The physical mechanisms deduced from the direct observation of inert gas bubbles in the electron microscope are used to evolve a model, or theory, for swelling. The model assumes the insolubility of the inert gases and the bodily migration of the gas bubbles, and relies upon the equation of state for the inert gas, upon the surface energy, the surface self-diffusion coefficient, and the lattice parameter of the solid, and upon some internal driving force, which is believed to be related to the line tension of dislocation lines. With values for these parameters and the irradiation and annealing conditions, estimates for the bubble size (r) and number/cm3(n) are made. From these the swelling (ΔVV) can be calculated, but the effect of grain size, grain boundary migration, and other structural features of the solid must be considered.

Electron microscopic examination of thin films from ordered AuCu I has revealed faulted regions which have been interpreted as twins in which only the super-lattice is rotated. The development of these “order” twins can be inferred, and it is suggested that they are generated by the movement of 12〈101〉 dislocations on successive {101} planes. This twinning, as a mode of stress relief, is discussed.

Rows of prismatic dislocation loops have been observed in thin films obtained from copper foils which have been water quenched from 1055°C. These rows emanate from precipitates in the copper and lie in the 〈110〉 directions, but only those lying nearly parallel to the surface contain many loops. The spacing of the loops is consistent with the theory of Bullough and Newman(1) which enables a value of the critical resolved shear stress to be estimated. This value indicates that the outside loops in the rows inclined to the surface of the thin film are lost to the surface by slip under the influence of their image forces. The loops, which are thought to be equivalent to platelets of interstitial atoms are too numerous to result merely from the differential expansion of the precipitate and the copper, which suggests growth of the precipitate during quenching.

Thin films of zone-refined aluminium have been bombarded at 80°c with a flux of 1.6 × 1010 fission fragments cm−2 sec−1 for various times. Defects can be discerned after 30 sec when the films are viewed in the electron microscope and after longer timès they enlarge into dislocation loops. The larger loops lie near {110} planes with 〈110〉 Burgers vectors. The configuration of the loops and their behaviour on heating suggest that they migrate by a combined conservative climb and slip process. At the higher doses loops lie within the larger loops and the diffraction contrast behaviour suggests that these loops have opposite Burgers vectors, i.e. a vacancy loop is encompassed by an interstitial loop and vice versa. In cases where two loops combine the dislocation interactions confirm this. It is concluded that both vacancy clusters and interstitial clusters form during the bombardment and that initially each type attracts its own; kind and repels the other so that each grows almost independently by a process involving both conservative climb and slip. As they enlarge slip becomes the more important process and near-by loops of the opposite kind, lying on similar slip cylinders slip together, partially cancelling one another. The slip and climb processes are compared, using a simple example where circular prismatic edge dislocation loops interact in aluminium, and found to be compatible with the interpretation. Some implications of the behaviour are given.

The helium produced by the B10 (n, α)Li7 reaction is expected to cause the volume of boron-containing materials to increase during irradiation at high temperature. In this investigation a 4 wt % boronsteel tube irradiated for a period of 30 months (1460 MWD/tonne) at 300° C in a Calder Hall reactor was subsequently heated. No metallographic change was detected until segments were heated to 800° C, when small bubbles 300–600 Å in diameter appeared in the Fe2B particles. After heating to 900° C the overall volume increased by ~ 5 % when the bubbles in the Fe2B phase had enlarged, except near the phase boundaries, and helium bubbles were also observed in the iron phase within the recoil distance (~ 4 microns from the phase boundary). Eventually, after annealing at 1000° C, large irregular bubbles were seen in phase boundaries and these were then the main cause of the overall volume increase; ~ 30 % after 24 h at 1050° C. The effects were most marked near the outer surface of the tube and became less marked towards the inside surface, but bubbles of helium, presumably produced by epithermal neutrons, were observed even on the innermost surface.

Electron transmission microscopy and selected area diffraction have shown that the dislocation loops in aluminium bombarded with 38 Mev alpha-particles lie generally upon the {110} planes with a [110] Burgers vector normal to the loop. These loops are only seen where the alpha-particles come to rest. The theory of diffraction contrast at dislocations (Hirsch et al. 1960) has been used to determine the sense of the Burgers vectors and shows that the loops are formed from interstitial atoms as previously inferred from their annealing behaviour (Barnes and Mazey 1960).

Hot-pressed beryllium has been examined after irradiation with more than 1021 fast neutrons/cm2 while at a temperature which varied between 280° C and 480° C. The smaller helium content (3.7 cm3 at NTP/cm3 Be), compared with the previously examined beryllium1), produced much smaller volume increases. The volume did not measurably increase during irradiation but did so only after heating for one hour at above 800° C. A small number of tensile samples were cut from the material and pulled at room temperature, 300° C, 450° C and 600° C. The 0.1 % proof stress and ultimate tensile stress had increased by a factor of more than 2, and the elongation reduced to practically zero at all these temperatures. Some recovery of the mechanical properties was achieved by annealing at temperatures above 800° C. The results are described in terms of the distribution of the helium bubbles which were observed in replicas and thin films from the material.

Abstract The electron microscope has been, used to examine point defect clusters produced in copper and aluminium foils bombarded with l·4 × 1017 alpha-particles cm−2. The foils were bombarded as a stack and, in copper, both those through which the alpha-particles had passed and that in which they came to rest appeared similar, containing dislocation loops (r ∼ 200 Å) and a background of more numerous and smaller dots (r∼20 Å). The behaviour at grain boundaries suggested that these two types of cluster were due to different point defects. Whereas the dislocation loops normally annealed out at about 350°c, in the foil in which the alpha-particles came to rest (containing ∼ 1015 atoms of helium cm−2) they grew to form a dislocation tangle, and eventually small helium bubbles (r ∼ 40 Å) appeared in the same number as the original small dots. These results indicate that the dislocation loops are due to the clustering of interstitial atoms and the small dots are clusters formed from vacancies. It is also inferred that the helium atoms are in interstitial positions during the bombardment, and form bubbles by nucleating upon the dusters.

Small pieces of a beryllium shim rod from the Materials Testing Reactor have been heated and changes in microstructure, volume and mechanical properties observed. The beryllium which was estimated to have received a dose of about 7.6 × 1021 fast neutrons/cm2 contained about 10 cm3 at N.T.P. of helium per cm3 and this precipitated as fine gas bubbles during heating for an hour at temperatures above 600° C. These bubbles grew, particularly at the grain boundaries, during further heating, until at 1100° C they were about 10−3 cm diameter. The precipitation and growth of the bubbles produced an increase in the volume of the samples, commencing at about 600° C and eventually reaching 30 % at 1000° C. The irradiated beryllium is very brittle, even above room temperature, and this may be aggravated by the large gas bubbles lying on the grain boundaries. The observations are discussed in relation to those already reported for beryllium containing large local concentrations of helium injected with a cyclotron.

Disks of extruded and cast beryllium have been injected with helium by bombardment with alpha-particles in a cyclotron. Microscopic observation of the region at the end of the alpha-particle track shows that the helium contained there precipitates as gas bubbles only if the beryllium is heated to about 800°C, when the grain boundaries act as sources of the vacancies necessary for the precipitation. Further heating causes the bubbles, which are initially about 4 × 10−6 cm in diameter, to enlarge and to reduce in number; this is accelerated at the grain boundaries. The large difference in grain size between the extruded and cast beryllium permits a lower temperature of precipitation and more rapid coarsening in the former.

Small pieces of a natural uranium fuel bar, after irradiation at below 300°C, were heated to higher temperatures under various pressures. This heat-treatment produced changes in both density and micro-structure. Heating in vacuum produced an increase in volume (i.e. swelling of about 2 per cent at 575°C, 4 per cent at 810°C, and over 20 per cent above 1000°C. Repeated thermal cycling, and reheating at low pressure after first heating at high pressure, produced swelling of 10 to 20 per cent at 810–835°C. Examination of the swollen metal with optical and electron microscopes revealed bubbles between 10−2 and 5 × 10−6 cm in diameter and also cracks; the latter were often associated with non-metallic inclusions and were particularly prevalent in thermally cycled specimens. The conclusion is drawn that the swelling is caused by the separation of the fission gases, krypton and xenon, to form gas pockets in the metal, and the mechanism of swelling is discussed.

The internal friction of copper single crystals remains unaffected by a short neutron bombardment at -195°c if no warning is allowed before measurement. Successive pulse anneals produce no marked change until 23°c is reached, when, the internal friction is greatly reduced.

Sandwiches of copper and nickel have been diffused under various hydrostatic pressures. This reduces the number of voids and above a critical pressure there are no voids. Similarly, voids previously formed by an anneal in vacuo are eliminated by applying pressure during a subsequent heat treatment, but the critical pressure is greater than that necessary to prevent them forming. It is concluded that the vacancy supersaturation is about 1 per cent and that the voids nucleate heterogeneously upon nuclei (about 10−5 cm in size) which are included in the weld between the two metals. There is no evidence of an alteration in the diffusion coefficient of copper into nickel with the pressures used.

Natural thorianite, containing large amounts of helium from the radioactive decay processes occurring in it, fragments explosively at 950°C when slowly heated. This behaviour is attributed to the helium precipitating as high-pressure gas at flaws in the mineral.

Gas bubbles produced internally in polymethyl methacrylate by heating or by irradiation have been studied. In the centre of a sample gas bubbles form after an incubation period and then grow, whereas in the surface layers, from which gas is evolved, no bubbles form. Pressure applied during the heating of samples can prevent the gas formed from precipitating. Samples in which the gas has been produced by heating contain a number of bubbles per unit volume which is a constant, but in irradiated samples this number varies depending upon the flux or the subsequent heating. The results are used to discuss the physics of the nucleation of gas bubbles in solids, and in particular the swelling which occurs in uranium upon irradiation with neutrons.

The nonlinearity of the resistivity/integrated flux curve obtained by Cooper, Koehler , and Marx (1955) during the deuteron bombardment of copper wires at 10°K is considered to be due to a bombardment induced annealing process quite distinct from the thermal annealing which is observed at higher temperatures. Three possible causes of this irradiation annealing are discussed. The local temperature rise around the track of the incident deuteron (the “thermal spike”) is of insufficient extent and duration to permit the observed amount of annealing. The local melting which might occur in displacement spikes also seems insufficient. If, however, an individual knock-on at the end of its track travels, not by itself moving from one interstitial position to the next, but by taking the lattice position of a neighbouring atom, forcing that atom into the neighbouring interstitial position, then the interstitial atom could effectively travel many atomic distances at low speeds, because the energy lost during such movement should be low. The result of such an interstitial atom passing within a critical distance of a stationary vacancy would be mutual annihilation, and a decrease in the resistivity. If this critical distance is three atomic spacings then it is sufficient that the interstitial atom travels about eighteen atomic spacings in the manner described to give the amount of annealing observed in the experiments.

The climb of edge dislocations in face-centred cubic crystals is described in terms of the atomic movements involved. The conclusion is reached that in a suitable environment such a dislocation will climb even when it is dissociated, although then it will climb less readily. A straight dislocation line will-be unable to climb when the departure from the equilibrium number of atomic defects is small, but jogs in the dislocation line will enable the line to climb even under these conditions. Dislocations will climb either to enlarge or reduce the area of the extra {110} planes of atoms which lie perpendicular to the Burgers vector, and with certain special dislocation configurations many parallel adjacent {110} planes of atoms can be either eliminated from, or inserted into, the crystal by a spiral mechanism. By examining the free-surface markings on crystals grown from the vapour it should be possible to distinguish surface steps caused by the climb of dislocations from those due to slip, and also to determine whether it is vacancies or interstitial atoms which are precipitated on cooling such a crystal.