Shock Waves and High-Strain-Rate Phenomena in Metals

Research on the reaction of metals to explosions and impacts has been largely stimulated and guided by practical problems. French militarists as early as the 1830’s were fragmenting hollow cannon balls under controlled conditions to establish optimum loading and material properties of the casings. Shortly before World War I, the British engineer Hopkinson made detailed observations on spalling of metals. In the 1930’s, the possibilities of using high explosives to form and project missiles was recognized. This led to the development of the metal-lined shaped charges used so effectively in World War II and later in the recovery of oil, and to the method used to trigger atomic explosions. Development of sophisticated electronic missile fuzing during and following the war emphasized the need for equally sophisticated fragmentation control, a field that has since occupied the attention of many engineers. The non-military use of explosives to work and deform metals started and began to flourish in the 1950’s with many small industrial operations springing up. The extensive engineering developments have heretofore been accompanied by only limited basic research efforts but recently these have greatly expanded in a concerted attempt to understand metal behavior under rapidly applied intense loading.

Since its introduction nearly two decades ago, the expanding ring test has shown considerable promise as a simple method of obtaining strain-rate-sensitive uniaxial material property data. The procedure is to monitor the kinematics of a uniformly expanding ring. The stress-strain-strain rate response of the ring material can then be calculated from the ring equation of motion and the recorded data.

Low strain-rate tensile test data (10−5 to 10−2 s−1) were obtained for 6061 aluminum alloy in the -T6- and -0- condition. Little, if any, strain-rate effect was observed over this range of strain rates. Thirteen expanding ring tests were performed on 1-in.- and 2-in.-diam 6061-T6 and -0- temper rings with maximum strain rates of 3000 to 23000 s−1. The flow stress at 2% circumferential expansion was determined from all of the tests and is shown to be dependent upon maximum strain rate. Furthermore, a size effect is observed, since flow stress at 2% strain and equal strain rates is greater in 2-in. rings than in 1-in. rings. This size effect may be due to shock hardening, but a definite conclusion has not yet been reached. The -T6- temper rings also show a size effect, but the difference between the 1- and 2-in. rings was not the same for all values of flow stress.

The strength and the ductility of an austenitic, a quenched and tempered and a maraging steel were investigated under tension over a wide range of strain rates till 5000 1/s using hydraulic, pendulum and fly wheel machines.

The collision of two hemishells is analyzed where one hemishell is clamped and the other impinges on the inner surface of the clamped hemishell. A model of rigid-perfectly plastic behavior is proposed for the flow of material in the region of contact. An expression relating velocity and deformation is obtained and compared with the results of experiments.

Sinter alloys containing a high percentage of tungsten (80 – 97.6%) are used because of their high density and great ductility. The deformation behavior of such materials has been investigated under hypervelocity impact conditions. Several types of projectiles have been fired into semi-infinite steel targets at impact velocities between 1000 m/s and 2100 m/s. A special shape factor has been defined in order to characterize the amount of deformation of individual tungsten grains within respective nickel-iron matrices. Within residual projectiles, the shape factors have been measured as a function of the distance from the stagnation point and impact velocity.

Sheet specimens of 304 stainless steel were tested in uniaxial tension over a temperature range of −173 to +100°C at low strain rate (10−3/s) and high strain rate (103/s). Two shook loading experiments were conducted; one at −105°C and one at ambient temperature. The amount of strain-induced α′ martensite was measured magnetically, the temperature increase during deformation was measured with a thermocouple, and the substructure was examined by transmission electron microscopy. At room temperature and small strain levels, more α′ martensite was produced at high rate than at low rate. At strains >0.25, low-rate deformation produced significantly more α′ martensite. Tests over a range of temperatures showed that the decrease in α′ at large strains and high strain rates resulted from an increase in temperature due to adiabatic heating. Micro structurally, we found that α′ nucleates at shear band intersections. At high rate, there are more shear bands, but the growth of the α′ embryos is restricted by the temperature increase. During shock loading, α′ martensite formed at both test temperatures, thereby demonstrating that transformation is possible at such high rates, and under the predominantly compressive stress state.

Observations have been performed on fragments recovered after the dynamic expansion and rupture of relatively thin metallic shells under the action of explosives. Two main types of rupture have been observed: a very ductile one, sometimes by chisel-point (e.g. in copper or aluminum alloys) and a brittle one associated with adiabatic shear bands (in uranium and titanium alloys). For these two types of rupture we have reviewed the different theories proposed to take into account the experimental results concerning the rupture elongation.

Dynamic fracture and quasi-static fracture have traditionally been studied by different communities, and correspondingly different terminologies have evolved. Nevertheless, the basic phenomenology is the same for all loading rates: nucleation, growth, and coalescence of microscopic voids or cracks. Successful prediction of the degree of micro structural failure produced by a given loading history thus requires an understanding of the kinetics of the microstructural failure processes. To attain this understanding will require the combination of three separate capabilities: (1) experiments with controlled stress and strain amplitudes and durations followed by posttest quantitative characterization of micro structural damage, (2) analytical models of the microstructural void or crack kinetics, and (3) computer codes to exercise the resulting, highly nonlinear, constitutive relations in numerical simulations of experiments that produce varying degrees of microstructural damage. This paper reviews the current status in developing and combining the above three capabilities. Some speculations are also made regarding future directions in linking continuum fracture behavior to microstructural material properties.

Internal fractures in spherical, paraboloidal, hyperboloidal and elliptical solids subjected to concentrated surface explosive loading are investigated. The initiation and growth of the internal fractures are associated with the focusing effect of P and S wave fronts reflected from the curved boundaries. High speed photography on perspex specimen is used for the verification of predictions of simple methods of geometrical optics and propagation of discontinuities. The speed with which these internal fractures grow are found to depend upon the geometry of the surface of the solid.

An analysis of fragmentation due to impulsive stress loading of solid materials is developed which results in analytic expressions for distributions in fragment sizes. The analysis is restricted to a linear (one-dimensional) distribution of material which is loaded uniformly in tension until fracture, and ultimately fragmentation, occurs. Concepts of survival statistics consistent with simple physical laws governing the fracture process are used to account for the spatial and temporal distribution in fracture nucleation sites. Analytic fragment distribution curves for ductile fracture are derived and found to provide a good representation of data obtained from impulsive fragmentation studies on aluminum rings.

Paper describes an experimental investigation of crack initiation at high loading rates. Three metals were tested. The material was subjected to various types of preloading, namely, uni-axial as well as bi-axial tension beyond the yield point and cyclic loading (fatigue). The specimen was made in the form of a thin-walled tube, loaded with interior pressure in a split Hopkinson bar of special design. Crack initiation data were evaluated, using linear and non-linear fracture mechanics (K_{C dyn} and J_{C dyn}). Results indicate that J_{C dyn} may be used in an initiation criterion. J_{C dyn} is reduced by pre-loading and it is smaller than the corresponding J_{C static} for the materials tested.

This paper briefly reviews the mechanics of the shear-control method, and describes the controlled failure processes which occur when this method is used in explosively loaded cylinders. Details of the failure processes were obtained by relating high-speed photographic studies of the cylinder expansion process to postmortem studies of the fragments using optical microscopy. The effects of various design parameters are discussed in terms of control effectiveness and the failure processes.

The strain localization phenomenon of “adiabatic shear” is generally attributed to a plastic instability arising from a thermal softening effect during adiabatic or near-adiabatic plastic deformation. High strain-rate adiabatic torsion tests indicate that the effective shear stress-strain $$(\bar \tau - \bar \gamma ) $$ relations for high-strength (rate-insensitive) steels can be described by a simple expression of the form: $$\bar \tau = {{\bar \tau }_0}(1 + \alpha \bar \gamma )\,\exp \,( - \beta \bar \gamma )$$ where $${{\bar \tau }_0}$$ is a constant, α and β are dimensionless hardening and softening parameters. The flow stress reaches a maximum at an instability strain, $${{\bar \gamma }_i} = {\beta ^{ - 1}} - {\alpha ^{ - 1}}$$. With parameters derived from the torsion tests, this relation has been used in computer simulations of the development of intense shear localization in a simple uniformly loaded body. Strain localization has been studied under conditions of both quasistatic and dynamic deformation. Application to the simulation of ballistic penetration is in progress.

Metallurgical effects on shear band behavior were studied by applying a state of uniform dynamic shear strain and then suddenly removing it by explosively expanding a hollow specimen cylinder to a diameter determined by a massive confining cylinder. Because geometry-induced stress concentrations are absent and because shear band development can be stopped at various stages, the procedure emphasizes the role of metallurgical features on shear band activity.

Optical Metallographic studies on the specimens deformed at different temperatures (Room temperature to 900°C) and mean strain rates of 0.015, 0.6, 11.0 and 470 s−1 using compression testing machine, double action hydraulic press, friction screw press and high velocity gravity drop hammer respectively were carried out. The specimens were heated to the required temperatures by means of a tubular furnace included in the tool set-up itself.

A criterion for thermo-plastic shear instability which includes heat transfer is derived from a system of equations describing plastic deformation, the first law of thermodynamics and Fourierr’s heat transfer rule.

Impact studies were carried out on several steels using flat- nosed projectiles. An air gun was employed to produce projectile velocities in the range of 100–250 feet per second. The target structures were studied using microhardness measurements and optical microscopy. Deformed shear bands were found in most steels and ferrous alloys. In most of the ferritic alloys studied, these deformed bands act as “precursor bands” to transformed bands which frequently follow when the strain concentration is sufficiently great or the strain rate sufficiently high.

Experiments were conducted to determine the shear strains and strain rates in adiabatic shear bands in a Ni-Cr steel. An explosively driven punch was used to shear plugs from steel plates and, thereby, create adiabatic shear bands for investigation. There were originally planes of chemical inhomogeneity (reference bands) in the plane of the steel samples. The slopes of the reference bands were altered by the plastic shear, and they were, therefore, a measure of the plastic shear strain. From measurements of the slopes of the reference bands, it was found that the plastic shear was approximately an exponential function of distance through an adiabatic shear band. Shear strains of at least 572 developed after the onset of adiabatic shear, and with measured rates of punching, it was estimated that shear strain rates as great as 9.4 x 107 sec−1 were attained within the adiabatic shear bands. The results suggest how heating occurs within the shear bands and that there is a plastic instability due to carbide and lath morphology changes and, possibly, a magnetic transformation associated with the deformation and heating.

A heat sensitive film technique has been employed to give spatial resolution to the observation of “hot spots” in the impact deformation of soft ionic or molecular crystals and in crystal-filled polymer composites. The temperature of these “hot spots” is estimated to exceed 250°C for even modest impact loads. The heating is interpreted to depend on local shear deformation.

The change of microstructure is a common process in impact-loaded materials. Around the path of projectile, for example, cracks and white streaks are visible. From the point of view of protection this deformation behavior of an armor plate must be considered an undesired process. In order to investigate these streaks a specimen was designed by which it is possible - using a Hopkinson bar - to produce white streaks in a laboratory with the advantage of good reproducibility. With different materials load versus time measurements using strain gauges were carried out. Comparing these results with those of the examination of the streaks by light and electron microscopy methods a relationship could be given between the load-time-diagram and the correspondent microstructure of the streaks. Furthermore the effects of material parameters on the intensity of deformation and the dimensions of the streaks are reported.

We present an elementary introduction to the art and science of uniaxial strain-shock recovery experiments. Subjects discussed include generation of planar stress waves, design of sample recovery experiments, stress-gage instrumentation, and temperature effects. The emphasis of the present paper is practical; we hope to provide the neophyte with the basic information needed to design and interpret well-characterized shock recovery experiments.

Solid samples have been routinely recovered for examination after having been subjected to high pressure shock loading. Such investigations have revealed many unique and interesting defect features and are essential if a detailed understanding of shock deformation processes is to be achieved. Nevertheless, examination of samples hours or days after they are compressed for only a few microseconds in a loading whose rise time may be subnanosecond fails to address the relationship between the residual defect structure and that existing during the loading. Electrical probes, and to a lesser extent optical probes, have provided reasonably direct measurements of defect states and some limited information on the evolution of these states. For example, measurements of the electrical resistance of metals provide an indication of vacancy concentrations. Similarly, measurements of shock-induced electrical polarization in insulating solids have provided evidence that large numbers of point defects are generated and displaced by the stress and velocity gradients within the shock fronts. Optical measurements of shock-induced bleaching of color centers in NaCl have provided some evidence for kinetics of the formation of higher-order point defects.

This paper presents the solution of the problem of the pressure field appearing in a metallic plate under the action of explosive load propagating over its surface with a subsonic velocity. The solutions are compared with the experimental results obtained by manganin gauges. The experimental methods are briefly described.

In this investigation, Y-cut quartz crystals are used as pressure-shear generators, and X-cut quartz and 6061-T6 aluminum are used as test specimens. The coupled longitudinal and shear motion generated within the Y-cut crystal can be transmitted into a test specimen bonded to the rear surface of the crystal, and monitored using velocity interferometer techniques. Experiments on X-cut quartz and 6061-T6 aluminum demonstrate applicability of the technique. Results obtained in this investigation are compared with numerical solutions obtained using the finite difference wave propagation code TOODY.

The strength of 6061-T6 aluminum was assessed over the stress range of 8–40 GPa using velocity interferometry to measure reloading and unloading profiles from the initial shocked state. These results show that the shear stress which can be supported in the shocked state increases by about a factor of five over this range. This observation is in agreement with previous investigations. However, an important new observation is that a substantial increase in shear stress occurs during reloading, resulting in a well-defined elastic precursor. This result indicates a significant departure from the elastic-plastic model and suggests that softening occurs during initial shock compression.

The attenuation of planar shock waves generated by plate impact was monitored by their decay throughout massive nickel blocks. This was accomplished, during the passage of the wave, by manganin piezoresistive gages connected to oscilloscopes and, in the post-shocked condition, by hardness measurements and TEM observations at various distances from the impact surface in the nickel blocks. The nickel systems exhibited different metallurgical microstructures before shock loading: preshocked (grain size 150 μm), annealed (grain size 150 μm)and annealed (grain size 32 μm). For each system two different initial shock pressures were used: 10 and 25 GPa. The pulse duration was held constant at 2 μs. The experimental records of oscilloscopes showed that there are no significant effects of grain size and pre-deformation on the attenuation in nickel. The observed attenuation was compared with the calculated one according to hydrodynamic theory and poor agreement was found, An “accumulation” model based on the conservation of energy is presented herein to explain the dissipative processes of shock waves in metals.

A pressure pulse of several kilobar amplitude and a fraction of a microsecond duration is required for experimental studies of dislocations moving in metal crystals at velocities of the order of the velocity of elastic waves. A technique for generating such pulses by means of an exploding aluminum foil is described. The aluminum foil is rapidly vaporized by a capacitor discharge while it is confined between elastic plates.

Experimental and theoretical results on isolated moving dislocations are used as the basis to deduce group behavior of dislocations within and just behind shock fronts. For strong shocks it is concluded that the shock front consists of a Smith interface of supersonic dislocations with a narrow zone immediately behind it of high density subsonic moving dislocations. The Smith interface does not exist in weaker shocks. The dislocation density increases as the square of the shock wave pressure for weaker shocks and becomes almost independent of shock pressure for very strong shocks.

A theoretical framework elucidating the generation of shock-induced defects is presented. Three different mechanisms responsible for the generation of point, line, and planar defects (twins and stacking faults), respectively, are described. Dislocations are homogeneously nucleated at or slightly behind the shock front by the powerful deviatoric stresses generated by the shock pulse; they are accelerated by the residual deviatoric stresses either towards or away from the front. Dislocation dynamical considerations limit their velocity to the velocity of shear waves in the medium. Their self-energy and stiffness are very high at the high velocities; hence, their ability to generate point defects upon intersecting each other is greatly enhanced, because the drag stress produced by the jogs is essentially independent of the velocity.

Shook waves in a crystalline solid cause intensive plastic deformation. Some models describing a strong shock wave in a crystal and peculiarities of generation and development of shears are analyzed. The results of measurements during the process of shock loading, investigations of structural changes, and computer modelling, make it possible to estimate the reality of existing concepts.

The effect of flat plate impact on pure (99.99%) aluminum is studied and compared to quasistatic deformation $$\left( {\dot \varepsilon = {{10}^{ - 3}}{s^{ - 1}}} \right)$$. Actual pressures range within 0.5 to 10 GPa. The major part of impacts were made at a pulse duration of 2 µs. No significant effect of increasing duration to 4 µs is observed. Post-shock dislocation density, -morphology and yield stress is obtained from eight pressures. The pressure is converted to plastic strain to amenable comparison with q.s. straining. A plot of yield stress versus square root of dislocation density reveals a friction stress in the shocked material. The influence of dislocation structure on flow stress is discussed. A quasiphenomenological model for dislocation net generation: $$d\rho /d\varepsilon = U - \Omega \rho $$ which has shown to be applicable to q.s. conditions for a number of single-phase metals is tested on shocked as well as q.s. deformed aluminum. The rate of generation of dislocations, U, is found to exceed that from q.s. compression at ambient temperature, while the rate of dynamic recovery, Ω, is minor or of the same magnitude.

The stationary shock wave structure in metals considered as cold plasma with degenerated electron gas is studied using the same approach as for the collisionless hot plasma. The shock wave parameters are derived from the cold ion-fluid equations, Poisson equation and Thomas-Fermi equation for electron density. The experiments to show the effects of the present approach are also discussed.

Response of MAR-M200 to shook loading shows that (a) the amplitude of the elastic precursor decays from 2.8 GPa at zero depth to .92 GPa at a depth of 7.8mm; (b) above the appropriate depth dependent Hugoniot Elastic Limit (EEL) the response of MAR-M200 shows a substantial reduction in its shear strength; and (c) the unloading of MAR-M200 from its shock compressed state appears to be elastic.

In this investigation it is shown that the hardening of shock-loaded metals can be studied using the technique of positron annihilation. The concentration of dislocations can also be satisfactorily determined.

The switching of a magnetic field while plastically deforming a zinc crystal causes a change in stress which is associated with electron-dislocation interactions. In this paper the dependence of this change in stress on strain-rate, field strength, and temperature will be presented. In addition, the connection between electron-dislocation interactions, dislocation velocities, and the occurrence of twinning in zinc will be established.

This chapter reviews the results of a large number of shook loading experiments utilizing a common design and a constant shock pulse duration of 2 µs. It is shown that dislocations, dislocation cells, planar dislocation arrays, stacking faults, twins, twin faults, and point defects all contribute in specific systems to residual shock strengthening. Shock-induced microstructures are determined primarily by the stacking fault free energy. High stacking-fault free energy metals and alloys are characterized by dislocation cell structures while low stacking-fault free energy metals and alloys (with the fcc structure) are characterized by planar dislocation arrays, stacking faults and twins in {111} planes. High stacking-fault free energy metals and alloys also twin according to critical shear stress criteria, and the (001) orientation is the initial orientation where twinning occurs. Residual shock microstructures and specific lattice defects induced by the peak shock pressure are shown to be related to resisdual hardness and engineering yield stress. Body-centered cubic metals and alloys are characterized by irregular dislocation arrays as a result of the more numerous slip planes, although twinning also occurs in bcc metals.

A high-energy, pulsed laser beam combined with suitable transparent overlays occn generate pressure pulses of up to 6 to 10 GPa on the surface of a metal. The propogation of these pressure pulses into the metal in the form of a shock wave produces changes in the materials micro structure and properties similar to those produced by shock waves caused in other ways. This paper reviews the mechanism of shock wave formation, calculations for predicting the pressure pulse shape and amplitude, in-depth microstructural changes and the property changes observed in metals. These property changes include increases in hardness, tensile strength and fatigue life. The increases in fatigue life appear to result from significant residual surface stresses introduced by the shock process.

The residual defect substructures developed in Cu-8.7Ge as a result of carefully controlled plate-impact experiments have been characterized and interpreted in terms of the mechanisms and kinetics of plastic deformation. In particular, the effects of a wide range of pulse durations at a constant pulse amplitude of 20GPa have been studied. Twinning has been found to contribute significantly to the plastic strain, as well as to the shock hardening. A maximum in hardening caused by bundles of extremely thin twins occurs for pulse durations near 70nsec. The observed plastic behavior can be accounted for by conventional dislocation dynamics without resorting to the concepts of supersonic dislocations or homogeneous nucleation of dislocations. It is also concluded that dislocation generation rates play a primary role in the plastic response of a material subjected to shock conditions and further, that short duration shock pulses constitute an important tool to study not only these generation rates, but the processes of plastic flow in general.

Shock experiments with α-phase Fe/30%Ni have been done at temperatures from 20°C to 400°C using the W.S.U. gas gun. Hugoniot elastic limits were found to be independent of initial temperature up to at least 275°C with mean value of 9.05 kbars. Amplitude of the phase transition wave varied with temperature in a manner similar to that reported by Rohde (1). Transition pressure at room temperature (uncorrected for shear) was 78 kbars, compared to 71 kbars reported by Rohde and 82 kbars reported by Loree et al. (2). At 400°C the wave profile consisted of two waves, not three. Amplitude of the first was approximately 5.5 kbars. Separation between first and second waves supports the hypothesis that the first is the transition wave, the second is the final wave3 and the elastic wave does not form because the transition collapsed the deviators. This must be regarded as tentative because of the poor quality of records at 400°C. With one exception recovered specimens showed no increase in austenite content for initial temperatures less than 200°C. Between 200°C and 400°C retained γ increased linearly. This contrasts with results reported by Rohde et al. (3) who found significant increases even below the transition pressure and at room temperature.

304 stainless steel and nickel sheet with grain sizes ranging from 14 – 93 μm were simultaneously shock loaded in sandwich arrays at pressures and pulse durations of 15 GPa, 2 μs, 30 GPa, 2 μs, and 30 GPa, 6 μs. Identical sandwiches were shock-loaded repeatedly at 25 GPa, 2 μs (two and three times) and at 30 GPa, 3 μs (twice). The residual hardness and tensile properties were measured and correlated with the microstructures observed by transmission electron microscopy. The effects of repeated shock loading were not observed to be additive nor to involve simple multiples of parameters characterizing either the structure or mechanical properties, or the shock loading parameters (pressure and pulse duration). The volume fraction of defamation twins increased in both materials for repeated shock loading at short pulse durations (2–3 μs), however, twinning occurred only in (001) oriented grains in the nickel at 30 GPa peak pressure, and were absent at 30 GPa, 6 μs. Correspondingly; the interwin spacing decreased with repeated loading. The volume fraction of martensite (α′) also increased in 304 stainless steel with repeated loading, and at increasing peak pressure and pulse duration. Dislocation cell sizes in nickel also decreased with repeated loading but the most prominent feature at lower pressure (15 GPa, 2 μs) was a pronounced thickening of the cell walls with repeated loading.

The residual microstructures and microstructural histories for Ni, Nig₈₀Cr₂₀ (Chromel A), Inconel 600, type 304 stainless steely and Mo, representing a sampling of fee and bcc metals and alloys (having a range of stacking-fault free energies ranging from roughly 21 to 128 mJ/m2 for the fcc materials) are described in detail for a range of peak pressures (8 to 120 GPa) and shook pulse durations (0.5 to 14 μs). Changes in the residual microstructures, which range from stacking faults, twins, and related planar arrays for low stacking-fault free energy fcc materials to dislocation cell structures for high stacking-fault free energy fee materials, are described and quantitatively related to residual mechanical properties (e.g., hardness and yield strength). The production of martensite and the mechanism of α′ - martensite formation are described in relation to shock pulse duration and repeated shock loading. The effect of repeated loading on twin production as well as the effect of shock pulse duration are also described for shock-loaded nickel and type 304 stainless steel. Shock pulse duration seems to have little effect on residual microstructure and hardness above about 1 μs pulse duration. These effects are illustrated in experimental summaries of a large spectrum of shocks-loaded data.

Solution annealed Cu(1.5 wt & Fe) alloys were precipitation annealed at 650°C and 750°C to produce size controlled dispersions of non-magnetic fccFe spherical precipitates. Discs of these alloys were then impacted in oriented magnetic fields at levels up to 5 GPa and then magnetically characterized after recovery. The impact transforms the fccFe to magnetic bccFe. Substructure in the precipitates before and after shock was characterized. Specimens with particle size distribution of 20–35 nm, 40–60 nm, and 80–110 nm having respectively single domain (SD) and superparamagnetic (SP), SD and small multidomain (MD), and MD particles were studied. Shock induced magnetic anisotropy was characterized by magnetic hysteresis loops and the stability of saturation remanence to demagnetization. The shock direction was located in the disk by measurement of angular variation of magnetic-coercivity. Magnetization reversal is unique because of the near ideal magnetic properties of fine Fe particles having a narrow size distribution, a mean interparticle spacing dependent on size distribution, no clusters or agglomerates of any kind, the same substructure in all Fe particles and no significant variation in particle shape. Therefore saturation magnetization reversal could be accomplished by application of a field equivalent to H_R (the remanent coercive force) in the opposite direction.

The structure and properties of copper containing from 1 to 5.6 volume fraction of A1₂0₃ after loading by plane shock waves are investigated. Pressure in the shock waves ranged from 175 to 350 Kb. Under loading the strengthening degree increases with pressure and decreases under equal loading pressures with increasing a volume fraction of a strengthening phase. The structure formed under the shock loading in large measure is determined by a volume concentration of a strengthening phase. The activation energy of the property recovery process is determined, and its connection with the structure is found.

Research efforts assessing the potential of shock TMP for a number of alloy systems are reviewed. Shock loading seems to be a promising deformation technique in TMP when (a) the initial strength of the alloy is such that conventional deformation is precluded and (b) when the shock wave induces property improvements that are significantly superior to those of conventional deformation.

The analysis of the applied and theoretical research concerning the influence of explosion upon the materials testifies to the fact that methods of explosive loading which earlier have been used only as an experimental method of high pressures physics are turned out into a perspective technological process in machinery and metallurgy industries.

The effect of an oblique reflection of the shock waves accompanying the shock compaction of porous metals is considered. Inhomogeneities are established to appear in the form of cavities and melts at irregular reflection regimes in compacted media near the reflecting surfaces. A formation mechanism of these inhomogeneities is shown to be connected with changes of aggregation state of that medium part which was compacted by Mach shock wave.

Present state of knowledge on “the direct electrical discharge compaction of metal powders” is reviewed and the response of a powder column subject to high voltage electrical discharge is described in terms of the electrical, physical, morphological and compositional characteristics of the metal powders.

This work investigates the formation peculiarities of the structure of powdered materials dependent on loading conditions. The samples were loaded by plane shock waves, according to the scheme with shock collision and with the acceleration of powdered-material plates at an angle to one another. It is shown that the microstructure is formed under the particle deformation in the process of compaction. It depends on the shock wave intensity, its duration and physical and mechanical properties of particles. The microstructure is formed as a result of shock wave interaction.

The microstructure at the half-radius position of a polycrystalline alumina rod formed by explosive compaction has been studied by transmission electron microscopy. The as-compacted lattice is composed of differently misoriented bands aligned predominantly in one direction. Such bands may correspond to frequently observed shock lamellae. The band edges are defined by dislocation arrays lying on the basal planes of the hexagonal alumina lattice. The dislocations have Burgers vectors of the type $$\left\{ {11\bar 20} \right\}and\left\{ {10\bar 10} \right\}$$, which are the Burgers vectors of slip dislocations in the basal plane. Basal plane twinning is also observed, and the twin boundaries are found to contain interfacial dislocations. While dislocation generation occurs primarily on basal planes, some dislocation activity is also noted on prism, $$\left\{ {1\bar 100} \right\}$$, and on rhombohedral, $$\left\{ {1\bar 101} \right\}$$, planes. Nonbasal twinning, however, has not been detected. The lattice damage is discussed in terms of possible dislocation and twinning mechanisms.

The discrete shook waves produced by the impact of a high speed punch can compact powders in a time of a few microseconds. Under optimised conditions, this rate of compaction allows inter-particle melting and welding, resulting in a strong compact with a strength that is equivalent to that of a wrought solid.

Scientific and technical problems related to the explosive welding and hardening of metals are presented. Fundamental physical principles, applied researches as well as industrial applications of the processes mentioned are discussed for each item separately. Design and construction of special equipment (explosive chambers) are dealt with as well. Specific lines of application of these technological processes in the national economy of the U.S.S.R. are also paid attention to and are illustrated with concrete examples.

A general review of explosion welding with particular emphasis on metallurgical effects is presented. Explosion welding is basically a solid-phase welding process, in which explosives are used to accelerate the parts to be joined into a high velocity oblique collisions, is fundamental to the welding process. The jetting action coupled with the collapsing flyer plate produce a unique weld interface, with very interesting metallurgical properties. The relationship of the welding parameters to the metallurgical properties is discussed.

The use of transmission electron microscopy (TEM) to examine the microstructure of the bonding zone in explosive welding is described. Plates of an aluminum copper alloy were heat-treated to produce a well defined microstructure. With such plates a series of claddings with different parameters and conditions were made. The microstructure of the bonding zone was then analysed. From the changes of microstructure within the bonding zone as compared to the microstructure of the material in the original state conclusions are drawn concerning the deformation process and the thermally activated reactions which take place during welding. The findings lead to a more comprehensive understanding of the explosive bonding process, including maximum temperature within the bonding zone, cooling rate, and deformation mechanism during the impact of the plates. It can be shown that bonding is achieved by short time melting followed by extremely rapid solidification and cooling of a very thin layer along the contact interface. A brief discussion on the generalization of the results to include materials with high melting temperature, and the correlation between the microstructure and properties of explosive welds is presented.

The morphology of the interface of aluminum - steel explosion welds obtained under different conditions was studied using different metallurgical techniques and the strength of the welds was correlated with the nature of the interface.

Explosive forming of sheet metal parts including large plastic strains may be done on several steps with intermediate annealing. The optimum conditions regarding the oven temperature, the heating time, the micro structure and the hardness of the recrystallized part depend besides on the degree of cold working on the strain rate, the pressure peak and the pressure duration.

Intensive work was carried out at UMIST on the fabrication of composites by one shot explosive bonding. Up to ten sheets of varying materials were arranged in parallel and the top was subjected to explosive detonation in the conventional explosive welding manner. Velocity measurements of each sheets as well as metallographic examinations of the interface waves, were carried out.

In the decommissioning of nuclear power reactors it is desirable to dismantle the highly radioactive pressure vessel by breaking it down into pieces of managable size. In order to minimize the exposure of the staff involved, the time required for this action should be reduced to a minimum. A method is proposed here which employs explosive technology to break the vessel into preformed fragments in a single pass. Experiments using simplified 1:20 scale vessels are described.