Pan-American Congress.Mining Engineering and Geology. 2nd Comm 3 p 35-49

Mielenz, R.C. 1954 Petrographic examination of concrete aggregtes. Proc. Am. Sot. Test. and Mat. 54pl188-1218

Mielenz, R.C. 1958 Petrographic examination of concrete aggregate to determine potential alkalireactivity. Highways Research Board Report. 18-C p 29-35

Mullick, A.K, et al. 1985 Identification of reactive mncrete aggregates containing strained quartz bySEM and IR. Proc 7th /nternation4 Conference on Cement Microscopy, p 316-332

Mullick, A.K, et al, 1986 Evaluation of quartzite and granite aggregates mntaining strained quartz. 7thInternational Conference. Concrete alkali-aggregate reactions.- Ottawa. p 428-433

Nakashiro, F.M. ,1965 Undulatory Range and C~tal Size of Quartz. The Canadian Mineralogist 8640-643

Paulitsch, P. and Arnbs, H. 1963 Undulation in quamgerollen. Tschermaks Miners/og. u. Petrog, Mit.p 579-590

P

8


Rae, LH, and Sinha, S,K. 1989 Textural and microstructural features of alkali reactive granitic rocks.8th International &nference on hkali-Aggregate Reaction. Kyoto, p 495-499

Rosenbusch, H. 1893 Microscopical physiography of the rock making minerals. Transl, J.P, Iddings.John Wiley, New York. 367 p

Staunton, D.E. 1940 The expansion of concrete through reaction between cement and aggrgate. Proc.Am. Sot. Civ. Engrs. 66 p 1781-1811

Tuttle, O,F. 1952 Origin of the contrasting mineralogy of extrusive and plutonic sialic rocks. Journa/ ofGeo/ogy. 60 p 107-124

Van Hise, C,R. 1890 me Pre-Cambrian rocks of the Black Hills. Bu//etin Geo/ogica/ Society of America.1 p 203-244

Vivian, H.E. 1951 Studies in cement aggregate reaction. XVI; The effect of hydroxyl ions on thereaction of opal. Aust. J. App/. Sci. 2 p 108-113

White, S. 1973 The dislocation structures responsible for the optical effects in some naturally deformedquartzites. Journal of Material Science. 8 p 490-499

Young, S.W. 1976 Petrographic textures of detfital polycrystalline quartz as an aid to interpretingcrystalline source rocks. Journal of Sedimentary Petro/ogy. 46 p 595-603

Undulatory Mnction of Quartz in Granites and Sandstones 47

Appendix AI.O.

Undulatory ~n~ion Measurement Pro~dures

NO=: Where possible the mean value of undulatory extintion for a sample should be obtained from atleast 20 measurements of different qua~ grains. This appties to afl the methods described.

Al.1. Universal Stage Method

The procedure is explained in full for the use of the unived stage assemb~ for the measurement of c-taxes orientation values.

Assembly Procedure for the Universal Stage.

The 5-axis universal stage is attached to the flat stage of a normal petrologic microscope. The5-axis aflows the central portion of the optid sphere to be orientated in almost any position (limited toinclinations of the thin section up to =600). figure Al. 1.1(a) shows a plan view of the 5-axes, theirrotation and notation.

figure Al. 1.1. ~sential features of the universal stage. (After Gay (1982), page 174, figure 12.2)(a) The axes of a 5-axis stage(b) The optical assembly

The optid arrangement of the univer- stage, (figure Al.1.1 (b).), enables the thin setion to bemanipulated into a position so that the c-axis of the portion of the grain being investigated is in line withthe optic axis of the microscope. To allow for the variation in the orientation of the grain, the thin sedionis mounted between two hemispheres. This optid arrangement atlows light to pass through the thin-on without being refraded, as would happen if the hemispheres were not used. The hemispheresand the thin sedion are assembled so that the grain under investigation is l~ted at the mntre of thesphere. Most universal stage opti~ dso have a gfassplate onto whiti the thin setion is pl~, theseare then santiched between the two hemispheres to m+e up the optid sphere. To attain a low levelof refraction at the glass boundaries a mdtum of tie same refractive index as the glass hemispheresmust be used. fir his proj~ hemispheres of refra~e index 1.= were used, * the g~fassanta~ made using dove oil, re~ index 1.= To enable tident woting distan~ whtiwould allow the maximum rotation of the universaf stage ~es, long woting d-m o~etives were


used, This allowed magnifications upto 60 times, which was more than adequate for the grain sizesmeasured.

Undulatory Extinction Angle Measurement using the Universal Stage Microscope (5-axis)

1.

2.

3.

4.

5.

6.

7.

8.

Ensure all axes are set to zero,

Locate quati grain for measurements,a.b.c.

d.

..Select first a ea of extinction,

{Rotate on A to extinction,Check that the area stays in extinctionMen A4 is rotated.If not then rotate A1 until the area is inthe other extinction position, and repeatstep c,

Rotate 45 degrees clockwise on A5.a. Rotate on A4 until extinction is reached.b. Record the amount of rotation on A4

(t away from operator)(- towards the operator)

Record the position of A3 (this should be at zero for the first area of extinction).

Reset A4 and A5 back to zero degrees.

Rotate on A3 to set the next area into extinction.a. Record the amo nt of rotation of A3 from the

3last position of A(C clocktise rotation)(A anti-clockwise rotation)

b. Check area stays in extinction when A4 isrotated.

c. If not then rotate A3 to the other extinctionposition, and repeat b.

Rotate 45 degrees clockwise on A5a. Rotate A4 until extinction is reached,b. Record the amount of rotation on A4

(+ away from operator)(- towards the operator)

Set A4 and As back to zero and repeat procedure from step 6.

1982),Figure Al.1 .2. reproduces the procedure described above in a diagrammatical form. (After Gaypage 176, Figure 12.3)

The procedure for the plotting and calculation of the angular difference between the two c-axesis explained as follows:

1. p[Ot A4 of the first area of extinction on the No~h-South Great Circle, t is above tieEast-West horizontal, - is below. Zero being at the centre of the stereonet increasing to 90° atthe two poles.

2. Rotate the current N-S line by the angular difference in the A3 remr ings. This will%either be clockwise or anti-clockwise depending on the direction of rotation of A of the Stage.

Undulato~ =tinction of Quati in Granites and Sandstones 49

me angle of rotation of the current N-S line can be measured using the degree graduationsaround the circumference of the stereone~

t\,(c) (d) (e)

figure Al.1 .2. Stereographic representation of the procedure for the measurement of theinclination and angular variation between two c-aes.(Reproduced from Gay (1982))

3. A new N-S line will now connect the two poles. Plot, as before, A4 for the second area,,of extinction on the new N-S line, as in 1.

4. Rotate the two points until they both lie on the same Great ~rcle line. me distancebetween the two points can now be read off using the degree gradations of the best fit GreatCircle. ~is angular difference is the corresponding angular difference betieerin the quartz grain.

me procedure explained previously is presented diagrammatically in figure Al.

Al .2. AU- Method

the two c-axes

.3.

me procedure for measuring this rotation is obtained as part of the Universal Stage Methoddescribed previously. me measurement is of the rotation needed to move between the No maximumextinction points, the rotation of A3 described in section 6. of the Universal Stage Method. As describedin the universal stage procedure qua~ grains are selected that show low birefringence, ie. the c-axis isalmost parallel to the optic axis of the microscope. me rotation measured is the product of the trueangular discrepancy between the two zones in the grain and the degree of inclination away from theoptic ais.

Al .3. DeHiii4Corva14n Method

~is method of measurement is the M of the ftat stage procedures. ~is method was firstd-bed in DeHiik and tiwaltis paper in 1984. Untikethe AU- method where the actual points ofmtimum etinction denote the start and finish of the etinction sweep, this method uses the first andM appearance of the extinction shadow in.the grain to denote the ea-nction sweep. In addition thegrains selected for measuring

N N

s

a) Plot the A4 value for the first aree of b) Rotate the firet N-S line, la Aal, theextlnotion on the N-S line (Great Circle). .The angular difference between A~l and A~2 in thleaquatorlal point rapregente the zero deer”eee. cage 10 detireeg anticlockwige, either clockwiseand each pole 90 deareea. In thie casa A41 in or anticlockwige dependina on the direction of+50 degrees hence plotting In the N-hamigphere. rotation of Aa on the universal ataaa. Plot

tha gecond araa of a:{tlnctlon valug, A42, onthe new N-S lina. In this caae A42 la +60daaraeg, aa ghown above.

c) To obtain the trua anaular differencebetween the two valuea for A4, the two Pointgnaed to ba rotated until thee botl~ lie on thasame areat Circle, In this caae 34 dearees W.Tl~e ansular dlffarance can now be read offdirectly ualnd the dearae graduatlona on theGreat Circle, in thie ceae the actual angulardlfferenca between the thg two c-a;<es (TUEA) la14 daareea.

Figure A1.1.3. Procedure for plotting unlvereal atege data ona etereonet.

.,

Undulatory ~nction of Qu~ in Granites and Sandstones 51

were those grains that showed the highest birefringence, ie. c-axes perpendicular to the optic axis ofthe microscope. The procedure, as described by DeHills and Corvd4n, is as follows:

Al .4.

1. The crystal was set in the position of highest birefringence;

2. The microscope stage was rotated until the first clear evidence of undulatory extinctionappeared; the reading on the stage was remrded;

3. The stage was rotated until it passed through complete extinction and further, until theundulato~ extinction band disappeared (the extinction bands should be barely visible); thereading on the stage was recorded;

4. The angle measured was calculated. (3,-2,)

Dolar-Mantuani Method

This procedure is best described in Dolar-Mantuani (1983). me method is a modification of theDeHills and CorvalAn method described above. The modifications to the previous method are that themeasurement takes into account the positions of maximum extinction as well as the first and lastoccurrences of the extinction. The method described is take for the reference quoted:

1, Select a grain showing high birefringence;

2. Rotate the flat stage until the grain shows the first signs of extinction; record the positionof the stage, this is recorded as C;

3. Rotate the stage until the grain reaches the first maximum extinction point record theposition of the stage, this is recorded as A

4. Rotate the stage until the grain reaches the last maximum extinction point; record theposition of the stage, this is recorded as B;

5. finally rotate the stage until the just before the last part of the extinction shadow leavesthe grain and record this stage position; this is recorded as D.

To obtain the extinction angle calculate the angular rotation between A. and D. and then theangular rotation between C. and B. The undulatory extinction angle for that grain is taken asbeing the mean of the two values.

me method described has been developed for the identification of potentially reactiveaggregates (in the ASR reaction) and includes the condition that in order to designate a samplewith a degree of undulatory extinction then at least 20 gr~ns should be measured and the meantaken.

Al .5. ROTA Method

This method of measuring the undulato~ extinction of quati grains is similar to the AU=method. As part of the Dolar-Mantuani method the position of the two maximum extinction points arerecorded, as A. and B. The angular rotational difference between these two figures is the ROTA angle.This method differs from the AU~ method in the fact that these measurements are made on grainsshowing high birefringence, and the AU~ method uses grains showing low birefringence.

Undulato~ ~tindion of Qu@ in Granites and ~dstones

Sample ~de

QtzP1KfBt&&khOreOtherVoid

Total

&artz

AGby

CGG–

371162369--24

--40

----

3444

1000

3?

603

63

162

20

64

183

20

254

32

E

467

583320

41

:20

GG

40581314--112157

----66

--

1000

40

603

10417420

8316320

324422720

577744920

:10320

SPG ~G

391261 ;:214 336119 431 11

---- 2:-- --

--: 8

-- 2

1000 1000

39 . 34

394 406

BNG

E348

1--

19------

319

1000

35

397

GSG

30126434881

--------

511

1000

30

435

Ufiversal Stwe Mea~ernents

2 5 8 72 4 2

: 9 15 101 2 3 3

14 20 15 14

4 4 5 56 2 2 1

23 7 8 81 2 3 3

14 20 15 14

Flat St-e Meawements

19 26 28 28-3 2

2; 2: 34 3314 22 24 2414 20 20 20

35 47 51 527 3 454 51 59 5:26 42 4414 20 ;: 20

2 4 62 :

i : 9 81 2 214 20 2: 20

TFG

388276248

62----

12----

146

1000

39

411

5293

20

4162

20

282

332520

533624720

31

-4

2:

53

KMG

500100388

10----

1--

1--

36

1000

50

411

5272

15

4162

15

272

322220

495

593920

3152

20

Undulato~ Mndion of Quati in Granites and Sandstones

CFG

183237346110--110334413

1000

18

406

6210316

404316

273312320

49

5;

:

4

;220

MG

34612050113

--81

--7

:

1000

35

414

6212220

517320

262302220

4a

5:4020

52a

2:

Bm

323117479--

3--77

----

2;

1000

32

1100

OPTIW DATA

Point@mt Data

Bm

ao62577a

----------842

1000

al

1100

45419a6272

--139------752

1000

45

2600

----------------------

--

30

2600

Universal Stage Measmements

a 6 93 2 215 10 13

4 52: 20 20

6 5 62 2 211 8 123 3 420 20 20

Flat Stage Meauements

293362320

302342720

55461

:

516

2:

53

A3720

6315320

8313214

7314214

37

5:3120

64a84

E

9315

2:~

WFSS

----------------------

--

95

1100

8213320

51

:20

2a3341820

506

80

E

418

2;

----------------------

--

40

29a

6211220

5

i320

23

$la20

43

5:

;

3

:

2:

54

WG

----------------------

--

30

29a

43

152

20

4172

20

212

24la14

394

453414

2141

14

Undulato~ ~notion of Ou@ in Granites and ~dstones

Sample@de

QtzP1KfBt

%MshOreOtherVoid

Total

martz

AGtiy

DEHIWeanDEHIWsdDEHI-DEHI~tiDEHIWn

HG

----------------------

--

25

298

416211

416211

203261510

36

5:2810

2~.

4110

WG

----------------------

--

35

298

528214

416214

242271920

4344a

;

41

:20

BDG

----------------------

--

35

--

OPTIC& DATA

Pointb~t Data

WG

----------------------

--

25

413

----------------------

--

35

600

BW

----------------------

--

10

--

Dm

----------------------

--

40

408

Ufiversal Stwe Measurements

31

;12

314112

5

1:216

316115

315215

Flat St-e Meauements

24 24 212 3 3

26 31 2a22 2114 20 i:

45 46 395 5

5: 57 5240 3114 z 20

3 3 21 1 14 6 41 1

14 20 2:

41

7215

4

:215

;123

20

.,.

BHSS

----------------------

--

95

235

73la320

528220

22 - 221 2

%.%20 20

42 403 44a 46

z ;

2 31

; 41

2: 20

4a

5;3a20

:102

20

Undulato~ ~notion of Qu~ in Granites and Sandstones

:

Sampletide

QtzP1KfBtk*MSkOreOtherVoid

Total

titz

AGEmy

AtieanAUWdAUEAm~AUEAminAUEAn

TUEAmenWEAsdTuEAmuTuEAmin

DOmeanDOWsdm~m-hmm

DEHI~eanDEHI~sdDEHI~DEHI=tiDEHIUn

~Amean~AsdmzmAmtimAn

BHQD

----------------------

--

35

643

7313320

529

2:

283332220

51559

:

529220

TDv

----------------------

--

95

298

9316

2;

629320

324392620

596694820

529320

BTSS

----------------------

--

qq

308

OPTIW DATA

Pointbut Data

BKSS

----------------------

--

-

208

Gss

----------------------

--

qq

308

HGSS

----------------------

--

~

308

Universal StMe Memements

5 4 6 63 1 3 212 7 11 112 3 3 320 15 20 15

5 4 43 :10 t ; 82 2 2 220 15 20 15

FlatStxe Meawements

26 21 30 235 3 4 236 29 41 2818 16 24 2020 20 20 20

46 38 53 427 6

5: 67 5:: 43 3820 E 20 20

5 43 ; ;

$ M 8 82 1 2 120 20 20 20

PGss

----------------------

--

v

303

6314220

;8120

253322020

455553620

:8220

Ssss

----------------------

--

%

386

5292

20

4162

20

21

2:1620

395

53

z

2141

20

56

Scss

----------------------

--

w

308

52

11

220

426120

213291720

37552

::

4

:220

. .

..

1111111111 I

1111111111 I

t II I I I I I II I

I Ill I I I I I I I

1111111111 I

1111111111 I

1111111111 I

1111111111 I


Appendix ~.O.

Crystallite size determination from X-ray line broadening

Introdudion

The history of this method of identifying apparent crystallite size and the theory behind it ispresented in Mug and Nexander (1974) and Henry, tipson and Wooster (1961).

In practice the apparent crystallite size of a polycrystalline material such as quati can becalculated from the X-ray diffraction line broadening, which is a product of the c~stallite size and thecrystal imperfections such as strain. To calculate the apparent c~stallite size the Scherrer equation isused:

D= KApcose

where D: apparent c~stallite sizeK: crystallite shape mnstant (0.9)A X-ray wavelength (~)~: Pure diffraction broadeninge: Peak position (Bragg angle)

Preparation

For theselection of theStrontian (GSG)quartz (HYD).

purposes of the project, apparent c~tallite size measurements were made for aquartz-bearing granites and metaquartzites, namely hnds End (EG), Shap (SHG),Granites, vein quati in the Bodmin Moor Granite ~D~ and man-made hydrothermal

The initial stage of preparation of the quartz was to remove the quartz grains from the otherminerals in the rock samples. This was done by fly-press crushing of the samples, down to 1-5mm insize, and hand pickingthe quadz grainsout of the crushed material using a binocular microscope and apair of tweezers, Roughly 5g is required. The separated quartz was then hand ground in acetone usinga tungsten-carbide pestle and mortar. This method was used as opposed to temering to minimise thepossibility of inducing or releasing strain by the heat generated whilst temering. The powders (5g) werethen micronised to obta;n the same size distribution for all of the samples, in order to avoid any possiblegrain size variations.

Procedure

A standard unorientated powder mount was prepared for each of the samples. Unfiltered Curadiation was used to obtain ~ lines of (100) reflections. Cu~ was used to avoid the a, az doublet.The wavelength of the Cu@ radiation is 1.3922A which gives the (100) reflection at a position of 18.8°26 for the 4.26 d spacing (intensity 35), the first peak for quartz.

The standard used to account for machine line broadening was that of Arkansas Stone (PhilipsStandard) which has a particle size range of 1-lO~m and is strain free. Measurement of line breadthwas taken at hdf peak heigth.

Measurement of line breadths were taken at the 18.8° 2e position. The conditions of operationwere as follows:

radiation Cu~ (no fitier) : 1.392Acurrent 30rnA, 40 kV


scanning rate 0.002°/sectime constant 5 seccount range max 2000 cps (2E3)scanning range 18.0°-19.2°26chart speed 50mmp 26

Before the equation can be solved a correction for the instrumental line broadening needs to bemade,. This is done by finding the relationship between bm and @/Bwhere:

b: instrument breadth at 1/2 height (Arkansas Stone) in ’26B: sample breadth at 1/2 hei ht in ’26

%~: pure diffraction breadth in 2e

The value of b/B is calculated and read off on the mrrection curve for low angle reflections inKlug and Nexander (1974) to obtain the respective @/Bvalue, The value of @is calculated as:

@is then converted into’8 (026/2) and then into radians

@rads = ~‘8 x n/180

before the Scherrer Equation is calculated to obtain the apparent crystallite size in ~.

The Scherrer equation requires the Bragg angle to be quoted in degrees 8 before cos 8 can becalculated. Therefore the value of the peak position, in degrees 28 needs to be halved and then the cosof the resultant found.

Example

The following is a worked example of the calculation described previously:-

Sample Code:- Hydro (Hydrothermal quartz sample)Standard sample:- Arkansas Stone (Philips Standard)

Peak position for (100) CUM for Hydro:- 18.85°28Wdth at % peak height of Hydro:- 0.1246°26 (B)

.

Mdth at % peak height of Std:- 0.1213°20 (b)

Instmment broadening correction:-

b/B = 0.121 3/0.1246= 0.9735

From Mug and Mexander (1974) the corresponding @/Bvalue can be read off from the graph, FigureA3,0.1. The pure diffraction bread~ @) is equal to:-

in this example

fl/B = 0.4

@= O.1246x0.4

p = 0.04984°28

Undulato~ =tinction of Qu~ in Granites and *dstones

1.0

0.8

0.2

0

1 4

Backreflections

I

o 0.2 0.4 0.6 0.8 1.0b/B

figure A3.O.1. Cuwes for correding X-ray diffradometer line breadths for instrumentbroadening. (After Klug and Aexander, 1974)

to obtain this vatue in ‘e then divide by 2

P = 0.02492° e

then convert the ‘e into radians

~rad~ = ~ x ~180

~rads = 0.0004

Dwide the pew position of Hydro by 2 to obtain the position in ‘e.

18.=0 28R.

6=9.429 e

Now dalate the @tite stie using the ~errer equation

D=~0.0004 @#x 0.9= (-q

D = 3175A

60


Appendix A4.O.

Samples”

The following is the total listing of the samples collected for the project. The codes show firstlythe region/sample location/rock type/number of sample, The short codes just show the sample locationand the rock type.

SW England

SW/BO/G/lSW/CWG/lSWICYIGIISW/GG/G/lSW/KD/G/lSW/CH/G/lSW/DUG/lSW/HG/G/lSWflD/G/lSWflDN/1SW/HD/G/lSWIWGI1SW/LWG/lSW/Sl/G/lSWflH/G/lSW/MUG/lSW/BH/SS/lSW/HH/QQ/lSW/PG/SS/lSW/LN/SH/l

Wales

W/CG/G/lW/G-/G/lW/N-/MG/WmH/QD/’

bke District

LD/sP/wlLD/SP/HF/lLDfl~G/1LD/HS/ST/lLD/BWSS/lLDlsslssllLD/Puss/l

Pennines

P~T/SS/lPICSBSIIPFs/ss/lPNG/SS/l

CLGCYGGGGKDGCHGDLGHGGTDGTDVHDGUGLEGSIGTHGMEGBHSSHHQQPGSSLNSH

Carnmenellis Granite

..

..Bodmin Mmr Granite

,.Bodmin Moor Granite Vein QuartzHingston Down GraniteSt Austell GraniteLands End GraniteStilly Isles GraniteGodolphin GraniteDartmoor GraniteBudleigh Salterton Pebble Beds

.. ..Carboniferous Sandstone (Bude Formation)Devonian Slates

CGG Coedana Granite (Mona Complex)G-G ,. .,N-MG Nanhoron MicrograniteBHQD Quartz-Diorite (Johnston Complex)

SPG Shap GraniteSPHF HornfelsTKMG Trelkeld MicrograniteHSST Silurian MetasiltstoneBKSS Permo-Trias Sandstone (St Bees Head)Ssss .. .. (Penrith)PLSS .. .. (St Bees Head)

BTSS Carboniferous Sandstone (Namurian)Csss .. ..FSSS .. .,HGSS .. ..

Undulatory Efiinction of Quartz in Granites and Sandstones 62

PPWSSI1PBc/ss/l

Scotland

S~D/G/lS~H/G/lS~N/G/lSICFIGIIS/CT/G/lS/CW/G/lSMM/G/lSflF/G/lS/GS/G/lS~MIG/1S/SH/G/lS~R/QDO/lS/CWQDO/lS/GT/QDO/lSIADILWIIsMw/1S~NNN/1S~UNN/1S/CBF/lSFN/MG/lS/CT/ST/lS/CG/SH/lsmmY/1SflG/PY/lS~D/GW/lsmP/ss/lSDM/SSllSIGUSSI1SkBES/lS/NB/SS/lSlsussllSISPISSIISMFISSI1

PESSScss

BDGBHGBNGCFGCTGCWGKMGTFGGSGRMGSHGBRQDOCKQDOGTQDOADLWMLWBNMNBUMNCBFFNMGCTSTCGSHBLPYTGPYDDGWBPSSDMSSGLSSLBSSNBSSSESSSPSSWFSS

Carboniferous Sandstone (Namurian)., .,

Broadlaw GraniteCove GraniteBanavie GraniteCriffel GraniteCreetown GraniteHill of Fare GraniteKemnay Granite

..Strontian GraniteRoss of Mull GraniteMoy GraniteQuartz-Dolerite

....

Lewisian Gneiss.,

Moine (Qtz-Feldspar Granulite).. ,.

FelsiteFurnace MicrograniteDalradian MetasiltstoneDalradian Quartzose-Mica-SchistSilurian Porphyrite

..Devonian Gre~ackeSandstoneSandstoneSandstonePermian SandstoneSandstoneTorridonian SandstoneDevonian SandstoneMoine Sandstone

Undulatory Mnction of Quartz in Granites and Sandstones 63

Appendix X.O.

Terminology

Birefringence

C--isCataclastic

FelsiteGneiss

GraniteGreywackeHomfelsUthic

Porphyrite

QuartzQuartz-dioriteQuartz-dolerite

the interference mlour produced by the crystal as cross polarized lightpasses through it.crystallographic axis in an trigonal crystal such as quati.cataclasis is the process of the mechanical fracturing of a rock Lassociated with metamorphism. A titaclastic rock is one which has undergonethis process,fine grained acid igneous rock.banded metamorphic rocks produced during high grade regionalmetamorphism.coarse grained acid igneous rock.sediment consisting of mainly Iithic fragments, often poody sotied.medium to fine grained rock produced by therma~contact metamorphism,fragments of rock occuring in later formed rocks. Microgranite medium grainedacid igneous rock.also known as microtonalite, medium grained rock containing mainly.feldspar phenoc~sts.common mineral made up of silimn and o~gen atoms, Si02.coarse grained intermediate igneous rockmedium grained basic igneous rock

pzz326 - trl.co.uk fileabo~ ~l the transport research laboratory is an executive agency of the...

Documents