Band 93, Heft 7

Conventional theory shows that if the mean free path of newly formed dislocations is independent of plastic strain, then the plastic strain is proportional to the square of the stress. If the mean free path decreases during deformation, being proportional to the distance between neighbouring stored dislocations, then the plastic strain is linear in the stress. Experiments on WC–Co hard metals show a region in which the plastic strain increases as the fourth power of the stress, followed by a catastrophic increase. This suggests that the mean free path of dislocations increases with strain and stress. A model is developed in which dislocations travel almost freely in the Co binder and form inverted pile-ups in the terminating WC grains, as in the analysis by Bilby, Cottrell and Swinden of the spread of plastic yield from a notch. The model reproduces the rapid increase of plastic strain with increasing stress reasonably satisfactorily and predicts a very high rate of work hardening as is observed.

The interaction of dislocations with finely-dispersed incoherent, non-shearable, hard particles governing the strength of oxide dispersion-strengthened (ODS) materials is briefly reviewed with particular emphasis on the basic features of the operating high-temperature mechanisms. The experimental determination of the particle-threshold stress within the yield stress- increment approach as well as its modelling in terms of the “interfacial pinning” mechanism are outlined and exemplified with the austenitic ODS superalloy MA 754 and its ferritic counterparts MA 956 and PM 2000. The available data of very recent high-voltage electron microscopic observations on MA 754 and MA 956 are re-evaluated. In situ straining experiments image moving dislocations under full stress and yield valuable quantitative informations characterizing the controlling (attractive) dislocation – dispersoid particle interaction which are not directly accessible by other methods, such as the (effective) obstacle density; the specific obstacle (particle) strength derived from the relevant interaction parameters; the locally acting shear stress estimated using the measured particle spacings, the curvature of the bowed-out dislocation segments in the stress-applied state, and their line tension calculated within the de Wit-Koehler model. The findings obtained in combination with the results of earlier conventional electron microscopy and those of critically analyzed macroscopic deformation data support convincingly our understanding of the elementary physical mechanisms underlying elevated temperature dispersion strengthening which promises, eventually, to put microstructural-based model description of the yield/creep strength on a more quantitative footing.

In order to characterise the influence of hydrogen on the mechanical properties of the β titanium alloy Timetal ® 21S, monotonic tensile tests as well as total strain-controlled fatigue tests were performed on cylindrical samples. Different microstructures were established by heat treatment prior to mechanical testing. Hydrogen charging was carried out from the gas phase during different steps of the heat treatment. The results obtained from hydrogen-charged and hydrogen-free specimens deviate strongly when the material undergoes microstructural changes during the heat treatment. A correlation of mechanical test results with electron microscopy observations indicate that the direct (intrinsic) influence of hydrogen is small and that the indirect (extrinsic) influence prevails. The main change in the microstructure is a result of the β-stabilizing effect of hydrogen, leading to a reduction in volume fraction of secondary α formed during aging.

In the present investigation, the cyclic deformation behaviour of TiAl6V4 and TiAl6Nb7 was characterized in constant-amplitude and load increase tests in laboratory air and quasi-physiological media by mechanical hysteresis and temperature measurements. Microstructural changes were evaluated by scanning and transmission electron microscopy for defined fatigue states and after specimen failure. The cyclic stress –strain curves of the alloys exhibit a cyclic softening and/or hardening behaviour depending on the stress amplitude. In constant-amplitude tests, the alloys show pronounced softening and/or hardening phases. In load increase tests with strain and temperature measurements, estimated values for the endurance limit were determined which correspond well with the results of Woehler tests. In the load increase tests, an influence of the media on the cycle-dependent damage state could not be detected.

Precision cast high-temperature components often contain specific imperfections like micropores or microcavities. The influence of the position, size and distribution of such micro cavities as well as of the parameters of the thermomechanical fatigue loading on the lifetime behaviour of the superalloy MAR-M247LC was investigated using two specimen series, containing a high and a low amount of microcavities and a state from which any microcavity was removed by hot isostatic pressing. Thermo-mechanical fatigue (TMF) tests with zero (“in phase”) and – 135° phase shift between temperature and mechanical strain were performed. The different damage behaviour at in-phase and – 135° TMF requires different methods of the assessment of microcavities. For the in-phase loading, they are regarded as cracks and their effect on the local strain distribution is determined by the crack opening displacement. This leads to very good results for both specimen states investigated. For the – 135° cycle, it is taken into account that the microcavities in the core of the specimens increase the loading in the microcavity-free shell of the specimen. This method corresponds closely to the effective damage mechanism and yields good results for both microcavity contents investigated.

Increases of residual stress stability and alternating bending strength of shot peened AISI 4140 are obtained by successive annealing treatments. This is caused by static strain aging effects, which lead to pinning of dislocations by carbon atoms and very small carbides. It will be shown that by well directed annealing of a quenched and tempered AISI 4140 it is possible to maximize the positive effects of static strain aging, without causing extended thermal residual stress relaxation. The amount of yield stress increases caused by static strain aging is quantified using tensile tests. Static strain aging is also found to be responsible for an increase of the quasi static and cyclic surface yield strength present after shot peening.

Cyclic deformation behaviour of ultrafine-grained (UFG) materials processed by severe plastic deformation is reviewed in the light of possibilities to enhance the fatigue properties of UFG materials. Based on the well-known laws from Coffin-Manson and Basquin the fatigue behaviour of UFG materials for both, the low-cycle fatigue as well as the high-cycle fatigue regimes, are analyzed. The beneficial effects of an annealing heat treatment or a back-pressure modified equal channel angular pressing (ECAP) technique on the fatigue behaviour of UFG materials are demonstrated and critically discussed. In particular, the positive influence of an annealing heat treatment on the fatigue behaviour is explained in terms of an improved cyclic stability of the microstructure after ECAP and of an enhanced ductility. A back-pressure modified ECAP technique was used in order to achieve a ultrafine-grained, high-strength material with increased ductility, which itself is related to a combination of crystallographic slip and grain boundary sliding. The potential of enhancing the fatigue properties of UFG materials by the back-pressure ECAP technique is discussed in the context of the Coffin-Manson and Basquin relations.

Conventional measurement of stress relaxation in tests with constant total (elastic plus inelastic plus thermal) strain requires high resolution and excellent control of strain and is perturbed by thermal fluctuations. A new method for measuring the stress relaxation behavior is presented which avoids these experimental problems. It is based on the fact that the stress relaxation behavior is fully contained in the relation between inelastic strain rate and stress at given mechanical strain. Consequently, it should be possible to measure stress relaxation in creep machines when stepwise unloading is performed such that the inelastic strain rate can be determined at constant mechanical strain. This is confirmed by tests on the Mg-base alloy AZ91.

The present work represents an exploratory study of creep in a Ni-rich NiTi shape memory alloy (50.7 at.% Ni) in the temperature range 470 – 530 °C at stresses between 90 and 150 MPa. Creep of Ni-rich NiTi alloys is characterized by stress exponents of 2 (for the minimum creep rate data) and of 4.7 (for creep rates measured after 10% of accumulated strain). During creep, Ni 4 Ti 3 precipitates (which nucleate and grow during heating under a small pre-load) coarsen. The interparticle spacing increases and this is why the resistance of the material to creep decreases. It was not possible to rationalize this creep softening on the basis of a traditional Orowan process. Instead a wrapping process, where particles were wrapped in by migrating sub-grain boundaries, was observed. The sharp minima in the creep curves and the strain dependence of the stress exponent as observed in the present study can be rationalized on the basis of an initial climb controlled bypassing of particles followed by this wrapping process, which becomes more intense as interparticle spacings increase. When comparing crept materials with materials which were aged at the same temperature for the same time there only is a small difference between the corresponding differential scanning calorimetry (DSC) charts. Therefore we conclude that the precipitation and coarsening of Ni 4 Ti 3 particles during creep is more important in affecting DSC chart features than the evolution of a creep dislocation structure.

Smooth and notched specimens of superalloy single crystals CMSX-4 with 〈001〉 orientation were subjected to creep and creep/high-cycle-fatigue loading at temperatures of 800 and 850 °C in air. Superposition of a cyclic stress component on the static stress component leads to a slowdown of the creep rate at small creep strains and to the formation of persistent slip bands running through both the γ matrix and the γ ′ ${{\gamma }^{\prime }}$ precipitates. Consequence of the formation of these zones is fatigue crack initiation at casting pores and shortening of lifetime at higher stress amplitudes.

A constitutive model for creep deformation in pure and dispersion-strengthened copper is presented. This description reproduces the characteristic features of steady-state and constant-structure creep with a single set of microstructurally founded parameters. In particular, the stress and temperature dependences of the creep resistance are adequately represented without introducing artificially high and, therefore, physically questionable values for the stress exponent and the activation energy of creep.

The temperature and strain-rate dependence of the critical resolved shear stress (CRSS) of high-purity NiAl single crystals in soft orientation (tensile axis near 〈111〉) was measured in tensile tests with constant strain rate, ε˙ $\dot{\varepsilon }$ p = 10 –4 s –1 , between 77 and 325 K. After pre-straining at room temperature, the crystals were plastically deformed by strain increments of Δε p ≈ 0.005 at successively decreasing temperatures. Below 325 K the CRSS on the 〈001〉{011} slip system slowly increases with decreasing temperature, followed by a steeper increase below about 200 K. The strain-rate sensitivity of the CRSS measured by stress-relaxation experiments exhibits comparable behaviour. The parameters for thermally activated dislocation motion, activation volume and activation energy, suggest that different mechanisms control the dislocation motion above and below 200 K. For low stresses, at temperatures above 200 K, an interpretation of the results within the framework of solid solution hardening theory is suggested. Below 200 K the results can be interpreted within the framework of the line-tension model of the kink-pair theory of flow stress. The evaluation of the data yields a kink-pair energy of 2 H k = 0.42 eV, a Peierls stress of 535 MPa, and a Peierls valley distance equal to the lattice constant. These data are generally compatible with a fundamental process of kink-pair formation on {110} or on {100} planes. The results are discussed with respect to recent atomistic modelling of dislocations in NiAl.

A very simple one-dimensional cellular automaton numerical model of plastic deformation that takes into account the local heterogeneity of strains, stresses and dislocation density has been developed. Results corresponding to an application of the model to low-temperature deformation (no long-range diffusion, no account for the influence of spatial strain rate fluctuations) are presented. The calculated stress – strain curves, mean dislocation density evolution, spatial fluctuations of internal stresses and dislocation density, are qualitatively and semi-quantitatively correct up to large plastic strains. In many ways, the results illustrate the validity of the ideas of Mughrabi’s “composite model”. It is of particular interest that the model predicts a very realistic stage IV after stage III for polycrystalline deformation as well as a convincing Hall-Petch effect for the flow stress – grain size relationship.

The behavior of the width of X-ray Bragg peaks recorded from dislocated crystals has been investigated both numerically and experimentally. Numerical simulations based on the Wilkens model of a restrictedly random dislocation distribution have shown that the width of the X-ray peaks increases linearly in the modified Williamson-Hall ( m WH) plot. The slope of the line is proportional to the square root of the average dislocation density. The method has been used to evaluate the dislocation densities in two texture components of a hot-deformed Al-2.5 wt.% Mg alloy, where peak broadening has additional sources.

The microstructure of three different nanocrystalline materials with hexagonal crystal structure are studied by X-ray diffraction peak profile analysis. The crystallite size distribution and the dislocation structure are determined in plasmathermal silicon nitride powder, sintered tungsten carbide and severely deformed titanium, and are compared with transmission electron microscopy (TEM) results. In the case of the silicon nitride powder, the particle size determined by TEM is in good correlation with the coherently scattering domain size provided by X-rays. In the case of bulk titanium, the crystallite size, given by X-rays, correlates well with the dislocation cell size obtained by TEM. It is found that in tungsten carbide the dominant dislocation slip system is basal, whereas in plastically deformed titanium the basal slip is practically absent with the dominance of pyramidal dislocations, in good correlation with other investigations of the literature.

In der Verkehrstechnik und vielen anderen technischen Bereichen bringt Leichtbau ökonomische und ökologische Vorteile. Daher werden Werkstoffe zunehmend nach ihrem Leichtbaupotential beurteilt. Im Falle der Strukturwerkstoffe, die gewissen Belastungen standhalten müssen, ist allerdings eine Einteilung nach der Dichte („Leichtmetalle‘‘) ungenügend. Der entscheidende Parameter ist vielmehr das Verhältnis von Dichte zu Festigkeit. Im Falle der Biegebeanspruchung wird der Sachverhalt kompliziert, da z. B. die elastische Steifigkeit solcher Bauteile (bei gegebener Länge) nicht nur von ihrem Volumen, sondern auch von ihrer Gestalt abhängt. In dieser Arbeit wird gezeigt, dass der übliche Vergleich von unterschiedlichen Werkstoffen zum Zwecke der Beurteilung ihres Leichtbaupotentials einseitig die weicheren Werkstoffe bevorzugt. Wenn dagegen der Vergleich unter der Bedingung vergleichbarer Außendimensionen gemacht wird, zeigt sich, dass ein und derselbe Parameter für einsinnigen Zug als auch für Biegebeanspruchungen gilt. Damit wird klar, dass das Leichtbaupotential von hochfesten Stählen und Aluminiumlegierungen vergleichbar ist – nicht nur unter einsinnigem Zug, sondern auch unter Biegebeanspruchungen.

Fatigue tests are performed on cylindrical, sheet and foil specimens to investigate the strain-induced martensitic transformation of austenitic steels. The tests set a variety of strain amplitudes and temperatures as well as various carbon concentrations for the investigated materials in order to broaden the scope of the martensitic transformation and thereby its characterization. It is confirmed that there is a threshold value of plastic strain amplitude for the investigated materials above which the incremental increase in martensite volume fraction in the specimen under continued cyclic loading is of significance. Block-loading tests indicate that cycles at amplitudes below this threshold do not contribute to the austenite transformation. Thermodynamic calculations are applied in order to provide information on the stability of the material with regard to carbon content. It is shown that carbon concentration can be established by gas charging of the materials so that the driving force for transformation can be adjusted purposefully.

The present study identifies key material phenomena that limit the predictive capabilities of current approaches employed in modelling the properties of heat treated components. Focus is on aspects related to the isothermal austenite-to-bainite transformation in low alloy steels. Specifically, experimental data are presented on (i) the stress – strain response of austenite prior to transformation, (ii) the magnitude and progress of transformation plasticity and (iii) the effect of externally applied loads on the kinetics of the bainitic phase transformation. The data obtained demonstrate a substantial stress-induced as well as a strain-induced acceleration of the bainitic transformation, which has to be accounted for in modelling. Moreover, stress – strain response of the supercooled austenite was found to be drastically different from that estimated based on the behaviour of stable austenitic steels. The results also point out that transformation plasticity can contribute significantly to distortion of heat treated components. Still, some of the assumptions widely used in modelling this phenomenon are clearly incompatible with the experimentally obtained data.

The damage evolution during fatigue of a particulate-reinforced metal matrix composite was investigated experimentally by two methods, i. e., by the measurement of the samples’ stiffness values within individual stress – strain hysteresis loops and by the measurement and evaluation of the acoustic emission during cyclic deformation. Both methods indicate that the first damage occurs during the first loading cycle, in tension as well as in compression. After a short period of a high damage rate which is due to fracture of large particles, a longer range of a smaller damage rate follows, in which smaller particles break. Finally, both methods indicate the growth of the main (and final) fatigue crack.

Heavily faulted Si flakes, which can be considered as crystallization products of eutectic droplets embedded in an Al matrix, were for the first time observed in an as-thixoformed A356 alloy, which contains 0.3 wt.% Sr. Large isolated particles as well as clusters of fine Si particles were observed. Evidence is provided that the faults on {111} planes of Si are not twins. Sr and Na are known to modify the morphology of the Si phase from flake to fiber morphology. Contrary to common perception, it is shown that the twin plane re-entrant edge mechanism does not operate during fiber growth. It is proposed that the modification process occurs during the nucleation of the Si phase by adsorption of modifying elements on twinned nuclei of Si. They prevent nucleus growth into stable twinned crystals, which is a prerequisite for the growth of Si flakes.

Scanning and transmission electron microscopy investigations and X-ray diffraction measurements were performed on cyclically deformed nickel samples showing average grain sizes D between 500 nm and 5 μm and containing different initial dislocation structures. The results are discussed in comparison with main microstructure features observed in fatigued nickel with D = 30 μm and D = 250 μm. A lower threshold value D TH ≈ 1 μm was found for the development of typical fatigue-induced dislocation patterns (FIDPs). The spacing d PR between dislocation-dense regions in the FIDPs is nearly independent of D down to D TH and depends only weakly on the strain amplitude ε pa ( d PR ≈ 500 nm at ε pa = 2⨯ 10 –3 ). While for D ≥ 3 μm no D dependence on the cyclic stress–strain curves could be observed, at D < 3 μm the stress level of the cyclic stress – strain curves seems to follow a Hall-Petch relation.

The tracer self-diffusion coefficients D T of implanted 71 Ge and 31 Si in both relaxed monocrystalline Si y Ge 1 –y epilayers and specimens made from bulk Si y Ge 1 –y have been measured as functions of temperature T (653 °C ≤ T ≤ 1263 °C) and composition y (0 ≤ y ≤ 1) by means of radio-tracer techniques, in which serial sectioning was done by ion-beam sputtering. For all compositions, the T dependencies of D T for 71 Ge and 31 Si are of Arrhenius type. The y dependencies of the pre-exponential factors of the self-diffusion coefficients and of the diffusion enthalpies show a break at y ≈ 0.65. This is interpreted in terms of a transition from interstitialcy-( y ≳ 0.65) to vacancy-( y ≲ 0.65) mediated self-diffusion.

The influence of corrosion on the strength of a commercial hardmetal and its behaviour under static and cyclic loading has been investigated. Tannic acid was chosen as the corrosive medium in order to simulate the environment to which hardmetal wood-cutting tools are exposed. The results show corrosive attack of unloaded specimens to proceed in a highly localised form, producing defects that reduce the strength of the material by up to 50 %. A further reduction in the maximum endured stress is observed when corrosive attack is combined with a simultaneous static load. This is considered to be due to hydrogen embrittlement. When the hardmetal is subjected to cyclic loading in the presence of the corrosive medium, an increase in the fatigue sensitivity of the material occurs, but noteworthy reductions in lifetime are only observed at low stress amplitudes.

We developed a model to describe the measured dependency of the radii of micropipes in 6H–SiC on their Burgers vector contents. This model permits to assess the internal stress that is present in a crystal during the early stage of seeded growth. This stress cannot be directly measured so far because of the high growth temperatures exceeding 2300 K. The evaluation yields stress levels of approximately 10 – 3 of the (room temperature) shear modulus.