Search

A.H. Cottrell led a large team of scientists and engineers as Deputy Director of the Metallurgy Division of the Atomic Energy Research Establishment at Harwell, between 1955 and 1958. The team sought to develop a fundamental understanding of the effect of aggressive operating conditions on the properties and behaviour of materials used in the construction of reactors, which provided important and critical insights into their design, operation and safety. The story is told through the experiences and research activities of one of Cottrell’s junior research staff. It offers an insight into the contributions made by Cottrell to the science and technology regarding materials used in the early days of building reactors for the generation of nuclear energy.

Optical microscopy has been used to study the sequence of events associated with drying of a silica-based sol-gel. Drops of sol containing 50 wt% silica were deposited on glass slides and dried in cells containing sulphuric acid as a drying agent. The drops, on the slide, were about 3.5 mm in diameter and before the onset of drying had a spherical cap shape. At the start of drying the rim of the drops became pinned to the glass surface. The humidity of the drying environment had a strong effect on shrinkage within the drop and the processes that occurred during drying. In very dry environments evaporation occurred primarily from the edges of the drop and resulted in a radial concentration gradient in the gel. Shrinkage resulted in a pattern of radial cracks that spread from the outside in a direction orthogonal to the iso-concentration profiles. Cracking was preceded by shear deformation. Drying was more uniform in humid environments and occurred from both the rim and the top surface of the drop. Wrinkling of the skin of gel that formed on the surface of the drop was followed by radial and circumferential cracking, except in very humid environments. The results are attributed to the influence of the radius of curvature of the drop surface on the rate of evaporation and the relative rates of flow and diffusion in the drying sol-gel.

The progressive development of primary arrays of straight and curved cracks during the drying of thin films of aqueous silica sol-gel is investigated experimentally by direct observations using light microscopy. Three main test configurations are used to generate a range of drying profiles in the sol-gel: (i) films between a glass plate and a flexible glass cover slip; (ii) films between two glass plates held apart by a rigid spacer, and; (iii) uncovered films on a glass plate. Cracks in the primary arrays grow normal to the iso-concentration water profiles. The arrays of primary cracks observed in basalt lava flows and described by Spry (J. Geol. Soc. Aust. 8 (1962) 191) i.e. parallel straight columns, chevrons, rosettes, fans, inverted fans and basins, are reproduced in the pseudo two-dimensional films of sol-gel. The development of both primary and secondary cracking patterns is strongly dependent on the constraints associated with the bonding of the sol-gel to the glass and the subsequent splitting of the gel-glass interfaces.

Observ. XV. illustrated by Schem. IX. Figur:1 (figure 1 of this paper) in Robert Hooke's Micrographia (1665)1 is a description of Kettering–stone ‘which is brought from Kettering in Northampton–shire, and digg’d out of a Quarry, as I am inform'd’. As Curator of Experiments for the Royal Society from 1662, Hooke was charged by the Society to bring in at every meeting one microscopical observation at least. The minutes of the Society2 record that on 15 April 1663 ‘Mr Hooke showed the Company two Microscopicall Schemas; one representing the Pores of Cork … the other a Kettering Stone, appearing to be composed of Globules; and those hollow ones, each having 3 Coatings, sticking to one another, and so making up one entire firm stone’.

The development of roughness on the fracture surfaces of a brittle, glassy, epoxy resin from the mirror-to-mist transition to macroscopic bifurcation has been investigated using optical microscopy, scanning electron microscopy (SEM) and contact and non-contact laser profilometry. Most of the observations were made on specimens fractured in edge-notched tension. In a series of tests the initial crack length was varied to obtain fracture surfaces formed by accelerating and decelerating cracks without macroscopic bifurcation (specimen A) and by cracks which accelerated continuously to macroscopic bifurcation (specimen B). Some observations were made on specimens tested in compact tension to study changes in fracture surface topography associated with crack arrest in stick-slip fracture. There was a close correlation between the topographical detail revealed by the different techniques. In specimen A the roughness increased progressively from the mirror-to-mist transition and reached a maximum before decreasing as the crack decelerated. The topographical features revealed by optical microscopy and SEM were the same for accelerating and decelerating cracks at the same roughness value. In specimen B the roughness increased continuously to macroscopic bifurcation. There was a close similarity between the topographical features at all levels of roughness. A simple model for the basic step involved in roughness formation is presented which involves an element of the crack tip tilting out of the plane of the main crack before stopping (micro-bifurcation). The scale of micro-bifurcation ranged from 3 μm in the early stages of mist, when the crack velocity was close to 10% of the shear wave velocity, to the full width of the specimen (6 mm) at macroscopic bifurcation. The micro-bifurcation process develops from crack surface undulations and does not involve micro-crack nucleating ahead of the main crack. It is concluded that the relationships between crack velocity and dynamic stress intensity, and the value of the limiting crack velocity, must be interpreted in terms of micro-mechanical processes at the crack tip which are strongly dependent on specific material characteristics.

A modification of Sommer's classical experiment [4] has been used to fracture rods of a brittle epoxy resin in mixed mode I/III stress conditions. The nucleation and growth of cracks in an increasing KIII/KI stress field have been investigated, particularly in relation to the formation of river lines and the evolution of multiple cracks between river lines to form smooth helicoid surfaces. The fractographic features associated with the progressive development of river line patterns involving crack bowing at river line steps, interaction between arrays of cracks and the coalescence of river lines are described. As KIII/KI increases the scale of the river line patterns increases but the patterns remain self similar. Using principles from differential geometry it is shown that helicoid surfaces can be generated entirely by crack evolution involving ‘tilting’ without ‘twiting’.

Cone fractures have been produced in epoxy resin samples in four different test configurations using a cylindrical indenter. The shapes of the cracks were determined using optical microscope focusing methods and it was found that they were strongly dependent on test geometry. Large changes in cone angle with crack length were observed. The evolution of the cracks was determined by mapping the fine river line markings on the fracture surface. In all cases crack nucleation occurred at point sources close to the edge of the indenter and then the crack grew in the material round the base of the indenter before expanding outwards. Numerous crack arrest markings were also mapped. It was found that the number and distribution of crack arrests was dependent on the test configuration and re-nucleation occurred after each arrest. The results are interpreted in terms of the no-twist growth constraint, which applies to crack growth in brittle solids. This leads to characteristic patterns on the fracture surface. It is shown that point nucleation results in a strongly non-symmetric fracture path and that the requirement that nucleation occurs at a point has a significant effect on the interpretation of fracture data in other test configurations.

Publisher Summary This chapter discusses jogs and intersection of dislocations in crystals. Since even well-annealed crystals usually contain a network of dislocations, it follows that every slip plane will be threaded by dislocations and a dislocation moving in the slip plane will have to intersect the dislocations crossing the slip plane. Trails of defects and prismatic loops are often produced during plastic deformation is left behind by moving dislocations and is a direct result of edge jogs on screw dislocations. The simple geometrical models of intersections have to be modified somewhat when the intersecting dislocations are extended into partials, as is expected to be the case in face-centered cubic metals with low stacking fault energy. The dislocations can only cut through each other by constricting to form perfect unextended dislocations in the region close to the intersection. The dislocations can then separate and elementary jogs are produced on the extended dislocations. It is clear that to form jogs on extended dislocations work will have to be done to constrict the dislocations and so the energy of the jog will depend on the stacking fault energy or, alternatively, on the spacing of the partial dislocations. The intersection of gliding dislocations with the forest dislocations that intersect their slip planes was treated as a short-range effect leading to jog and kink formation in the dislocation core. Dislocations can interact at long range, however, by virtue of the stress fields they produce. A particularly interesting dislocation structure, which provides a method for direct experimental measurement of stacking-fault energy in face-centered cubic metals, arises when the dislocations forming an attractive junction are extended.

Publisher Summary This chapter discusses the movement of dislocations in crystals. There are two basic types of dislocation movement. Glide or conservative motion occurs when the dislocation moves in the surface that contains both its line and the Burgers vector: a dislocation able to move in this way is glissile, one that cannot is called sessile. Climb or nonconservative motion occurs when the dislocation moves out of the glide surface, and thus normal to the Burgers vector. Glide of many dislocations results in slip, which is the most common manifestation of plastic deformation in crystalline solids. It can be envisaged as sliding or successive displacement of one plane of atoms over another on so-called slip planes. Discrete blocks of crystal between two slip planes remain undistorted. Further deformation occurs either by more movement on existing slip planes or by the formation of new slip planes.

Publisher Summary This chapter deals with dislocation arrays and boundaries in crystals. Plastic deformation of crystalline materials leads to the formation of three-dimensional arrays or distributions of dislocations that are characteristic of crystal structure of the material being deformed, temperature of deformation, strain, and strain rate. Additionally, such features as grain boundaries, precipitates, and stacking fault energy affect the distribution of the dislocations. The hardening of crystals during plastic deformation is due to the increase in dislocation density and the mutual interaction between dislocations. The simplest boundary is the symmetrical tilt boundary. Line defects in interfaces between crystals can play an important role in transformations involving the growth of one crystal at the expense of another, e.g., twinning, martensitic transformations, and precipitation. In the simplest picture, the defect is akin to a step and its motion over the interface is the mechanism that transforms atoms of one crystal to sites of the other. For the movement of a boundary containing interfacial dislocations, several conditions must be satisfied. First, if slip is involved the dislocation must be free to glide on its slip plane, i.e., the plane that contains the line and its Burgers vector. Second, matter must be conserved if a diffusional flux of atoms is required for motion of interfacial defects. Third, the thermodynamic condition that movement results in either a reduction in the energy of the boundary or, in the case of movement induced by an externally applied stress, that the stress does work when the boundary moves. Fourth, the driving force due, for example, to the externally applied stress or the excess vacancy concentration must be sufficient to produce dislocation movement.

Many common metals such as copper, silver, gold, aluminum, nickel, and their alloys, have a face-centered cubic crystal structure. The deformation behavior is closely related to the atomic structure of the core of dislocations. In face-centered cubic crystal structure, the atomic structure of the core of dislocations is more complex than other crystals. The shortest lattice vectors, and therefore the most likely Burgers vectors for dislocations in the face-centered cubic structure, are of the type 1/2〈110〉 and 〈001〉. Since the energy of a dislocation is proportional to the square of the magnitude of its Burgers vector b 2 , the energy of 1/2〈110〉 dislocations in an isotropic solid will be only half that of 〈001〉, i.e., 2 a 2 /4 compared with a 2 . Thus, 〈001〉 dislocations are much less favored energetically and, in fact, are only rarely observed. Since 11/2〈110〉 is a translation vector for the lattice, glide of a dislocation with this Burgers vector leaves behind a perfect crystal and the dislocation is a perfect dislocation . Another dislocation arrangement has been observed in metals and alloys of low stacking-fault energy following treatment that produces a supersaturation of vacancies.

Publisher Summary When annealed crystals are deformed there is a rapid multiplication of dislocations and a progressive increase in dislocation density with increasing strain. After large amounts of plastic deformation the dislocation density is typically in the range 1014 to 1015 m–2. In the early stages of deformation dislocation movement tends to be confined to a single set of parallel slip planes. Later, slip occurs on other slip systems and dislocations moving on different systems interact. The rapid multiplication leads to work hardening. The origin of dislocations and main mechanisms of dislocation multiplication are described in this chapter. There are two main sources of dislocations in freshly grown crystals. First, dislocations or other defects may be present in the “seed” crystals or other surfaces used to initiate the growth of the crystal. Second, “accidental” nucleation may occur during the growth. When a dislocation is created in a region of the crystal that is free from any defects the nucleation is referred to as homogeneous. This occurs only under extreme conditions because a very large stress is required. To account for the large plastic strain that can be produced in crystals, it is necessary to have regenerative multiplication of dislocations. Two mechanisms are important: Frank-Read type sources and multiple cross glide.

Publisher Summary This chapter discusses dislocations in major metallic crystal structures, other than face-centered cubic crystals. In body-centered cubic metals (e.g., iron, molybdenum, tantalum, vanadium, chromium, tungsten, niobium, sodium, and potassium), slip occurs in closepacked 〈111〉 directions. An important feature of dislocations in ionic solids is that electrical charge effects can be associated with them. Dislocations in covalent crystals affect both mechanical and electrical properties. In materials with layer structures, slip occurs readily in planes parallel to the layers and is almost impossible in nonlayer planes. The dislocation arrangements and Burgers vectors are therefore confined mainly to the layer planes. The weak binding between layers can result in stacking faults of low energy and hence unit dislocations are usually widely dissociated into partial dislocations. Plastic deformation of crystalline and semicrystalline polymers involves the mechanisms well established for other crystalline solids, namely dislocation glide, deformation twinning and, in some cases, martensitic transformations.

For a growing crack the conventional definitions of tilting and twisting are inadequate. New definitions are proposed, based on differential geometry, and it is shown that, in homogeneous, isotropic, brittle solids, non-planar crack growth must occur entirely by tilting movements. Examples are given of the growth of cracks on curved surfaces which illustrate that the no-twist condition produces significant constraints on the path of fracture. The development of fracture surfaces when cracks are subject to mixed-mode loading conditions are described with particular reference to the influence of mode III, twisting conditions. It is shown that the requirement that only tilting can occur leads to many characteristic fractographic features.

The investigation of damage tolerance of composite materials comprises a spectrum of topics. A review of research in the area of damage tolerance critically exposes the need to identify basic damage mechanisms in order to develop more damage-tolerant materials and structures. The importance of characterizing damage mechanisms is emphasized and examples are given to illustrate the approaches which can be used. New results on the characterization of impact damage in carbon fibre reinforced epoxy resin composite laminates are presented. The discussion leads to the identification of a number of issues that require further research for the development of a mechanism-oriented rational damage-tolerant design approach.

Publisher Summary This chapter discusses the elastic properties of dislocations. The atoms in a crystal containing a dislocation are displaced from their perfect lattice sites, and the resulting distortion produces a stress field around the dislocation. The dislocation is therefore a source of internal stress in the crystal. The region above the slip plane contains the extra half-plane forced between the normal lattice planes, and is in compression: the region below is in tension. The stresses and strains in the bulk of the crystal are sufficiently small for conventional elasticity theory to be applied to obtain them. This approach only ceases to be valid at positions very close to the center of the dislocation. Although most crystalline solids are elastically anisotropic, i.e. their elastic properties are different in different crystallographic directions, it is much simpler to use isotropic elasticity theory.

Publisher Summary This chapter describes the techniques of analyzing dislocations. A wide range of techniques has been used to determine the properties of individual dislocations and study the distribution, arrangement, and density of dislocations. Among the experimental techniques, transmission electron microscopy has been dominant for the past 50 years. In this method, dislocations are examined either singly or in groups at high magnification in specimens ∼0.1 to 4.0 μm thick. Four other experimental techniques have proved useful for providing information about dislocations either in particular materials or prior to the development of transmission electron microscopy: Surface methods, which reveal the point of emergence of a dislocation at the surface of a crystal; Decoration methods, in which dislocations in bulk specimens transparent to light are decorated with precipitate particles to show up their position; X-ray diffraction, in which local differences in the scattering of X-rays are used to show up the dislocations; and field ion microscopy and atom probe tomography, which reveal the position of individual atoms. The power of computers can be exploited to investigate properties of dislocations that are not open to direct study by experiment. The techniques involved can be divided into two broad categories. One simulates dislocations and other defects in crystals modeled at the atomic scale. The other treats the crystal as an elastic continuum with no atomic discreteness.

Publisher Summary This chapter discusses the defects in crystals. Dislocations are an important class of defect in crystalline solids; therefore, an elementary understanding of crystallinity is required before dislocations can be introduced. Metals and many important classes of nonmetallic solids are crystalline, i.e. the constituent atoms are arranged in a pattern that repeats itself periodically in three dimensions. The actual arrangement of the atoms is described by the crystal structure. The crystal structures of most pure metals are simple, the three most common being the body-centered cubic, face-centered cubic, and close-packed hexagonal structures. The arrangement of atoms in a crystal can be described with respect to a three-dimensional net formed by three sets of straight, parallel lines. The lines divide space into equal sized parallelepipeds and the points at the intersection of the lines define a space lattice. Every point of a space lattice has identical surroundings.

Publisher Summary The theoretical shear strength of a perfect crystal is many orders of magnitude greater than the observed critical shear strength of real crystals that contain dislocations. This chapter considers the factors affecting the strength of crystals as an introduction to the way that an understanding of the properties of dislocations can be used to interpret the behavior of crystalline solids. Aparts from effects associated with diffusion at high temperature, plastic deformation occurs by the glide of dislocations and hence the critical shear stress for the onset of plastic deformation is the stress required to move dislocations. This is measured usually by a tensile test in which the specimen is elongated at a constant rate and the load on the specimen is measured simultaneously with the extension.

The previous two chapters are concerned with the elastic behaviour of composites containing fibres which are, in effect, infinitely long. The preparation of composites containing short fibres (or equiaxed particles) allows scope for using a wider range of reinforcements and more versatile processing and forming routes (see Chapter 11). Thus, there is interest in understanding the distribution of stresses within such a composite, and the consequences of this for the stiffness and other mechanical properties. In this chapter, brief outlines are given of two analytical approaches to this problem. In the shear lag model , a cylindrical shape of reinforcements is assumed, and the stress fields in fibre and matrix are simplified so as to allow derivation of straightforward analytical expressions for the composite stiffness. The Eshelby method , on the other hand, is based on the assumption that the reinforcement has an ellipsoidal shape (which could range from a sphere to a cylinder or a plate). This allows derivation of an analytical solution which is more rigorous than that of the shear lag model, but with the penalty of greater mathematical complexity. In the treatment given here, attention is concentrated on the principle of the Eshelby approach; sources are suggested for readers needing more mathematical details. The shear lag model The most widely used model describing the effect of loading an aligned short-fibre composite is the shear lag model, originally proposed by Cox (1952) and subsequently developed by others (Outwater 1956, Rosen 1960, Dow 1963), which centres on the transfer of tensile stress from matrix to fibre by means of interfacial shear stresses.

The model for the craze controlled fracture process in polystyrene has been developed further by taking into consideration the micromorphology of the crazes in which the nucleation and propagation of cracks occurs. The micromorphology of crazes formed in thin films of polystyrene, some of which had fractured, has been characterized by means of transmission electron microscopy. The observed micromorphological detail has been shown to be consistent with the micromorphology of the fracture surfaces of bulk specimens. In particular, the slow and fast regions of crack propagation which result in distinctly different fracture surface morphologies have been shown to be associated with differences in micromorphology which occur along the length of a craze.

Filament-wound pipes pressurized in one of two modes, to give biaxial or uniaxial stress conditions, have been examined using standard microscopic techniques. Whitening of the pipes which develops during testing has been correlated with different types of cracking. A mechanism for weepage in filament wound pipes by the crossing of transverse cracks in adjacent laminae is described. The types of crack present in the pipe at loads up to final failure were found to be consistent with predictions of the stresses acting in the pipe walls.

The ability of polymer-based fibre reinforced composite materials to absorb energy when subjected to high stress depends on the micromechanics of deformation and fracture of the individual laminae and the interaction between laminae. The micromechanismsdepend on the properties of the matrix, fibres and fibre-matrix interface. The capacity of material to absorb energy is important in many applications and particular reference is made to the response of structures to impact.

The molecular and structural parameters controlling the mechanical properties, deformation and fracture of spherulitic nylon 6 have been investigated. The nylon was prepared by the anionic polymerization of e-caprolactam and the polymerization conditions were varied to give samples having a range of spherulite diameter, molecular weight and degree of crystallinity. The tensile properties and fracture mode of the nylon varied considerably with degree of crystallinity and polymerization temperature. High crystallinity and low polymerization temperatures below 423 K gave a brittle material. Polymerization above 423 K resulted in a ductile material which showed a yield drop. In this material final fracture was preceded by the formation of inter and trans spherulitic cracks which coalesced to form a large cavity that led to final failure. In nylon having a low degree of crystallinity, fracture was fibrillar in nature and occurred by the ductile drawing of the material to strains greater than 250%.

A modified fracture toughness test has been used to measure the growth of cracks in unidirectional glass fibre/polyester resin composite materials in the presence of a dilute hydrochloric acid. Crack growth rates perpendicular to the fibre axis have been measured over a range of stress intensities. Scanning electron microscope studies of the fracture surfaces have shown that the micromechanisms of nucleation and propagation are dependent on stress intensity. The use of crack growth data to predict component lifetimes and the existence of inherent flaws in the material are discussed.

The deformation and fracture behavior of injection molded plaques have been determined, and the results interpreted in terms of the effect of molecular orientation on the crazing and shear yielding behavior. The molecular orientation was characterized by optical birefringence. A range of injection molding conditions and two mold thicknesses were Used and this resulted in a large variation in the molecular orientation, particularly through the sheet thickness. Tensile tests were made on samples cut at different angles to the injection molding direction. The moldings are considered to consist of a composite of layers of material with different orientation, and the properties of the samples cut from the molding are analyzed in terms of the properties of each layer. Results from material oriented unidirectionally by hot drawing have been used to predict the composite properties, and good agreement has been obtained.

Tensile and compression tests have been carried out over a range of temperatures between 78 and 360 K to determine the micromechanisms of deformation and the mode of fracture. In compression, deformation occurred by shear band propagation whereas, in tension, deformation was restricted to crazing followed by fracture. Anomalously low crazing stresses were observed for tests on specimens immersed in liquid nitrogen. Analysis of the mechanism of fracture confirmed previous work relating crazing and fracture and the fracture model has been modified slightly to take account of additional features observed in failure of crazes at high temperatures.

The deformation and fracture behaviour of sheet polystyrene compressed edge-on between flat dies has been studied at room temperature. Using this test procedure it is possible to produce shear bands and crazes in the same specimen. Fracture occurs by nucleation of a crack in a craze which lies in the plane normal to the plane of the sheet and parallel to the compression axis. The amount of shear band formation depends on the prior heat treatment. No shear bands were observed in quenched material and larger compression displacements were required to nucleate crazes and fracture. Intense shear bands were formed in slowly cooled material and their interaction resulted in easier craze nucleation and fracture. Quench ageing studies were made between 293° and 360°K. There was a smooth increase of the yield stress with ageing temperature but the displacement to fracture showed an abrupt decrease which corresponded with the ageing treatment required for shear band formation. Fracture surface studies have...

Thin sections of pre-crazed bulk specimens of polystyrene and thin sections containing crazes which were obtained by straining fresh films of polystyrene have been examined in the electron microscope. The use of reinforcing impregnants, a cold stage or a diamond knife was not required for ultramicrotomy. The microstructure of crazes in polystyrene has been shown in the later stages of craze deformation to consist of a well-defined fibrillar structure. The fibrils have a characteristic diameter of 200 to 400 A and in the electron microscope the fibrils exhibit a beaded contrast. The beaded contrast can be considered as evidence for the existence of a domain structure in amorphous polymers. Some preliminary observations of the structure of crazes formed by straining in the electron microscope and some observations on the fracture surfaces of crazes are also reported.

The effect of pre-existing micro-cracks on the fracture of [010] tungsten single crysta's has been investigated. Cleavage cracks on (010) planes were introduced into the edge of the specimens by a spark-machining technique. Tensile tests at 77°K showed that the fracture stress was dependent on the crack length according to the Griffith criterion. At this temperature the specimens failed in a completely brittle manner. A surface energy of 6300 ergs cm−2 was obtained in close agreement with the theoretical value for the true surface energy. Metallographic examination of the fracture surfaces showed that the markings on the surface depend on the stress level at fracture. One test was carried out to determine the effect of prestraining at 295°K on the subsequent fracture behaviour at 77°K. The effect of temperature on the fracture of specimens containing a constant crack length was determined between 77 and 478°K. As the temperature increased, the crack propagation mechanism changed from fast cleavag...

The microstructure of compression moulded TPX, a semi-crystalline copolymer based on poly(4-methylpentene-1), has been examined using several techniques and has been shown to consist of small clusters of lamellar blocks about 5 μm in diameter. It is suggested that the clusters represent the early stage of spherulitic growth. The tensile properties of this material have been measured at temperatures between 77 and 420K. There is a sharp change in tensile ductility at about 290K which corresponds with a change in the microdeformation processes. Optical studies have shown that between 285 and 330K ductility is associated with stress whitening and that above 330K large scale drawing occurs. Optical observations together with examination of thin sections in the electron microscope indicate that whitening is due to the formation of a large number of very fine irregular crazes. The microstructure of the crazes resemble those formed in polystyrene and other amorphous polymers.

Fracture surface studies have been made on pre-cracked tungsten single crystals pulled in [010] at temperatures between 200 and 3000K. The cracks were introduced by a spark technique and were atomically sharp. The appearance of the cleavage face varied with crack propagation stress and temperature. At low stresses and low temperatures a large proportion of the surface was ‘mirror’ smooth. Using oblique illumination the smooth surface was found to be covered with Wallner lines which started either at the edge of the crystal or at small cleavage steps. As the stress and temperature increased the area of smooth surface became progressively smaller and was restricted eventually to a narrow spike along [001]. The density of river lines also increased. A number of boundaries separating regions of different river line density were observed which spread out from the side of the fracture surfaces and were attributed to interaction between the stress field at the tip of the crack and a transverse elastic stress wave. The velocity of the cleavage cracks was estimated from the shape of the Wallner lines and from the shape of the river line boundaries. The limiting crack velocity approached 0.60 and 0.82 of the velocity of transverse elastic waves at 77° and 20°K respectively. At higher temperatures, in the presence of river lines, the velocity was between 0.20 and 0.55 of the transverse elastic wave velocity.

Abstract Cleavage cracks have been introduced into 〈100〉 axis single crystals of tungsten by a modification of the spark discharge technique used by Hull, Beardmore and Valintine (1965). It has been possible to insert cracks in specimens at 77°K and to subsequently test these specimens in tension at 77°K without an intermediate rise in temperature. Various other test conditions have also been used. From the values of fracture stress and crack length obtained, values of the effective surface energy, γ, have been determined. In analysing the results, account has been taken of the shape of the spark crack and specimen geometry and dimensions using theoretical analyses of the stress concentrations associated with cracks. Both the temperature of crack initiation and the temperature of crack propagation were found to affect γ. For specimens in which both initiation and propagation of the cracks took place at 77°K, a value of γ = 1700 ± 300 ergs/cm2 was obtained.

The effect of temperature on the tensile fracture stress of polycrystalline tungsten of average grain diameter 2d = 0.03 mm has been investigated in the temperature range 77°–480°K. The onset of the ductile-brittle transition, that is the brittle-semi-brittle transition, occurred at 350°K. Below this transition the fracture stress increased with decreasing temperature. Above 473°K grain-boundary cracking was observed on the specimen surface. The effect of spark-induced microcracks on the fracture stress was studied as a function of microcrack length at 77°K and testing temperatures between 77° and 480°K. The fracture stresses at 77°K followed the Griffith relation to give a value to the surface energy that was close to the theoretical value for the true surface energy. The effect of blunt grain-boundary cracks on the fracture of tungsten was shown by pre-straining the material to 2 and 4% strain at 473°K prior to testing at lower temperatures. The temperature dependence of this fracture stress was less than for sharp spark-induced cracks.

The crazing behaviour of a grade of polystyrene which crazes profusely when tested in tension has been studied at 293°K using quantitative optical, electron and electron scanning microscopy. Tensile tests on this material showed that a high density of crazes was produced at room temperature. The crazes formed in a zone which spread progressively down the specimen. The plastic strain associated with crazing was sufficient to cause craze yielding and the phenomenon is compared with discontinuous yield in mild steel. Craze-yielded specimens were used to determine the stress—strain curve of craze material. Microstrain techniques were developed for this work and the total volume of craze was determined from the density, distribution and thickness of the crazes. It was found that the stress—strain characteristics of polystyrene crazes are similar to those observed by Kambour and Kopp (1969) for solvent-induced crazes in polycarbonate.

Single crystals of a 3·25 % silicon iron were deformed in tension between 20 and 293 °K. The orientation was varied systematically between [010] and [110], to determine the orientation dependence of slip and twinning. The operative slip and twinning systems were measured by two surface analysis. At 20 °K all specimens twinned and fractured; at 77 °K crystals within 11° of [010] twinned and fractured and the remainder yielded before fracture; at 195 and 293 °K all specimens yielded. Yielding occurred by the formation and propagation of slip bands along the specimen. At 77 °K slip was confined to {011} planes but at higher temperatures slip occurred on the plane containing the slip direction with the maximum resolved shear stress independent of whether or not it was a low index plane. The yield propagation stress varied with orientation in close agreement with that expected for slip on an {011} system at all temperatures. Twinning occurred on those systems with the maximum resolved shear stress on the twin plane in the twinning direction and the results support a critical resolved shear stress law.

The electron microscopy techniques which can be used to provide definite information about fracture processes in amorphous thermoplastics are summarized. A comparison is made of the micromorphology of crazes formed in thin films of polystyrene and the micromorphology of the fracture surfaces of bulk specimens. This shows that the fracture behaviour of thin films simulates closely the fracture behaviour of the bulk in craze controlled fracture and that the former procedure may therefore be used for high resolution studies of fracture processes. A brief description is also given of the precautions that have to be taken in the study of electron beam sensitive materials and of the developments in electron optical instruments which should result in an improvement in the facilities available for electron microscopy studies of polymers.

Tungsten is usually considered to cleave on the {100} plane although several theoretical calculations predict a lower surface energy for the {110} plane. It is shown that {110} clearvage can be produced in ⟨110⟩ axis single crystals of tungsten. A single experimental value obtained for the {110} plane implies that the surface energy of this plane is lower than that of the {100} plane. It is noted that relative values of surface energy do not alone dictate the cleavage plane in the completely brittle fracture of a single crystal. The direction of applied stress must also be considered.

The deformation of single crystals of iron 3% silicon having orientations between [001]-[111] and [111]-[110] has been studied as a function of orientation and temperature. At 293°K the operative slip systems, determined by slip trace analysis, were close to those of maximum resolved shear stress though the orientation dependence of the yield propagation stress was compatible with slip on {110} systems at all orientations. The variation of the yield propagation stress with temperature was studied for crystals which slipped on {110} and {112} slip planes respectively at and above 293°K. A slight difference in temperature dependence was observed at lower temperatures. The results are compared with those obtained in recent similar work and it is concluded that they may best be explained in terms of a model of thermally activated composite slip in which all dislocation movement is on {110} slip planes.