Search

The potential theoretic methods recently developed at NRL for solving the diffusion equation are applied to the problem of lamellar eutectic solidification. This approach leads to a set of coupled nonlinear integro- differential equations for the shape of the solid-liquid interface and the solute concentration on the interface. The general characteristics of the solutions to these equations are discussed, and in particular it is shown that lamellar solutions may not be possible when the ratio of thermal gradient to freezing rate is less than some critical value which depends on the phase properties and the phase fraction. Selected numerical results are presented for the Pb-Sn eutectic system and compared to the results of Jackson and Hunt, who assumed a planar interface model. It is shown that the planar interface model can be a source of considerable error under certain conditions, and that the self-consistent model provides additional theoretical insights.

An analysis is presented which predicts the hammer forces, specimen bending moments, and tear energies, generated during the NRL Dynamic Tear test. The theoretical predictions agree quite well with experimental observations for relatively brittle materials. The tear energy per unit net area calculated from this analysis is shown to vary significantly with initial impact velocity and specimen length for a brittle material. It is relatively independent of the specimen thickness and notch depth to width ratio. (Author)

The importance of dynamic fracture toughness as a controlling influence on the design and integrity of high performance structures is becoming more generally recognized. Since tear test procedures involve dynamic loading, a rigorous analysis of this test method can provide direct translation of test results into structural parameters such as fracture stress and stress intensity values of fracture mechanics in addition to tear energy values. This report concerns the development of a dynamic analysis for the two principal elements in the tear test, namely, the hammer and the specimen. Several methods of analysis for predicting the hammer force-time curves are discussed and an analysis based on simple bending theory which predicts the beam bending moments and stresses is presented. The calculated force-time relationship for the hammer was in good agreement with experimental results. The results indicate that the first peak in the hammer force-time curves is not directly related to fracture but is caused by inertia effects. For the specimen, the calculated stresses and bending moments predicted by the dynamic analysis agree fairly well with experimental values. Both calculated and experimental values of dynamic stresses and moments are lower than those calculated from a static analysis; and therefore, in the dynamic tear test a static analysis cannot be used to calculate fracture stresses or stress intensity factors. (Author)

Conditions for crack initiation as a function of test specimen geometry, test material, testing conditions and internal structure of the material were investigated. Lithium fluoride was used to give information concerning the dislocation moments and arrangement at and near existing cracks. The results include the effects of dislocation arrangements in lithium fluoride single crystals near existing cracks.

The stress field interactions between neighboring stress raisers (simulated flaws; a variety of edge notches, and holes) were investigated under uniaxial tensile load to obtain the basic information required for a fracture analysis of inhomogeneous brittle materials, such as ceramics. For the quantitative analysis of the stress distributions, photoelasticity techniques were used in transmission on Homalite CR-39 specimens. A numerical solution for the stress field of a doubly periodic array of elliptical holes was also obtained.

Crack initiation and propagation under altering loading has been investigated for a number of technical alloys (AISI 301, PH-15-7Mo, D6-AC, HP9-4-25, Ti-6A1-4V). The parameters n(+) and N sub o relating to the crack propagation and initiation respectively have been introduced through the use of a crack propagation law suggested by Weiss and have been correlated with the specimen geometries and material properties. The parameter n(+) is thought to depend in some way on the strain hardening coefficient of the material and is found to provide an indication of the material's resistance to fatigue crack propagation. The parameter N sub o is related to the crack initiation times and a relation between N sub o, the geometry of the specimen, and the parameter n(+) has been proposed. (Author)