Search

The strain-hardening of α-brasses is examined in terms of the condition for the generation of a new slip band at the head of a pile-up of dislocations against a grain boundary. The flow stress of high-zinc brasses appears to be determined by this condition, but the development of intragranular dislocation cells eventually makes it inappropriate for low-zinc brasses. The effect of strain on the various factors involved in the slip band generation is examined experimentally and theoretically.

The tensile flow stress at 77 K has been studied for super-pure and commercially pure aluminium and for an Al-21/4 wt% Mg alloy in relation to, strain ε, grain size d and dislocation density ρ.It is concluded that a principal contribution to strain hardening, particularly in the alloy, is the long-range stress from accumulated dislocation pile-ups against grain boundaries. In the theoretical model, this is linear with ε and with ρd. There is also hardening due to dislocation intersection and shorter-range interaction, and this depends on ρ1/2 and ε1/2.

The tensile flow stressσε has been studied for super-pure and commercial aluminium and for an Al-21/4% Mg alloy in relation to strain, grain size d, subgrain or cell size and dislocation density ρ.At room temperature and 200°C, the flow stress depends on both ρ1/2 and d-1/2. This is interpreted an evidedce of a requirement inσe that slip can be transmitted across a grain boundary. There is extensive dislocation annihilation during deformation at these two temperatures, particularly in the two aluminiums and particularly in fine grains. The behaviour of the commercial aluminium is unusual. At room temperature, the retained ρ is higher in coarse than in fine grains. However, the flow stress remains higher in the fine grains because of the slip transmission requirement. At 200°C, the dislocation density is so much higher in the coarse grains that the flow stress is also higher.At 400°C, most of the dislocations are annihilated and sharp subgrains form with planar network boundaries. It is proposed that the flow stress is determined by the boundary network and the size of the subgrains rather than by the three-dimensional dislocation network within the subgrains.

The columnar-equiaxed transition in superpure aluminium requires not only the presence of nucleants or crystal fragments, but also fulfilment of the conditions for growth on these particles ahead of the solid interface. Experimental evidence emphasizing these factors is presented.

The surface energy γr of a running crack in thin plates of Al2O3, MgO and glass is deduced from the crack length at which branching. occurs. If the stress on the running crack is low, the surface energy approaches the intrinsic value. But, if the stress on the running crack is raised, γr can become a number of times greater than the intrinsic value, suggesting that plastic deformation accompanies the crack.

It is shown experimentally that nitrogen-looking is appreciably stronger than carbon-locking. This is contrary to the customary assumption and it is suggested that there may be chemical as well as elastic interaction with the dislocations. Boron is found to produce even stronger looking than nitrogen. These variations in the locking strength alter the duetile-brittle transition temperature. The stronger the locking, the higher the transition temperature.

A criterion for the transition in fracture mode from ductile to cleavage is calculated for fracture at a notch. It is concluded that the temperature-dependence of this transition probably arises more from that of the Peierls-Nabarro stress required to move a free dislocation in α-iron than from the temperature-dependence of the locking of e dislocation source. The transition temperature is found to be a function of grain size, the friction on a free dislocation, the strength of the dislocation locking, the degree of triaxiality of the applied stress and the elastic constants.

The expressions obtained in a previous paper for the fracture-mode transition temperature at a notch are shown to agree well with experimental measurements of its variation with grain size and with the friction on a free dislocation. A calculation is given of the effect of the dislocation locking strength on the transition temperature and this is in reasonable agreement with the observed temperatures for steels with weakened locking due to the presence of manganese. It is suggested that strain and strain-ageing raise the transition temperature because of the increased tensile stress available at fracture for the propagation of a dislocation crack.

Alumina is normally regarded as truly brittle below ∼ 1000°C. Contrary to this view, it is argued that plastic deformation at the crack tip plays a necessary part in the fracture. This is used to explain the minimum observed at 250°C in the fracture stress of polycrystalline plates containing drilled cracks and in the stress intensity required for crack-branching. Other details of the fracture of single crystal and polycrystalline alumina, including changes in the proportion of intergranular and transgranular fracture with temperature, crack length and crack-branching, are also explained by the involvement of plastic deformation.

Gibbs' adsorption equation and the Griffith and dislocation theories of fracture are used to calculate this lowering. The possibility that such adsorption causes the embrittlement of iron by hydrogen is discussed and is shown to be in good agreement with measured fracture stresses.

For a number of metallic polycrystalline aggregates, it is shown experimentally that s`f, the flow stress at constant strain, is related to the grain diameter l by [image omitted] where s`0 and k are constants. This has the same form as the relationship between the lower yield point, when this occurs, and grain size. An explanation is given by an extension of Taylor's theory to allow for the resistance at the grain boundary to the formation of a slip band. Dislocation-locking and a small number of slip systems are shown to favour a strong grain-size dependence of s`f and this explains the variation of this dependence amongst the common metals and alloys.

It is shown that the portion of this stress due to interaction with random C and N atoms is temperature-independent. At low temperatures, another, very temperature-dependent portion is dominant, and this is interpreted as evidence of a sizeable Peierls-Nabarro force in α-iron.

Evidence that ductile fracture requires a compressive stress on the dislocation arrays close in value to that required for cleavage is presented. It is shown that the ductile fracture stress s`f is related to the grain diameter l by the equation s`f=s`0+kl-1/2, where s`0 and k are constants. This is the same form of relationship as already found for cleavage and the slope k is practically the same for both types of fracture.

The inhibition of plastic deformation by dislocation locking promotes brittleness. The influence of manganese in iron is studied in detail and it is concluded that the ductility conferred by the manganese is associated with weakened dislocation locking as well as with grain refinement. The weakening of the locking appears to be due to interaction between nitrogen and manganese atoms.