Showing papers in "Materials Science and Engineering A-structural Materials Properties Microstructure and Processing in 1993"

Structure and properties of ultrafine-grained materials produced by severe plastic deformation

Abstract: Strain-heat methods of obtaining ultrafine-grained (UFG) metallic materials with grain sizes as small as 20 nm and peculiarities of their structure are considered. It is shown that intercrystalline boundaries are the main element of the structure of UFG materials and that they are typically in a non-equilibrium state. The formation of a special grain boundary phase, i.e. a thin near-boundary layer with high dynamic activity of atoms, has been found. This unusual structure leads to the manifestation of promising new elastic, strength, superplastic, damping and magnetic properties of UFG materials.

Hall-Petch strengthening in nanocrystalline metals

Abstract: The tensile behavior and microhardness of nanocrystalline Cu and Pd were measured as a function of grain size In the case of Cu, an increase in strength with grain refinement continues down to the finest-grain material tested, whereas Pd shows little dependence of strength on grain size in the nanocrystalline regime Both nanocrystalline Cu and Pd are significantly stronger than conventional grain size material A literature search and experiments show that negative Hall-Petch slopes at ultrafine grain sizes observed in some studies are not associated with room temperature creep but rather with the use of a single sample annealed repeatedly to change the grain size

Titanium carbonitride-based cermets: processes and properties

Abstract: Cermet is a structural material in which approximately equiaxed fine grains of ceramic hard phase embed in a matrix of metal or alloy binder. Titanium carbonitride-based cermets were first invented in the 1930s, but the boom of the cutting grades really started in the early 1970s when titanium carbide-based cermets were established. However, because of their superior properties, the Ti(C,N)-based cermets are now in a process of replacing the TiC-based cermets for cutting tool applications. In traditional titanium carbonitride-based cermets, polybdenum is regarded as the indispensable ingredient for wettability and sinterability, at the expense of grinding machinability. With the recent invention of the pre-sintering solid-solution treatment of the ceramic hard phase, the materials development of titanium carbonitride cermets has come to a new stage where molybdenum is no longer indispensable. For cermets of very high nitrogen content, however, the combination of solid-solution treatment and a moderate molybdenum addition is predicted to be the way out.

Wear regimes and transitions in Al2O3 particulate-reinforced aluminum alloys

Abstract: The effect of the applied load on the unlubricated sliding wear behaviour of a 6061 Al alloy reinforced with 20 vol.% Al2O3 particles was studied. Experiments were performed using a block-on-ring type wear rig. Wear of the control alloy, i.e. the unreinforced 6061 Al, and the wear of the counterface (AISI 52100 steel) were also studied. Three wear rate regimes were observed in the composite: in region I, i.e. at low loads (less than 10 N) the wear rates were less than 2 × 10−5 mm3 m−1. This was followed by a transition region where the rates increased by a factor of 102. In region II that covered mid-range loads the wear rates raised steadily, 10−3 to 10−2 mm3 m−1, up to 230 N where a second transition took place to a severe wear regime (region III). In the unreinforced 6061 Al, only regions II and III were observed. At low loads the wear resistance of the composite (region I) was two orders of magnitude higher than that of the unreinforced alloy. In region II there was no significant difference between the wear rates of the unreinforced and the Al2O3-reinforced alloys. However, the transition from region II to III occurred at a lower load in the unreinforced 6061 Al (60 N). Metallographic studies performed to delineate the rate controlling wear mechanisms revealed that low wear rates in region I resulted from the load-bearing capacity of Al2O3 particles and the formation of transfer layers on the contact surfaces of the composites. When the applied load exceeded the fracture strength of the particles, particles at the surface were fractured and wear occurred by a process of subsurface crack growth. Al2O3 particles promoted crack nucleation and growth and acted as third-body abrasives resulting in wear rates similar to those in the unreinforced 6061 Al. The transition to severe wear rate regime is shown to be controlled by frictional heating to a critical temperature. This temperature was higher for the composite material for which the transition was postponed to higher loads.

Synthesis and properties of nanophase materials

Abstract: Nanophase materials, with their grain sizes or phase dimensions in the nanometer size regime, are now being produced by a wide variety of synthesis and processing methods. The interest in these new ultrafine-grained materials results primarily from the special nature of their various physical, chemical, and mechanical properties and the possibilities of controlling these properties during the synthesis and subsequent processing procedures. Since it is now becoming increasingly apparent that their properties can be engineered effectively during synthesis and processing, and that they can also be produced in quantity, nanophase materials should have considerable potential for technological development in a variety of applications. Some of the recent research on nanophase materials related to their synthesis and properties is briefly reviewed and the future potential of these new materials is considered.

Principles of particle selection for dispersion-strengthened copper

Abstract: A new fundamental approach to the design of high strength, high thermal conductivity dispersion-strengthened copper alloys for applications in actively cooled structures is developed. This concept is based on a consideration of the basic principles of thermodynamics, kinetics and mechanical properties. The design requirements for these materials include a uniform distribution of fine particles for creep and fatigue resistance, a high thermal conductivity, thermodynamic and chemical stability at temperatures up to 1300 K, a small difference in the coefficients of thermal expansion between the particle and matrix, and low particle coarsening rates at the processing and service temperatures. The theory for creep of dispersion-strengthened metals developed by Rosler and Arzt is used to predict the optimum particle size for a given service temperature and to illustrate the need for a high interfacial energy. Resistance to coarsening leads to a requirement for low diffusivity and solubility of particle constituent elements in the matrix. Based on the needs for a low difference in the coefficients of thermal expansion to minimize thermal-mechanical fatigue damage and low diffusivity and solubility of the constituent elements, several candidate ceramic phases are compared using a weighted property index scheme. The results of this quantitative comparison suggest that CeO2, MgO, CaO and possibly Y2O3 may be good candidates for the dispersed phase in a copper matrix.

Dislocation loops at crack tips: nucleation and growth— an experimental study in silicon

Abstract: The nucleation of dislocation loops at crack tips and the development of the plastic zone were studied in single-crystal silicon samples precracked at room temperature and loaded at T ⩾ 900 K under well-controlled conditions (mode I loading, constant loading rate). Several crystallographic orientations with different cleavage planes and crack front orientations were investigated. In situ observations by synchrotron X-ray topography were supplemented by chemical etching after fracture. Special attention was paid to the early stages of plastic zone formation. Dislocation nucleation appeared to be very heterogeneous along the crack front and may be favoured at free surfaces and cleavage edges. Activated slip systems and dislocation arrangements are discussed. It is shown that considerations based on the crack tip stress field to not suffice to account for the observed slip systems. The ledge crack mechanism of Zhou and Thomson has probably been observed but cannot be proved to be the main emission mechanism.

Statistical distributions of fracture strengths of cast Al7SiMg alloy

Abstract: Al7SiMg alloy vertically cast test bars were produced in dry sand moulds by top filling (TF), turbulent bottom filling (TBF) and turbulence-free bottom filling (TFBF) with filtered metal. The test bars were heat treated prior to tensile testing. Scanning electron microscopy examination of polished sections showed that all the test bars contained tangled networks of oxide films which are seen to constitute cracks within the metal. They are more prevalent in the TF and TBF castings. A large number of anomalous features are observed on all fracture surfaces. These are oxide films produced during filling and entrained in the casting. The best description of the skewed distributions of tensile strengths of each set of castings was obtained with the Weibull distribution. Weibull slopes of 11, 20 and 38 were obtained for the TF, TBF and TFBF castings respectively.

Processing nanocrystalline ceramics for applications in superplasticity

Abstract: The production of nanocrystalline ceramics for subsequent use in superplastic forming operations requires that the ceramics be made in large quantities, with high densities, and under stringent grain growth control. To make large amounts of nanocrystalline starting powders, two popular wet chemical techniques (precipitation from salt solutions and alkoxide hydrolysis) can be used and are described in this paper. Unfortunately, pressureless sintering of these powders does not typically lead to the high densities and ultrafine grain sizes desired in the final product. Sintering data suggest that pore shrinkage occurs only when grains reach a critical size with respect to the pore size; thus, if the ceramic contains large pores, densification can require significant grain growth. Separation of large pores from grain boundaries may also occur and lead to incomplete densification, even at extremely large grain sizes. In all cases the pressureless sintering behavior of the nanocrystalline ceramics appears to adhere to well established theories used to explain the sintering of conventional, larger-grained ceramics. During both pressureless sintering and sinter-forging experiments, the grain size of a nanocrystalline ceramic is identical to the average spacing between open pores in the sample. Pressureless sintering results in the closure of these pinning pores by about 90% density and thus3leads to a substantial grain growth at densities greater than 90%. Sinter-forging, however, often allows one to maintain a stable population of small open pores (for pinning purposes) throughtout sintering, while preferentially eliminating the large pores which detract from the sample density. The deformation regime in which sinter-forging is performed has a decided effect on whether large pores or small pores are eliminated preferentially and, consequently, on whether a high density and fine grain size combination is achieved or not.

Delamination wear in ductile materials containing second phase particles

Abstract: Plastic deformation and damage accumulation below the contact surfaces play an important role in sliding wear of ductile materials. In this study, metallographic techniques have been developed and used to determine the magnitude of the shear strains and the microhardness gradients in near surface regions in an aluminum-7% silicon alloy. Under dry sliding wear conditions, both the magnitude of plastic strains and the depth of heavily deformed zones increased with sliding distance and applied load. The flow stress and the plastic strains in the deformed zones are shown to obey a work hardening law which can be expressed in the form of a Voce type constitutive equation. Scanning electron microscopy (SEM) studies on the longitudinal sections below the worn surfaces indicated that thin flake-shaped debris particles were generated by a process of subsurface delamination occurred via cracks which originated from silicon particles within the deformed zones (but not at the contact surface) and propagated parallel to the surface. A model based on the hypothesis that delamination cracks are formed by the coalescence of voids at a critical depth below the worn surfaces has been proposed. It is shown that the critical depth for maximum rate of damage accumulation is determined by a competition beetween the plastic strain which enhances void growth and the hydrostatic pressure which suppresses it.

Grain boundary design and control for high temperature materials

Abstract: Recent experimental work has been reviewed concerning the effects of grain boundary structure on grain boundary sliding and migration, and other boundary phenomena involved in high temperature plasticity. A basic knowledge of the structure-dependent intergranular phenomena is important for a full understanding of the heterogeneity and characteristic features of high temperature deformation and fracture, because the structural effects of grain boundaries enhance the boundary-induced heterogeneity of deformation and fracture in polycrystals. The grain boundary character distribution (GBCD) has been shown to be important and useful in explaining and controlling high temperature plasticity, superplasticity and fragility in polycrystals. The potential use of grain boundary design and control for high temperature materials has been discussed on the basis of recent work concerning GBCD, texture and other microstructural factors. The recent achievement of grain boundary design and control for high temperature materials has been briefly introduced.

Estimates from atomic models of tension-shear coupling in dislocation nucleation from a crack tip

Abstract: The normal stress distribution across a slip plane has the effect of reducing the critical loading required for dislocation emission from a crack tip. The reduction by normal stresses was found to be very significant for Si, based on properties estimated for it using density functional theory, to be large for Fe as modeled by the embedded atom method (EAM), and to be smaller in AI, Ni and ordered Ni3A1, estimated using the EAM. The general dependence over a wide range on parameters characterizing the tension-shear coupling was also determined. In the context of a Peierls model for dislocation nucleation at a crack tip (J. R. Rice, J. Mech. Phys. Solids, 40 (1992) 239), our approach was to search for onset of the dislocation nucleation instability based on the numerical solution of the system of non-linear integral equations describing an incipient dislocation. The incipient dislocation consists of a distribution of sliding and opening displacements along a slip plane emanating from the crack tip; these displacements are related to the shear and tensile stresses across the slip plane by constitutive relations based on the atomic models mentioned. Results from the atomic models are used to parametrize constitutive relations involving a Frenkel sinusoidal dependence of shear stress on sliding displacement at any fixed opening displacement, and a Rose-Ferrante-Smith universal binding form of dependence of tensile stress on opening displacement at any fixed shear displacement. These relations then enter the system of integral equations, solved numerically, which describe the elasticity solution for a non-uniform distribution of sliding and opening along the slip plane. The results show that tension-shear coupling will often significantly reduce the loading for dislocation emission from the value estimated on the basis of an unstable stacking energy Yu~ determined with neglect of such coupling, in a shear-only type analysis. For the EAM models of the metals considered, a simple and approximate method to account for the tension effects is to use a modified quantity yo iu*~, which is an unstable stacking energy for lattice planes which are constrained to a fixed opening A0*, corresponding to that for vanishing normal stress at the unstable shear equilibrium position. Moreover, it is found that the normal stress effect can be described well in these cases by replacing the unstable stacking energy 7u~ in the shear-only model by a tension softened ~'u~(q~), which depends on the phase angle ~p of the combined tension-shear loading along the slip plane according to the stress intensity factors of the elastic singular solution. The same simple procedures for accounting for tension effects on nucleation are less suitable for lattices with strong coupling such as Si.

Functional interfaces for oxide/oxide composites

Abstract: Potentially useful interfacial coatings for composites of sapphire fibers in polycrystalline aluminum oxide matrices are discussed, subject to the constraints of phase compatibility in high temperature oxidizing environments and low mechanical interface strength ( i.e. to allow pull-out). Results of preliminary tests to investigate feasibility of some candidate coatings ( β -aluminas, LaPO 4 and CaF 2 ) are presented. These systems exhibited interfacial debonding under certain conditions, as required for high toughness composites, although several issues remain to be resolved before they can be used successfully.

A unified solute diffusion model for columnar and equiaxed dendritic alloy solidification

Abstract: A unified solute diffusion model is proposed for columnar and equiaxed dendritic alloy solidification, in which nucleation, growth kinetics and dendrite morphology are taken into account. Various applications to a uniformly solidified system are demonstrated, with emphasis on three special cases: complete solute mixing in the liquid, columnar growth with significant dendrite tip undercooling, and equiaxed dendritic growth. Theoretical predictions of microsegregation, eutectic fractions and cooling curves are compared with a number of previous theoretical and experimental results, and good agreement is found.

Superplasticity in fine grained ceramics and ceramic composites: current understanding and future prospects

Abstract: Superplasticity in ceramics has now been reported in a wide range of materials with elongations to failure of more than 100%. Although the experimental observations of large deformation are in some ways similar to those reported in numerous metallic alloys, there are significant differences in the mechanical properties and cavitation failure characteristics of superplastic ceramics. This paper provides an overview of superplastic deformation and failure in ceramics, with specific emphasis on a 3 mol.% yttria stabilized zirconia and a zirconia-20wt.%alumina composite. It is demonstrated that there is a transition in the deformation behavior of zirconia which is dependent on the grain size and the impurity content of the material. Many of these materials fail by the nucleation, growth and interlinkage of cavities, so that the ductility is governed by the imposed stress and the grain size. Potential areas for additional research on superplastic ceramics are highlighted.

Determination of dendritic coherency in solidifying melts by rheological measurements

Abstract: The instant, during equiaxed solidification, when the dendrites start to impinge and form a network is called the dendritic coherency point. The paper presents a method for measuring the dendritic coherency point. The method utilizes the fact that the shear strength of the solidifying material increases sharply at the coherency point. A probe is solowly rotated in the melt during controlled solidification and the torque and temperature are measured. The coherency point can be deduced from the derivative of the torque/temperature curve. Coherency measurements of solidfying aluminium alloys have shown that the coherency point for a given alloy is reproducible for given cooling conditions, but varies greatly with alloying composition, grain refinement and cooling rate. It has been found that the fraction solid at which coherency occurs varies between less than 0.1 for poorly grain refined Al foundry alloys and more than 0.5 for fine grained Al alloys.

On the effect of nitrogen on the dislocation structure of austenitic stainless steel

Abstract: The strengthening and work hardening characteristics of high nitrogen austenitic steel have been investigated by TEM in order to explain the outstanding mechanical properties, which are very high yield strength and good toughness. It is shown that plastic deformation of those steels always occurs by a combination of planar glide and mechanical twinning. However, the critical stress/strain conditions for the onset of mechanical twinning depend strongly on the actual nitrogen content. Specifically, as the nitrogen content is increased, the onset of deformation twinning is shifted to lower strains and higher stresses, i.e. the more important becomes the contribution of deformation twinning to the total strain. The observed behaviour is explained by the influence of nitrogen on internal friction and stacking fault energy. It is concluded that the high yield strength of cold worked nitrogen-bearing steel is essentially due to tight stackings of twins and stacking faults. Apart from regular structure evolution, inhomogeneous structures are observed, which can be explained in terms of texture formation and the dynamic properties of stacking faults under stress.

A new method for activation volume measurements: application to Ni3(Al,Hf)

Abstract: A new method is presented that allows the measurement of microscopic activation volumes which are typical of thermally activated dislocation processes. It consists of a repeated relaxation test with constant durations, which start at a given stress level. The advantages of this method, as compared to a previous one which uses constant stress drops, are outlined. The new method is successfully applied to the determination of the transition stress between micro- and macro-plastic deformation in the alloy Ni 3 (Al,Hf) at room temperature.

Cracking and stress redistribution in ceramic layered composites

Abstract: Problems are analyzed that have bearing on cracking and survivability in the presence of cracking of layered composite materials composed of brittle layers joined by either a weak interface or a thin layer of a well-bonded ductile metal. The problems concern a crack in one brittle layer impinging on the interface with the neighbouring brittle layer and either branching, if the interface is weak, or inducing plastic yielding, if a ductile bonding agent is present. For the case of a weak interface, the effect of debonding along the interface is analyzed and results for the stress redistribution in the uncracked layer directly ahead of the crack tip are presented. Debonding lowers the high stress concentration just across the interface, but causes a small increase in the tensile stresses further ahead of the tip in the uncracked layer. A similar stress redistribution occurs when the layers are joined by a very thin ductile layer that undergoes yielding above and below the crack tip, allowing the cracked layer to redistribute its load to the neighbouring uncracked layer. The role of debonding and yielding of the interface in three-dimensional tunnel cracking through an individual layer is also discussed and analyzed. Residual stress in the layers is included in the analysis.

Model for the prediction of the mechanical behaviour of nanocrystalline materials

Abstract: A model is proposed in the present paper to account for the mechanical behaviour of nanocrystalline materials. In this model, the distribution of the grain size in nanocrystals is simulated with a logarithmic normal distribution, and one dislocation per grain is assumed. The plastic yielding for nanocrystalline materials is considered to be controlled by the stress required to attain dislocation loops (the Frank-Read source) in a set of larger grains with their critical semicircle configuration. The dislocations in the rest of the smaller grains are considered to be in the subcritical configuration, which produces a reversible deformation and only contributes to an inelastic deformation. The model presents a very good agreement with the σyvs. Dav−1/2 relationships for five nanocrystalline materials; of these, three metals exhibit a negative Hall-Petch slope and two a positive Hall-Petch slope. The model also predicts a decrease in Young's modulus with diminishing grain size, which is in agreement with experimental results for nanocrystalline copper and palladium.

Kinetics of precipitation in AlSc alloys and low temperature solid solubility of scandium in aluminium studied by electrical resistivity measurements

Abstract: The precipitation kinetics of Al 3 Sc phases in Al-0.090at.%Sc and Al-0.15at.%Sc alloys were investigated in the aging temperature range from 533 to 733 K mainly by electrical resistivity measurements. The average Sc concentration c Sc of the Al-rich matrix was calculated from the value of the resistivity. The results of aging time dependence of c Sc and fraction of solute transformed were analysed from the viewpoint of time scaling, the Johnson-Mehl-Avrami equation and the Ostwald ripening theory. The rapid growth stage of the Al 3 Sc phases was successfully scaled and nearly characterized by the model of diffusion-controlled growth of a fixed number of particles. The solid solubilities of Sc in aluminium from 733 K down to 643 K were estimated by the analysis of c Sc of the Al-rich matrix during the Ostwald ripening of the Al 3 Sc precipitates.

Effects of electromagnetic vibrations on the microstructure of continuously cast aluminium alloys

Abstract: The influence of electromagnetic vibrations imposed during solidification on grain refinement in the 1085 and 2214 aluminium alloys, characterized by a narrow and a wide freezing range respectively, has been examined. The vibrations were produced by the simultaneous application of a stationary magnetic field B0′ and a variable magnetic field B of frequency 50 Hz in the sump of continuously cast ingots. Extensive grain refinement has been observed in both alloys with increasing magnetizing force. This study shows the mean grain size obtained by this vibrational technique is always smaller than that produced by the recently developed CREM (carting, refining, electromagnetic) process.

Tribological properties of SiCp-reinforced Al-4.5% Cu-1.5% Mg alloy composites

Abstract: SiC p -reinforced Al-4.5% Cu-1.5% Mg composite specimens were processed through vigorous stirring of the carbide in a semi-solid alloy slurry, remelting and casting. The specimens were examined microstructurally and their tribological properties were evaluated both in the as-cast and heat-treated conditions by pin-on-disk tribometry and automatic scratch testing. The specific wear rate and the friction coefficient between a diamond stylus and the polished specimen surface decreased, and the composite microhardness increased with increasing volume fraction carbide, and decreasing carbide particle size and spacing. A solution and ageing treatment led to an increase in wear resistance and composite microhardness, and a reduction in friction coefficient.

Low cycle fatigue damage in nickel-base superalloy single crystals at elevated temperature

Abstract: Low cycle fatigue tests on AM1 nickel-base superalloy single crystals were conducted under axial strain control at 650, 950 and 1100 °C. The behaviour of the [001] orientation was investigated at the three temperatures, that of the [111], [101] and [213] specimens was studied at the two lower temperatures. The orientation dependence of fatigue life-total strain range curves was mainly due to variations in Young's modulus with orientation. Most cracks grow in stege II mode whatever the temperature. Cracks nucleate at micropores and in the interior of specimens at low temperatures; surface cracks induced by oxidation are dominant at high temperatures and low strain ranges. Most of fatigue life is spent in microcrack growth.

Predicting slip behavior in alloys containing shearable and strong particles

Abstract: Deformation behaviour is predicted by combining slip intensity calculations for shearable precipitates (δ′) with critical particle size (for dislocation shearing-to-looping transition) estimates for strong precipitates (T1, S′). The deformation behavior of the microstructures correlates fairly well with the predictions. Calculations imply that significant volume fractions of shearable phases affect the shearability of strong precipitates. Although specific examples are drawn from AlLiCuMg alloys, the general theory is applicable to most precipitation-hardened alloys. The predictions of the mechanical behavior made for the various microstructures correlate well with the available data.

The modelling of the mechanical alloying process in a planetary ball mill: comparison between theory and in-situ observations

Abstract: In order to understand the mechanical alloying (MA) process in a planetary ball mill (PBM), the kinetics of a ball during milling have been analysed and the calculated ball trajectory is compared with in-situ observations of ball movement. Three different milling modes are deduced from the modelling. Although an experimentally efficient milling mode can be explained from the modelling, there is no good agreement between theory and in-situ observations in the mode of operation of most commercially available PBMs. The present study shows that, in these mills, the MA process should be described in terms of attrition and wear and not in terms of impact as is usually done.

Deformation twinning in intermetallic compounds—the dilemma of shears vs. shuffles

Abstract: In this paper the geometric description and general theory of mechanical twinning are reviewed, the twins in general lattices and superlattices are summarized, and the kinematic process by which mechanical twins form is revisited. A case study of mechanical twinning of HfV2 + Nb, a C15 (cubic) Laves phase, is presented and the synchroshear of selected atomic layers is proposed to explain the physical process of twin formation. If the twins form in this way, then long shear vectors and/or atomic shuffles are not really necessary.

High speed laser cladding: solidification conditions and microstructure of a cobalt-based alloy

Abstract: Laser cladding experiments with a hypoeutectic Stellite 6 alloy have been performed with scanning speeds ranging from conventional values (1.67 mm s−1) up to very fast values (167 mm s−1). The evolution of the secondary dendrite arm spacing is discussed using the solidification conditions deduced from a two-dimensional heat flux model and the quaternary CoCrCW phase diagram.

A method for simulating electron microscope dislocation images

Abstract: An image simulation program has been developed to quantitatively evaluate transmission electron microscopy (TEM) images of dislocation configurations. It takes into account many beams, includes linear anisotropic elasticity and uses the column approximation. The program simulates the image of 1 to 4 parallel dislocations and their associated planar faults under bright field, dark field and various weak-beam TEM conditions. Simulations of a variety of dislocations are presented here and it is shown that the use of a many-beam calculation is essential in the case of weak-beam identification of weakly dissociated dislocations.

Thermal stability of ultrafine-grained metals and alloys

Abstract: Grin growth in ultrafine-grained elemental metals (Cu, Ag, Pd) and alloys has been studied by differential scanning calorimetry and transmission electron microscopy. The samples were prepared via the inert gas condensation technique, followed by uniaxial high pressure compaction. Abnormal grain growth is observed in all the pure elemental samples. The onset of this secondary recrystallization is at or slightly below the recrystallization temperature known from the respective conventionally cold-worked metals. The activation energies for grain growth were determined by the Kissinger method and correspond to the typical values for grain boundary diffusion. For samples measured directly after preparation, a two-step process was found. The low temperature reaction can be attributed to relaxation processes, because no microstructural changes can be observed. The reaction at higher temperature is due to abnormal grain growth. Principally the same behaviour was found in samples with increased residual porosity (up to 15 vol.%) and concentration of substitutional solutes (up to 5 at.% Au). Gaseous impurities (oxygen) increase the onset temperature for abnormal grain growth. Normal grain growth is observed in duplex microstructures, as demonstrated using AgCu as a model system.