ALBERT

Notes: The results of research into short fatigue crack behaviour in shot peened, medium carbon steel specimens under reversed torsion are presented. Characteristic features of short crack growth were established on the basis of optical and electron-optical observations of the replicas and micro-sections of the samples. Surface crack growth analysis, supported by plots of crack development, crack growth rate and crack length distributions against cycle ratio show that significantly slower crack growth and lower crack densities occur in shot peened specimens than in non-treated samples. That results can be linked to the effect of grain distortion, the packing of laminar grains, a high dislocation density in the plastically deformed surface layer and the introduction of compressive residual stresses. A fractographic analysis of the fracture surfaces has assisted an understanding of the mechanisms of fracture in shot peened specimens.

Notes: Abstract— An experimental study within the Canadian Offshore Corrosion Fatigue Research Programme was performed on the early development of fatigue cracking along the wavy toe of manual fillet welds between structural steel plates. Stress relieved and as-welded cruciform joints were tested under R =−1 and R= 0 loading at different stress amplitudes. The depth and the opening level of cracks as small as 10–20 μm were monitored using miniature strain gauges installed along the toe apex, in combination with beach marking. Most of the “initiation life” (25% to 50% of total life), conventionally defined by a crack depth of 0.5 mm, is consumed in short crack propagation. Three types of short crack development for different combinations of local mean stress and stress range are identified and analyzed. Growth rates in as-welded specimens are faster than in stress relieved specimens, which results in shorter “initiation lives”. This is associated with a higher effective stress range, particularly under R = - 1 loading where cracks are open over nearly the full stress range. The V-notch stress intensity factor is a promising parameter to rationalize the crack “initiation life”. It takes into account the thickness effect experimentally observed. Under R = - 1 loading of as-welded joints, using R = 0 data and taking the whole stress range gives a reasonably conservative approximation of the crack “initiation life”.

Notes: Abstract— For continuously welded structures subjected to cyclic loading, the highly stressed zones where cracks initiate and lead to failure are usually located at weld toes. At these critical points, called hot-spots, the very local stress states are difficult to determine so that standard fatigue criteria are very difficult to apply for fatigue life prediction.This work presents a fatigue design criterion for continuously welded thin sheet structures, based on a unique S-N curve. The approach, which refers to the hot-spot stress concept, defines the design stress S as the geometrical stress amplitude at the hot-spot.In practice, the geometrical stress state is calculated by means of the finite element method (FEM) using thin shell theory. Meshing rules for the welded connection, which can be applied methodically to any welding situation, allow the hot-spot location, and therefore the design stress of any structure, to be determined.Experimental data and FEM calculations show that a unique S-N curve can be obtained whatever the geometry of the welded structure and the loading mode.

Notes: Abstract— The paper summarises existing fatigue data for welded aluminium tubular joints and attempts a first classification and formulation of design values on the basis of statistical-regressional analysis to be utilised in new design code drafts. A comparison with results in steel is performed and the need for further investigations outlined.

Notes: Abstract— The effect of axial misalignment on the fatigue strength of load-carrying transverse cruciform welded joints was investigated using experimental and fracture mechanics methods. Where failure occurred by cracking from the weld toe, misalignment significantly reduced the fatigue strength. The reduction could be predicted using a nominal stress concentration factor (SCF). Misalignment had less effect where failure was due to cracking through the weld metal; an expression was deduced for the SCF in this case. For fracture mechanics assessments, an expression for an effective stress intensity factor using the SCF and stress intensity factors for aligned welds was shown to agree with the finite element (FE) results. Predictions of the effect of misalignment using the FE results agreed with experimental data. Misaligned transverse load-carrying cruciform joints should be assessed for fatigue failure from the toe using the same SCF as for a butt weld with the same misalignment. For failure through the throat, an alternative expression for the SCF is recommended. Fracture mechanics assessments of misaligned joints should be carried out using an effective stress intensity factor derived from the SCF and stress intensity factors for aligned joints. These recommendations are now incorporated in British Standard PD 6493:1991.

Notes: Abstract— Forged components of ferritic steel can be protected by a welded austenitic stainless steel clad. Intergranular cracking can take place in the ferritic phase close to the ferritic austenitic interface. After developing a technique for fabrication of these cracks, the formation conditions are studied. Auger electron spectroscopy investigations of specimens containing a real crack opened inside the vacuum chamber are used for interpretation. Sulphur segregations embrittle the grain boundaries which are cracked by residual and thermal stresses during the postweld heat treatment.

Notes: Coil springs made of silicon nitride (Si3N4), which has higher strength and heat resistance than other ceramics, have recently been developed. By examining the range of strengths and the heat resistance of the coil springs, it is shown that the coil springs can be used at temperatures up to 1000°C. We derive an equation to calculate the effective volumes of coil springs and examine the influence of size on the strength of coil springs. In addition, using a process zone size failure criterion, proof tests were conducted and analysed. The results show that the mean strength of coil springs decreases with increasing effective volume. Therefore the strength of coil springs can be estimated from the value of the effective volume, and this relationship can be applied in design calculations. Furthermore, by conducting proof tests, coil springs that have relatively large defects can be eliminated and so only highly reliable coil springs can be placed in service.

Notes: Abstract— Constant and variable amplitude (VA) loading fatigue studies were carried out on a 6261 aluminium alloy using cylindrical plain hour-glass specimens. Crack growth was monitored via surface replication using cellulose acetate.Crack growth results at constant amplitude loading show the typical intermittent high and low periods of growth rate associated with crack-microstructure interactions. Acceleration in growth rate during an overload block depends on crack length and stress amplitude ratio. It appears to pass through a maximum at a crack length corresponding to the first microstructural barrier. Microstructural-based modelling is therefore required for small fatigue cracks, rather than solely closure-based modelling. The Navarro-de los Rios model of short fatigue crack growth appears able to provide good indications of crack growth rates under VA block loading, and gives reasonable life predictions.For short cracks (surface length 〈 80 μm) and a small overload ratio (6.7%), crack growth may show severe retardation during the overload block. This is ascribed to crack tip blunting being more important than the increase in stresses when closure is low. It appears from a Miner's rule type exercise, that VA block loading has its major effect on growth at a surface crack length of 20 μm. This means that the crack initiation period cannot be ignored in life prediction models for small fatigue cracks.

Notes: Abstract— The behaviour of physical short mode I cracks under constant amplitude cyclic loading was investigated both numerically and experimentally. A dynamic two-dimensional elastic-plastic finite element technique was utilised to simulate cyclic crack tip plastic deformation. Different idealisations were investigated. Both stationary and artificially advanced long and short cracks were analysed. A parameter which characterises the plastically deformed crack tip zone, the strain field generated within that zone and the opening and closure of the crack tip were considered. The growth of physically short mode I cracks under constant amplitude fully reversed fatigue loading was investigated experimentally using conventional cast steel EN-9 specimens. Based on a numerical analysis, a crack tip deformation parameter was devised to correlate fatigue crack propagation rates.

Notes: Abstract— Curry's model of the WPS effect has been applied to the results of a previous paper, and is extended to treat warm prestressing in blunt notched test-pieces. The effect of more complex prestress cycles is also predicted by an extrapolation of the model. The effects of the load-cool-fracture, LCF, cycle can be reasonably predicted for both sharply precracked and blunt notched specimens. For the sharply precracked specimens the effects of the load-unload-cool-fracture, LUCF, cycle at — 196°C are consistently overpredicted and this may be due to a decrease in the cleavage fracture stress at — 196°C of the material at the crack tip which has been subjected to repeated plastic straining by the combination of loading cycles. Modifications to the model are suggested which reduce the overproduction but a wide degree of scatter is observed in the experimental observations. Blunt notched specimens show a reasonable correlation between prediction and theory for the tensile LUCF cycle. Problems have been found in predicting the effect of various prestress cycles in different specimens due to the inherent variability in baseline fracture behaviour of the weld metal. It is concluded that the general trend of results is adequately explained by superposition models but that a greater understanding of local flow properties at a crack tip is required to achieve reasonable predictive success for weld metals such as A533BW.

Notes: Abstract— The development of anisotropic material properties due to torsional cyclic plastic straining of tubes with sustained axial loads was examined for a type 316 stainless steel at room temperature and at 500°C. The effect of the cyclic strains and the cumulative ratchetting strains on the axial tensile properties was determined and the results show a significant increase in the tensile strength at both 20°C and 500°C, with more pronounced hardening at the higher temperature. The cyclic shear stress-strain response of the material is shown to be extremely temperature dependent and the hardening ratio is much greater at 500°C, which is consistent with the dynamic strain ageing observed previously for this material. The ratchetting strains are controlled by the cyclic shear strain hardening, by the axial hardening resulting from the cyclic shear and the cumulative axial strains, and by the ratio of the secondary shear stress to the primary axial stress.

Notes: Abstract— An experimental study of the surface evolution during controlled plastic strain amplitude single-step and two-step loading tests reveals the important damage mechanisms for 316L stainless steel. In the first stage, the cyclic plastic strain is concentrated into persistent slip bands (PSBs) and a surface relief is formed consisting of extrusions and intrusions. The frequency of occurrence and the total density of PSBs has been assessed using systematic observations in a scanning electron microscope. The relative volume occupied by PSBs determines the fatigue damage in this stage. Two-step loading has only a small effect on the PSB damage evolution and nearly equal saturated values (corresponding to the applied plastic strain amplitude) were achieved in the single-step and the two-step loading investigations.

Notes: Abstract— The effects of specimen thickness, stress intensity levels and R ratio following single tensile overloads, multiple overloads and overload/underload events have been investigated in a BS4360 Grade 50D steel. The amount of fatigue crack growth retardation increased with both decreasing applied baseline ΔK and increasing overload block size. Smaller increases in retardation were obtained for overload/underload block events compared with block overloads. The data suggests that crack flank plasticity resulting in crack closure adequately accounts for much of the observed behaviour. Near crack tip plasticity was thought to play a more important role in generating crack closure than that remote from the crack tip.

Notes: Abstract— Fatigue crack growth rates in corner notched specimens of forged Ti-6A1-4V, used in the manufacture of rotating aero-engine components, have been measured for fatigue loadings that combine major and minor stress cycles. The loadings are simple representations of the flight cycle and the potential in-flight vibrations, each loading block consisting of high-frequency minor cycles superimposed on the major cycle dwell at maximum load. The crack growth rates are dependent on the number and stress ratio of the minor cycles, but only when they individually contribute to the growth of the crack. Estimates of the fatigue threshold values and near-threshold growth rates associated with the minor cycles have been made, all potential load history effects having been ascribed to the minor cycle component of the loading. Using this data, satisfactory crack propagation life predictions have been demonstrated for a wide range of test conditions involving the conjoint action of major and minor stress cycles.

Notes: Abstract— A new fatigue domain diagram, recently introduced by one of the authors, makes it possible to demonstrate the fatigue behavior of specimens under varying stress amplitude loading both qualitatively and quantitatively. The diagram is briefly reviewed and crack propagation and damage summation of steel specimens under two level and multi level tension-compression loading are simulated and discussed. Typical patterns of low-high and high-low sequence levels are explained, predicted and, with careful classification, shown to follow certain cumulative damage trends. Correlation with experimental results is shown and discussed. The main conclusion is that one can show repeatable trends in H-L and L-H two-step and multi-step loading sequences, only for cases where the local material properties are not drastically changed, and the failure pattern is similar (critical crack propagation or gross yielding) in all stages of the tests. Damage accumulation, expressed as additional crack length, is clearly shown on the general fatigue diagram.

Notes: Abstract— The behaviour of short cracks and the lifetime of unnotched specimens of a normalized medium carbon steel is observed under constant cyclic loading and under cyclic block loading with changing mean stresses. The greatest part of the life is covered by the growth of short fatigue cracks. During cyclic block loading, the length of the loading blocks has a marked influence on the crack density and the lifetime of the specimens. Short blocks lead to high crack densities and low lifetimes. This effect can be explained by a short crack growth model which takes into account the crack arrest before microstructural barriers and crack linking in the case of high crack densities. The lifetimes for cyclic block loading are calculated on the base of the constant level data and found to be in good agreement with the measured lifetimes.

Notes: Abstract— Generally engineering components are subjected to multiaxial variable amplitude loading, which may be non-proportional. One of the best known low cycle fatigue design codes that addresses non-proportional loading problems is the ASME Boiler and Pressure Vessel Code, but this does not include a cycle counting method and it does not allow for the presence of a mean stress. Traditional rainflow methods, incorporating a strain range and a mean stress, are applicable only where there is a single load varying with respect to time, or possibly where multiaxial loads are proportional. So a multiaxial non-proportional cycle counting method and a fatigue damage calculation procedure are proposed here, based on plastic deformation response and a critical plane hypothesis that incorporates the effect of mean stress. A cyclic deformation model, based on the Ramberg-Osgood equation and a multiaxial memory rule, is used to calculate the mean stress response under variable amplitude loading.The proposed procedure is assessed with combined tension/torsion tests on En15R steel under variable amplitude loading. Fatigue life predictions are compared for analysis with and without mean stress corrections, to assess the sensitivity to mean stresses of non-proportional fatigue endurance in the low cycle regime.

Notes: Abstract— In the paper a previously proposed stochastic description for curvilinear fatigue crack growth is extended and adopted to the characterization of crack retardation due to overloads. The model has the form of a cumulative process with random elementary crack increments and random angles of the crack deflection. The basic statistical characteristics of the model-process are related to fracture mechanics and to the parameters known from traditional experimental predictions.

Notes: This paper presents the possibility of using the reverse magnetostriction (Villari) effect in fatigue testing of ferromagnetic materials. The tests were conducted on cyclically loaded nickel with no auxiliary external magnetic field. The following properties were determined: magnetic induction B, magnetic field strength H, energy of the magnetic hysteresis loop ΔM (in the B–H coordinate system), plus such mechanical quantities as stress σ, strain ɛ, plastic strain ɛp and energy of the mechanical hysteresis loop ΔW (in the σ–ɛ coordinate system). A variety of magneto-mechanical characteristics are presented and their susceptibility to loading parameters of the fatigue process are discussed. A relationship between ΔW and ΔM is demonstrated. The Villari effect is shown to be especially useful in determining the cyclic yield limit.

Notes: Four types of toughness were measured over a range of temperatures using CVN, COD, 4PB and PCI specimens made of low alloy steels and weld metals. It was found that there were unified correlations in transition temperatures among the various toughness parameters which could be realized by the dependence of the yield strengths of the test materials on both temperature and strain rate.The intrinsic reason for the correlations could be attributed to local cleavage conditions, which control the ductile-to-brittle transition during ductile fracture processes, whether the specimens were notched or precracked, under static or dynamic loading conditions. It is suggested that the ductile fracture process itself was independent of test temperature, but the local cleavage condition remains to be more fully understood in future work.

Notes: A three-parameter fracture mechanics model is proposed to theoretically analyse the propagation of an elliptical-arc part-through flaw in a round bar subjected to constant cyclic amplitude axial or bending loads. The edge flaw presents an aspect ratio α=ael/bel (ael, bel= ellipse semi-axes) and a relative crack depth ζ=a/D, where a and D are the depth of the deepest point on the crack front and the bar diameter, respectively. Additionally a parameter s=ael/a (ellipse shifting) defines the distance of the ellipse centre from the bar circumference. The surface flaw growth occurs according to preferred patterns which tend to converge to an inclined asymptotic plane in the diagram of α against s and ζ.

Notes: Two titanium alloys TA6V and TD5AC were tested. Tensile tests were performed, under static and dynamic loadings, on cylindrical notched and fatigue precracked specimens. The visco-plastic constitutive equations of the alloy were found by fitting finite element computations with the experimental results. The dynamic fracture toughness was obtained by applying the convolution method of Bui and Maigre. Results for the TA6V alloy did not display significant variations of fracture toughness with loading rate, whereas for the TD5AC alloy an increase was measured. The critical void growth was found to be independent of the strain rate. Fair predictions of the fracture toughness under static as well as under dynamic conditions could be achieved by finite element computations using these experimental critical void growth values.

Notes: A new test specimen geometry was advised to investigate unstable crack propagation and crack arrest. This geometry is a cracked ring which is subjected to a compressive load applied to its poles while the crack is located on the equatorial plane at the outer surface of the specimen. The main interest of this geometry lies in the variation of the stress intensity factor, K, with crack length which follows a bell-shaped curve numerically determined. The increasing part of the curve enables us to study unstable propagation and the decreasing one ensures crack arrest. This experiment has two major advantages in comparison with other specimen geometries; the boundary conditions are well controlled during the propagation, and the loading conditions of the crack are therefore precisely known. The round shape of the ring reduces wave reflection effects from free boundary surfaces. It is therefore shown that a static analysis can then be used to investigate crack arrest behaviour.

Notes: Principles of Microstructural Fracture Mechanics (MFM) are used to develop a model for the characterization of environment-assisted short fatigue crack growth.Fatigue cracks are invariably initiated at corrosion pits formed at inclusions, hence the analysis includes stress concentration effects at pits that lead to the propagation of fatigue cracks the rates of which are considered to be proportional to the crack tip plastic displacement. This plasticity is constrained by microstructural barriers which are overcome in a non-aggressive environment at critical crack lengths only when the applied stress is higher than the fatigue limit. However, the superposition of an aggressive environment assists fatigue damage via crack tip dissolution, enhancement of crack tip plastic deformation, the introduction of stress concentrations at pits and a reduction of the strength of the microstructural barrier. These environment effects are manifested in a drastic reduction of the fatigue limit and higher crack propagation rates.The model is compared with fatigue crack propagation data of a BS251A58 steel tested in reversed torsion when submerged in a 0.6M NaCl solution.

Notes: Abstract— Two scalar parameters relating to stored energies are defined along an arbitrary cyclic loading path for an isotropic elastic-plastic continuum. The definitions of the parameters are introduced in the one-dimensional case. The tensorial generalization is developed firstly under the mechanical point of view and secondly focusing attention on the thermodynamical rates. The case of the usual cyclic traction test is studied in order to suggest an interest in the proposed parameters in the quasi-reversible range and in the subsequent transition range between reversibility and quasi-perfect plasticity.

Notes: Abstract— Criteria exist in the field of multiaxial fatigue for predicting the cyclic lifetime for a high number of cycles. They can predict the fracture of a part, but they are not able to calculate the damage level introduced. Since such theories are often used in Research and Development departments, we have based this present study on these criteria. Experiments were made on more than 1000 specimens that now allow us to propose a general procedure for the use of multiaxial criteria after pre-cycling, i.e. for non-virgin materials. We also suggest the introduction of a threshold, which depends on the damage level, to filter out small cycles.

Notes: Abstract— Gas transmission pipes are sometimes subject to external damage due to interference by excavators. Optimized grinding of gouges may offer a solution in repairing pipes. Since pipelines may be subjected to internal pressure variations, the Dang Van criterion has been used to size the allowable grinding depth. The criterion's boundary has been determined from uniaxial fatigue testing and extrapolated to the higher hydrostatic pressure which occurs in practice. The aim of this paper is to check by means of biaxial fatigue tests on specimens, that this extrapolation ensures a lifetime of 105 cycles for the ground pipes. The state of stresses in the central area of the biaxial specimen was calculated from an elastic finite element simulation. The test machine had independent motions of the two perpendicular axes of loading. The fatigue testing was limited to 105 cycles and was carried out at a hydrostatic pressure of 200 MPa. The results validated the extrapolation technique for the Dang Van criterion.

Notes: Abstract— Mixed-mode fatigue crack growth has been studied using four point bend specimens under asymmetric loads. A detailed finite element analysis provides the stress intensity factors for curved cracks under different mixed-mode load conditions. Both fatigue crack growth direction and crack growth rate are studied. The maximum tangential stress and the minimum strain energy density criteria were found to provide satisfactory predictions of the crack growth directions. An effective stress intensity factor was used to correlate the fatigue crack growth rates successfully. It is found that the use of mode I fatigue crack growth rate properties results in a conservative crack growth rate prediction for mixed-mode load conditions.

Notes: Abstract— In this paper are presented the results of fatigue crack propagation tests on angled-slit, three point bend mixed-mode (I + III) specimens manufactured from a low pressure steam turbine rotor forging. The path of crack propagation has been studied for two mixed mode (I + III) loading conditions. It has been observed that crack growth occurs by a mode I mechanism and a model has been developed to correlate crack growth rates in mixed mode (I + III) specimens with data from pure mode I fatigue tests.

Notes: Abstract— Fatigue tests of sheet specimens having a central crack were carried out to study the effects of holes and dimples on the arrest of fatigue crack propagation. Two holes were drilled at some distance from, and at either side of, a crack tip, and the dimple of a certain diameter was introduced by pressing steel balls in the specimen at a crack tip. Results showed that the two holes produced an increase in crack propagation life (about 3 times) when the holes were drilled at an appropriate distance. On the other hand, the effect of a dimple on the fatigue strength was remarkably large, i.e. in the greatest case a 2.2 times increase in the fatigue endurance limit of cracked specimens and about a 50 times increase in the crack propagation life, at stresses above the fatigue limit. The main reason for the remarkable recovery of fatigue strength was the residual compressive stresses produced by the dimple. To evaluate the effect of residual compressive stresses on the da/dN vs. δK relation, a simple model is proposed. By using this model, the effect of residual stresses on crack propagation can be estimated quantitatively. Furthermore, the fatigue life of dimpled specimens was estimated based on the model.

Notes: Abstract— A review of variable fatigue loading of carbon fibre/epoxy laminates indicates that for two-stage loading, low-to-high loading is generally more damaging than high-to-low loading. Damage accumulation under two-stage loading appears to produce linear damage in some CFRP materials and non-linear behaviour in others. The behaviour of those materials exhibiting linear damage may be predicted by the Palmgren-Miner rule. A model is presented for determining the longitudinal split growth in notched unidirectional carbon fibre/epoxy test coupons. The model is able to predict the non-linear split growth for low-to-high loading whilst the linear damage accumulation for high-to-low loading may best be determined by use of the Palmgren-Miner rule.

Notes: Abstract A study was completed on irradiated and unirradiated (offcut) material from Pickering Nuclear Generating Station (NGS) A Unit 3 pressure tubes which were fabricated from ingots of 100% recycled material (quadruple-vacuum-arc-melted). The toughness of the material is shown to be generally higher than that of previous material fabricated from 〈 100% recycled material (double-vacuum-arc-melted). Fractographic and microstructural studies demonstrate that variations in toughness of the low chlorine (〈 1 ppm) material are predominantly due to preferential void nucleation at particles, in particular aligned zirconium phosphides. In the absence of such particles, very high toughness levels can be maintained after irradiation.

Notes: Abstract Monte Carlo simulations of fatigue in welded joints have been performed using an approach which combines a model for initiation with a multiple surface crack propagation model. The results have been compared with experiments on T-plate welded joints in which initiation and propagation of surface cracks were monitored using potential drop techniques. Predictions of initiation life using a local strain approach were conservative. Despite this underprediction of initiation life, predictions of total fatigue life were very good as a consequence of accurate simulations of propagation life and the fact that initiation represented on average only 12 to 22% of total life, depending on stress level. The initiation model considered variability due to the local weld toe angle and radius, and material strain–life behaviour. The only variability considered in the propagation analysis was the position and timing of initiation events, which leads to variability in coalescence. The underprediction of the variability in propagation and total lives was attributed to the underprediction of initiation life and the fact that out-of-plane coalescence effects were ignored.

Notes: Abstract An experimental investigation into the growth and closure behaviour of small corner cracks has been conducted on a low carbon steel (ASTM A516 Grade 70). This ferritic/pearlitic steel has been chosen to investigate the complexities of small crack behaviour in the dual phase material. Innovative procedures are used for compliance measurement and opening load determination. The closure and growth behaviour of the small corner cracks are presented in terms of shape and size. The results are divided into two stages, the first pertaining to the “stationary crack”, and the second dealing with the crack as it grows. The cracks are shown to exhibit an irregular growth rate, which is in part, shown to correspond with microstructural features of the material. In the short crack regime, the linear fracture mechanics parameter, ΔK does not correlate with the data and the use of a ΔKeff based on the effective applied load provided little improvement.

Notes: Abstract Approximate equations were developed to describe the stress fields in stress concentration regions. A uniaxial tension loading condition was considered. For infinite bodies, the stress equations contain two parameters, the stress concentration factor Kt and the notch root radius ρ; for finite bodies, a finite size correction factor fw which involves the notch depth, t, and the length of ligament ahead of the notch tip, d, was introduced. After the approximate equations reach their minimum value, a uniform distribution is assumed. The accuracy of this description of the stress field was verified by comparison with analytical solutions and the results of numerical analysis. The new equations give better estimations of stress fields than currently available approximate expressions.

Notes: Abstract The structure and mechanical properties of electrophoretically deposited t-ZrO2/Al2O3 lamellar composites are described. The fracture behavior of, and crack paths in, such composites with strong interfaces has been evaluated using indentation and 4-point bend tests. The effect of residual and induced stresses is considered. It is shown that significant crack deflection can be obtained in t-ZrO2/Al2O3 composites by incorporating a number of closely spaced, thin Al2O3 layers.

Notes: Abstract The growth of semi-elliptical cracks emanating from single surface defects under cyclic bending loading conditions has been investigated. Experiments to determine crack shape development during fatigue have been conducted on specimens containing spark-machined starter defects of various shapes and sizes. The results appear to indicate that the size and shape of the initial starter defect only affects the developing crack shape until the crack depth is approximately 20% of the specimen thickness; upon reaching this depth, all crack shapes (independent of initial size and shape) appear to be very similar. A mathematical model, based on the Newman and Raju stress intensity solution for semi-elliptical cracks, has been utilised to predict effectively the shape of the cracks developing from the various single starter defects with aspect (a/c) ratios as large as 3.

Notes: Abstract This paper addresses the problem of the determination of J-intergral from experimentally measured quantities for homogeneous and overmatched single edge notched bending specimens (SENB) or through thickness centre cracked panels loaded in tension (CCT). Commonly, the experimental J-integral is calculated from the area under the load versus load-line displacement curve. Nevertheless, in the case of gross-section yielding, which currently occurs for short cracked specimens or overmatching cases, this methodology mis-estimates the real J-integral. A new proposal, based on analytical considerations and finite element calculations, is made to estimate the J-integral from the area under the load versus CMOD curves.

Notes: Abstract This study was aimed at assessing the influence of stress state upon the life of specimens as well as on the failure mechanisms under thermal cycling. The experiments were carried out on thin-walled tubular specimens (made of austenitic steel). They were loaded in pure tension and in pure torsion as well as under complex loading. A constant value of an equivalent stress was obtained in each case. The influence of stress state on both the specimen life and the equivalent steady strain rate has been found. The Monkman—Grant failure criterion was used as a relationship between time to failure and equivalent steady strain rate. The analysis of the results obtained showed that the low proportion of the shear stress component in the equivalent stress does not change the values of constants in that failure criterion as is also the case for the Norton-Bailey law. The prevalence of the shear stress component is a cause of a significant change in the coefficient of the Monkman-Grant failure criterion while the exponent remains constant for all examined stress states. Initiation and propagation of cracks as well as their surface character were found to be affected by the stress state.

Notes: Abstract—Three different fracture mechanics approaches were applied to two full-scale penstock model tests. The two models were produced, using a Sumiten 80P HSLA steel (minimum yield strength 700 MPa and minimum ultimate tensile strength 800 MPa), in the form of pressure vessels. The first experiment was a burst test performed on a pre-cracked model to determine crack arrest properties. The second experiment was a hydro-pressure test on a model with no cracks and this enabled a post-yield experimental analysis of the undermatched weld metal, when cracks did not initiate. Crack driving forces, obtained numerically, and J-R curves, obtained by the J-integral direct measurement on tensile panels, were used to predict the residual strength. The overall behaviour of a welded penstock under load was analysed on the basis of the results of these three approaches, allowing an evaluation of the significance of cracks.

Notes: Abstract—Fractographic features related to fatigue crack growth in a Ti-6Al-3Mo alloy are studied using compressor disks tested on a hydraulic test bed and which simulate operational multiaxial cyclic loading conditions. The hold-time of a cycle results in the formation of a fracture relief which reflects mainly the two-phase (α+β) lamellar structure of the titanium alloy and a fragmentary fatigue striation formation. Correlation between the number of fatigue striations on the fracture surface and the number of applied blocks of loading (imitating the service conditions of compressor disks) has been obtained. The hold-time duration of the cycle does not affect the crack growth rate and the formation of the fracture relief in this material. An analytic expression is suggested to describe the relationship between fatigue striation spacing, δ, and the stress intensity factor KcI as applied to quarter-ellipse-shaped cracks; it is of the form δ= C[f(τ, FCi)KcI]4, where f(τ, FCi) accounts for the hold-time, τ, and the programmed loading together with their influence on the fatigue crack growth behaviour. The particular threshold value of stress intensity factor (KcI) is established at 20 MPa m. The work indicates that the role of τ manifests itself via a considerable acceleration of crack growth.

Notes: Abstract—Optical fractography was used to estimate growth of small cracks at notches under programmed FALSTAFF loading in an Al-Cu alloy. Crack sizes as low as 25 microns and growth rates over two orders of magnitude could be resolved using this technique. Randomized MiniFALSTAFF load sequence was modified into a programmed load equivalent with major loads either preceding or following marker loads. Crack growth rate under programmed FALSTAFF spectrum as estimated by optical fractography conformed to compliance based estimates on a SE(T) specimen. Long crack growth rates under programmed and randomized MiniFALSTAFF spectrum were essentially similar. Spectrum load fatigue crack growth was studied in central hole coupons under notch inelastic conditions. Scatter in growth rates for small notch cracks was found to be of the same magnitude as that of long cracks. Multiple fatigue cracks are observed at the notch root, and they appear to influence each other.

Notes: Abstract—High temperature low cycle fatigue tests, with and without strain dwells, were conducted at 750°C, 950°C and 1050°C, on single crystal SRR99 nickel base superalloy, with different crystal orientations. At 750°C, SSR99 exhibited cyclic stability regardless of cycle type. The presence of strain dwells caused cyclic softening at 950°C compared with continuous cycling tests. At 1050°C, cyclic softening was observed for all the tests. The introduction of strain dwells produced significant stress relaxation at 950°C and 1050°C, but not at 750°C for the strain ranges in this study. Significant mean stress was observed at the three temperatures for tests with tensile or compressive strain dwells. The crystal orientation was found to have a dominating influence on the cyclic stress strain relationship and stress relaxation response. A simple approach is developed to correlate the effect of orientation on the cyclic mechanical response.

Notes: Abstract— A unified fracture mechanics test method for quasi-static testing of materials is described. It builds on the ESIS Procedures P1 and P2 and introduces additional aspects, such as the δ5 crack tip opening displacement, non-standard CT and SENB specimen configurations, centre cracked tensile specimens, testing of weldments, and guidance for statistical treatment of scattered data.

Notes: Abstract— A version of the endochronic theory of plasticity for the modelling of nonproportional cyclic loading has been developed. To describe an additional hardening of a material a new engineering method for defining a nonproportionality parameter is proposed for a wide class of cyclic strain paths with a prescribed maximum range of plastic or total strains. This parameter makes it possible to establish an unambiguous linear dependence between the cyclic strain path shape and the stress level in a stabilized state. Conjugation conditions have been formulated to describe complex histories of nonproportional cyclic loading. The results of the modelling are shown to be in fair agreement with the experimental data.

Notes: Abstract— Statistical fatigue tests have been conducted on a structural medium carbon steel, S45C, in room air and in 3%NaCl solution, using five cantilever-type rotary bending fatigue testing machines which were specially manufactured for the purpose of the present study. Fatigue life distribution was examined at three and five stress levels in air and in 3%NaCl solution, respectively, and twenty specimens were allocated to each stress level. In room air, it was found that fatigue life distributions followed the three-parameter Weibull distribution, which were closely related to fracture morphology. In 3%NaCl solution, they also followed the Weibull distribution, but the scatter in fatigue life was smaller in comparison to that in air. It is suggested that the decrease in the scatter of fatigue life may be attributed to a smaller fraction of crack initiation life in 3%NaCl solution. The growth of corrosion pits was investigated using a laser microscope. The distribution of corrosion pit depths followed the log-normal distribution, and the corrosion pit depths increased with increasing time or the number of cycles. It was found that the growth of corrosion pits was accelerated by stress cycling and the depths increased with increasing stress level. Based on these results, a growth law of corrosion pits, including the effect of stress cycling, is proposed.

Notes: Abstract— We aim to develop a systematic method of designing structures, by finite element methods, for high cycle fatigue under periodic constant load systems. After having defined a precise terminology, we quickly list those multiaxial fatigue criteria which can be found in the literature. Some criteria, derived from a microscopic approach (Dang Van's, Papadopoulos' and Deperrois' criteria) are extensively presented. The criteria which can be reasonably retained for numerical analyses of structures are underlined and compared to one another. As a conclusion, we describe a high cycle fatigue CAD system which can be derived from this analysis.

Notes: The local approach method is used to calculate the fatigue crack initiation/early crack growth lives (Ni) in high strength structural steel weldments. Weld-toe geometries, welding residual stresses and HAZ (heat affected zone) cyclic mechanical properties are taken into account in the Ni estimation procedure. Fatigue crack initiation lives are calculated from either a Basquin type or a Manson-Collin type equation. The local (HAZ) stress and strain amplitudes and the local mean stress are determined from an analysis based on the Neuber rule and the Molski-Glinka energy approach. The accuracy of the different methods is evaluated and discussed. Finally the previous methods are used with HAZ cyclic mechanical properties estimated from hardness measurements.

Notes: This paper presents analytical and numerical solutions to the analysis of welded specimens when loaded in three-point bending, and compares the results with those obtained experimentally. In each case the crack is located within the weld material, and runs parallel to the weld. Two analytical models are presented for deriving limit loads using slip-line field theory. Due to the welding process, the material behaviour in the weld-base interface is complex, and this is described in the analytical solutions using a number of material zones. The analytical solutions also provide the η and d∞ coefficients which are used to determine the J and crack tip opening displacement (CTOD) parameters. These solutions are then compared with numerical results obtained using the finite element method.Good agreement is obtained between the numerical and analytical results, and it is shown that for overmatched specimens a better analytical solution can be obtained by using a slip-line field geometry which passes predominantly through the base material. When the analytical η solutions are applied to the experimental results it is shown that, in the case of undermatched weldments, J can be used as a fracture characterising parameter, but it cannot be used in the case of an overmatched specimen.

Notes: A previously proposed single crystal hardening law is applied to the prediction of responses of polycrystalline material under non-proportional cyclic loading. In this paper, the Kroner, Budiansky and Wu model is adopted and the relevant numerical schemes for both the iteration related to the non-proportional loading paths and the search of active slip systems are established. Two typical engineering materials: oxygen-free, high-conductivity (OFHC) copper and 316 stainless steel, which differ greatly from each other in microstructure, are used for predictions and comparisons with experiments. Loading paths include the symmetric tension-compression cycle, the circular cycle and the rectangular cycle. The behaviour of 316 stainless steel, at both room and elevated temperature is modelled. Comparisons show that the predictions are in quantitative agreement with the corresponding experiments for all the cases mentioned above. In addition, comparisons of different single crystal hardening laws are also presented.

Notes: Abstract— A study has been undertaken into short crack growth behaviour of AISI type 316 stainless steel under creep-fatigue conditions at 550°C within the high strain range of 0.9 to 2.5% and including a 60 min hold-time. During the high-temperature, reverse-bending tests, surface crack initiation and growth on both the tensile-hold and the compressive-hold sides of circular-section specimens were monitored by means of a plastic replication technique. Detailed analysis revealed that under creep-fatigue conditions, the initiation and growth behaviour of many individual cracks and their subsequent coalescence to form a major Stage II (tensile) crack was the dominant feature in the failure process. A life prediction model is proposed which incorporates the process of short crack coalescence. Satisfactory predictions of creep-fatigue lifetimes are derived from the model.

Notes: Abstract— A study of the dependence of the fracture toughness on the microstructural morphology of polycrystalline materials is presented. The microstructure of several aluminas was characterized by thermal etching and image analysis. The fracture toughness of these materials was then determined by bridge-indentation pre-cracking followed by a four-point bending test procedure. A SEM investigation of the mechanisms of crack propagation was also carried out. The observed microstructure-related fracture toughness responses are discussed.

Notes: Abstract The dual boundary element method for the analysis of cracks in linear elastic materials has been previously generalised by the authors to allow for automatic remeshing when crack tips intersect other cracks or boundaries, and initiation and growth of small cracks at positions of high stress concentration. The new cracks are assumed to result from sudden events such as an overload or the subsequent stress redistribution when cracks intersect other cracks or holes.In this paper a crack at the edge of one hole in a row of pin-loaded holes is investigated; various values are considered for the stress at which new cracks may initiate. Two rows of aligned or staggered holes are examined also. The spacing between the holes was typical for lines of holes in overlap joints in plates. For the same load transfer between the plates, new cracks are initiated less readily and grow more slowly for a double row of aligned holes than for a single row or for staggered rows.

Notes: Abstract The service life of gears with a crack in a tooth root can only be determined by numerical methods. An initial fatigue crack is assumed on the tensile side of a tooth root at the site of maximum reference stress and is assumed to commence growth perpendicularly to the surface. An initial assessment can be made using a two dimensional analysis, which is relatively quick and cheap to perform. However, if we wish to take into account the influence of the contact area of load transfer, which can be distributed along the tooth width in different ways, the gear has to be treated by a three dimensional finite element analysis. Crack profile advance is made in stages, each using the strain energy release rate criteria and giving ultimately the stress intensity factor as a function of average crack depth. With known fracture mechanics material characteristics of different gear layers, through which the crack propagates, the service life of a gear is then determined by numerical integration of Paris' equation. A one-sided contact area causes the crack to propagate several times faster than the preferred load distribution across the middle of the tooth.

Notes: Abstract Fifteen kinds of mullite/SiC samples with different microstructures were prepared in order to examine the effect of Sic volume% and Sic grain size on mullite morphology and mechanical properties. Special attention was paid to the effect of heat-treatment on fracture stress. It is shown that these materials have damage self-healing characteristics. The best mullite/SiC system, within the given test conditions, is 20% by volume of Sic, having a grain sue of 0.56 μm, and the best condition for damage healing is a 1 h heat treatment at 1300°C in an air atmosphere.

Notes: Abstract Micromechanical fracture-toughness models are applied to experimental results for a metal-matrix composite (2009/SiC/20p-T6) to understand the temperature dependencies of toughness and fracture mechanisms, as well as to test quantitatively a continuum fracture-mechanics approach. Models which couple the crack-tip strain field, characteristic fracture-process distance and measured intrinsic micro void-fracture resistance predict the temperature dependencies of fracture-initiation (KJICi) and crack-growth (TR) toughnesses from 25°C to 316°C. The temperature dependencies of KJICi and TR result from the interplay between the fracture resistance and the crack-tip strain field, each being temperature-dependent. Strain-based models are equally valid for void nucleation- or growth-controlled fracture. A scenario for fracture is nucleation-controlled damage within Sic-particle clusters, corresponding to KJICi, followed by cluster-damage growth to coalescence under increasing stress intensity. Void growth is stabilized increasingly at elevated temperatures.

Notes: Abstract The loads for yield of plates and cylinders with semi-elliptical surface flaws have been evaluated by a weight-function method incorporating stress intensity factors using results from finite element calculation. A new weight function for loads partially distributed along a chord is proposed. The yield load is defined as that at which the plastic zone first reaches the back surface. The results show that the effects of the cylinder geometry parameter t/R on yield load is small, and it is proposed that the plate results be used also for cylinders over the range of geometries 0 ≤t/R≤ 0.1.

Notes: Abstract A study has been conducted on the initiation and growth of fatigue cracks in the three principal directions of an aluminium alloy 2024-T351 plate tested under stress control (R, minimum to maximum stress =– 1). Early and multiple fatigue crack nucleation from broken Al7Cu2Fe second phase particles resulted in shorter lives for the longitudinal direction specimens in the medium to long life regime. Although fatigue cracks nucleated in large surface grains, rather than at broken particles, the lives of the short-transverse direction specimens were marginally longer. Cracks also nucleated in large surface grains in the transverse direction specimens, yet the average fatigue lives were about twice as long. This was the consequence of wider slip bands and fewer initial micro-cracks.

Notes: Abstract Stress intensity factors for circumferential surface cracks in pipes have been derived using the finite element method. Both cracks located at the in- and outside of the pipes have been analysed. The derived solutions cover a wide range of geometry and load configurations and are presented in a tabular form that defines influence functions for the stress intensity factor along the whole crack front. The solutions show good agreements in comparisons to other published solutions.

Notes: Abstract A fundamental understanding of dynamic delamination in composites is sought through the application of theoretical and experimental approaches familiar to dynamic fracture mechanics. Analysis of steady-state fracture in an infinite orthotropic strip yields a simple solution which can be used to evaluate numerical procedures and experimental results. The analogous specimen consists of a single edge notched composite strip bonded to stiff steel substrates to enforce the desired displacement boundary conditions. Delamination velocities of the order of 10 to 1000 m/s were measured using a graphite gauge technique. Quasi-static and dynamic finite element methods are applied to investigate the behavior of the specimen and to determine static initiation and dynamic delamination toughness. The experimental observations cannot be explained by linear elastic fracture theory. The absence of a unique G(ȧ) relationship might be rationalized by a simple model relating matrix crack zone size to fiber bridging mechanisms.

Notes: Abstract In brittle-matrix composites cracking of the matrix is often accompanied by bridging of the crack surfaces. The bridging will reduce the net stress intensity factor at the crack tip and consequently increase the toughness of the composite material. The bridging mechanism is due to for example unbroken whiskers, fibres, ductile particles or interlocking grains.Analysis of the bridging mechanism in cracked structures is conveniently carried out using the concept of cohesive zone modelling. In this case the action of the bridging elements is replaced by a distribution of forces, so called cohesive forces trying to close the crack. The commonly used approach in such modelling has been to replace the action from individual bridging elements by a continuous spatially independent distribution of closing tractions whose magnitude is a function of the crack opening displacement only.In this paper the influence of the spatial distribution of bridging elements is considered for plane crack problems. The cross section of the bridging elements is assumed to be circular and the distance between the different bridging elements is determined by the volume fraction, the radius and the geometrical distribution of the bridging elements.Damage resistance curves have been calculated for typical whiskers-reinforced ceramic composites, and the results from the present spatially dependent models are compared with results from calculations with spatially independent models. The influence of the radius of the bridging element, the volume fraction of whiskers and the material properties are illustrated and the use of spatially independent models is discussed.

Notes: High-cycle-fatigue/creep experiments were performed on a 9%Cr-1%Mo temperered marten-site ferritic steel at 873 K in air. The stress ratio R=σmin/σmax ranged from-1 (“pure” fatigue) to 1 (“pure” creep). The maximum stress σmax was kept constant at 240 MPa. The lifetime depends on the stress ratio R in a non-monotonic way.In the stress ratio interval 0.6 〈 R 〈 1.0 both the creep strain rate and the lifetime are controlled by mean stress σmassof the stress cycle. In the stress ratio interval — 1 〈 R 〈 0.2 the lifetime is controlled by the stress amplitude na. The fatigue/creep interaction occurs in between these intervals.The fatigue/creep loading induces transformation of the tempered martensite ferritic structure into an equiaxed subgrain structure. The resulting subgrain size depends strongly on the stress ratio.

Notes: A high speed steel, processed by two powder metallurgy routes and heat treated to give a range of microstructures, was investigated in 4-point bending at room temperature using smooth and precracked specimens. The finer microstructures were in the material from gas atomised powder which was hot isostatically-pressed, commercial ASP60 alloy, while the coarser microstructures derived from laboratory vacuum sintering of water-atomised powder. The resultant hardness values Hv50 were in the rage 780 to 1050, prior autenite grain sizes, 5 to 25 pm and maximum carbide sizes, 6 to 32 μm.Only some of the uncracked samples exhibited macroscopic yielding, at stresses in the range 1.64 to 2.59 GPa; the finer microstructures being asSociated with the higher strengths. Macroscopic plastic deformation never exceeded 0.33%; fracture strengths were in the range 1.46 to 2.75 GPa. Fracture toughness, Klc, varied from 12 to 17 MPa√m in the Hy50 range 920 to 800 for the directly sintered steel and only from 10 to 12 MPa√m in ASP60. The insensitivity of K1c to macroscopic hardness in ASP60 is asSociated with the plastic zone size of 1.5 μm which approximates to the average carbide spacing.Nucleation and growth of natural, i.e. stress-induced, microcracks in un notched specimens was studied by surface replica microscopy. Crack nucleation took place at stresses between 0.5 and 1.5 GPa, i.e. below those for yielding and for fracture, σF, and was by debonding of inclusions (alumina and calcium-alumino silicates) or cracking of carbides. In the coarsest microstructure monotonic stepwise subcritical crack growth was observed from stress levels of ∼1.3 GPa, i.e. ∼0.8 to 0.9σF. Similarities to the behaviour of short fatigue cracks in metallic materials and the R-curve behaviour of ceramics are referred to.

Notes: Fatigue damage in two austenitic-ferritic duplex stainless steels, with the structure of a natural composite and different levels of nitrogen content, was studied in low-cycle fatigue. Both steels show initial cyclic hardening followed by softening and a long stabilisation period. The cyclic stress-strain curve increases with the nitrogen content while Manson-Coffin curves of both steels intersect at medium fatigue lives. The study of the surface relief reveals intensive slip markings both in ferrite and in austenite. Their density is influenced by the nitrogen content. Both the intensity and density of the persistent slip band (PSB) markings are higher in the ferrite. Crack initiation was found to appear predominantly in PSBs in the ferritic grains at the low strain amplitudes, and in the ferritic and austenitic grains at the highest strain amplitudes. The level of the cyclic stress-strain response and the fatigue lives are discussed in terms of the cyclic strain localisation and of the effect of texture and nitrogen content on the strength and fatigue damage. The increased strength of the austenitic phase, due to high nitrogen alloying, results in cyclic slip localisation in the ferrite, and the decrease of fatigue life, compared with the steel with the lower nitrogen content.

Notes: Experimental determination of the fatigue endurance (S-N) curve of either a given material or a machine element imposes the choices of sample size and of stress levels to be tested. Since referenced recommendations do not apply when the data include run-outs or have variable scatter, this paper studies the confidence of fatigue tests, taking into account the effect of sample size and of a statistical model. In this research, a great number of rotary bending fatigue tests were carried out, using a heat treated carbon steel with relevant scatter of fatigue lives. The statistical analysis allowed one to obtain confidence limits of the S-N diagram estimates and to propose some criteria for a better formulation of test schedules.

Notes: Abstract A high cycle fatigue model for the prediction of component lifetime in elastic rolling contacts is developed and applied. Varying magnitudes and positions of the contact loads are described by use of discretized statistical distributions. Longitudinal and lateral adhesion are included. The Hertzian contact pressures are analytically found, and the corresponding subsurface stresses are calculated using a numerical integration scheme starting from the exact point force solutions of Boussinesq and Cerruti. Triaxial fatigue with rotating directions of principal stresses is studied using the Dang Van fatigue initiation criterion, together with the Palmgren-Miner damage accumulation law. The full model has been implemented in a computer code. A wheel/rail contact problem is treated and the results are compared to previous numerical and experimental data.

Notes: The whole fatigue domain has been divided into six Werent zones, each governed by a separate fatigue regime. Some of these regimes coincide with a known classification of fatigue, and others are new regimes, where new prediction methods have been introduced. The proposed predictions are to be considered as practical curve fitting relations for special cases, but those cases can also be fairly general and useful for design purposes. The proposed fatigue diagram can be used as a basis for a quantified explanation of several known fatigue phenomena.

Notes: A high-cycle fatigue criterion suitable for multiaxial non-proportional stress loading is proposed in this paper. The criterion is based on some microscopic considerations related to the crystalline structure of metals. The purpose of the present paper is mainly the application of this criterion in two loading cases: (a) biaxial loads involving two normal stresses or one normal and one shear stress, and (b) triaxial load with two normal stresses and one shear stress. Stress states of these kinds are very common in piping assemblies. Application of the proposed criterion in the case of triaxial loading, where the three stress components are of the same frequency, but out-of-phase, leads to a simple analytical formula. This formula is the equation of a bounding surface that delimits in the space of the above three stresses the safety domain against fatigue crack initiation. A remarkable theoretical result concerns the phase difference of the shear stress, which does not appear in the derived formula. Consequently, according to our proposal the safety domain (i.e. the limiting fatigue endurance) under combined out-of-phase biaxial normal stress loading and torsion is independent of the phase difference of the torsion. Obviously this result holds also for the simpler case of axial load and torsion. On the contrary the phase difference between the two normal stresses has a strong detrimental effect on the fatigue endurance of a metal. As is shown these theoretical conclusions are in good agreement with fatigue limit test data found in the scientific literature.

Notes: In order to evaluate the notch fatigue strength and notch sensitivity of aluminum-lithium, 2090 and 8090, alloys, rotary bending fatigue tests have been carried out using circumferentially notched specimens with different stress concentration factors. The results were compared with those of traditional aluminum, 2024T4 and 7075-T6511, alloys. It was found that 2090 and 8090 alloys showed superior notch fatigue strength in comparison to the conventional aluminum alloys. The notch sensitivities to the crack initiation limit of the aluminum-lithium alloys were lower than those of 7075-T6511, while they were nearly equal to those of 2024T4 for blunt notches. The notch sensitivities to the crack propagation limit were also lower in aluminum-lithium alloys, in particular the 8090 alloy, than in the conventional aluminum alloys. It was suggested that the decreased notch sensitivities of the aluminum-lithium alloys were attributed to both the crack propagation mode and the excellent propagation resistance related to their microstructures.

Notes: Abstract— Imitating Garwood's 3-parameter technique, an experimental parameter Jmax was introduced to predict fatigue crack growth rate (da/dN) over a wide range including small scale yielding and large scale yielding. It was found that for a ΔK-increasing fatigue test condition, Jmax is a valid parameter. A significant crack growth acceleration, caused by a transition of fracture mechanism, occurs when Jmax=JIC The fracture mechanism involving striation formation when Jmax 〈 JIC becomes ductile tearing when Jmax 〉 JIC Equations to predict the effect of stress-ratio on Jmax as well as on da/dN are given.

Notes: Abstract— An analysis is made of striation spacing measurements as observed on fatigue crack surfaces of specimens and longerons of helicopter rotor-blades made from the aluminium alloy AVT. It is shown that, for the combined cyclic tensile stresses at various R-ratios and static rotation angles, α, striation spacings were controlled by an equivalent intensity factor Ke=K1F(α, R). Values of the F(α) for Ri= const were computed.The crack growth life in terms of the number of flights and stress equivalents were computed for various longeron sections of the helicopter rotor-blades which had failed in service.〈section xml:id="abs1-1"〉〈title type="main"〉SUMMARY AND CONCLUSIONS(1) A fatigue striation spacing analysis was made from measurements on AVT Al-alloy specimens tested in air under combined cyclic tension and static-rotation loads at various a-angles of rotation, and stress ratios, R. Fractographic analyses have shown similarities in terms of striation spacing formation in the investigated range of α-angles and R ratios.(2) The similarity in the fatigue fracture mechanisms attested to the possibility of characterizing the striation spacing formation by a unified kinetic diagram via a Ke= K1F(α, R) approach for different α-angles and R ratios. The function F(α, R) was determined and it can be used to calculate the tensile stress equivalent according to values of σi measurements taken from components broken in tests or in service.(3) It was a longeron of the rotor-blade of helicopters manufactured from AVT Al-alloy broken in service and tested cyclically that were investigated fractographically. An analysis of the formation of striations and their spacing has shown similarities with the investigated longerons fatigued in tests and in service. The crack growth life versus number of flights and stress equivalent was calculated according to values of the σi measurements.(4) It was shown that a fatigue crack will not nucleate in longerons in service if the depth of defects on the external surface have a size less than 0.3 mm. The stress for tested longerons is near to the calculated equivalent for longerons that failed in service.

Notes: Abstract— The ACPD-method was used to measure crack growth in low cycle fatigue tests at elevated temperature. The material which was investigated at T= 550°C was the ferritic steel 2.25Cr–1Mo. Characteristic values of representative hysteresis loops were determined. The crack growth data were correlated with the effective part of the fracture mechanics parameter ΔJ. A new method for the calculation of ΔJ is proposed.

Notes: Abstract— The effect of a laser surface treatment on the high-cycle fatigue behaviour of the austenitic stainless steel AISI 316L was investigated. The specimens were subjected to a surface melting using a CO2-laser. Although this treatment introduced tensile residual stresses in the melted and resolidified surface layer and intensified the surface roughness, the fatigue limit could be increased by 20% with respect to the as-received specimens. It is suggested that grain refinement in the rapidly resolidified surface layer is mainly responsible for this improvement.

Notes: Abstract— The cyclic behaviour and micromechanisms of 3Y-TZP ceramics were investigated at a stress ratio, R, of 0.1 and a frequency of 0.1 Hz at room temperature for both nanocrystalline (100nm) and submicron polycrystalline (0.35μm) 3Y-TZP ceramics. The nanocrystalline material displayed cyclic softening, but the submicron polycrystalline material displayed cyclic hardening. The cyclic hardening resulted mainly from saturation of tetragonal to monoclinic phase transformation which induced a higher internal stress in the material; the cyclic softening resulted partly from the formation of microcracks along the grain boundary of the tetragonal phase.Superplastic deformation was first found by the Scanning Electron Microscope and the Atomic Force Microscope (AFM) in some micro-areas of the fatigue fracture surface of the nanocrystalline material. Furthermore it was observed by AFM that there were a large number of slip lines and an early stage for the formation of an extrusion on the side edge of the fracture. From this it was deduced that the micromechanism of local superplasticity resulted from the slip motion of dislocation.

Notes: A series of strain-controlled, low-cycle fatigue experiments have been conducted on 42CrMo steel under various loading paths including circular, square, cruciform, and rectangular paths. Present experiments have shown that there is additional hardening under non-proportional cyclic loading. Non-proportional cyclic additional hardening also results in a shorter life for multiaxial low cycle fatigue. A non-proportionality measure of strain path based on both a physical basis and macromechanical phenomena is proposed. The loading path effect on additional hardening is also described well. Low-cycle fatigue damage accumulation and the evolution process under non-proportional loading is analysed via the Continuum Damage Mechanics Model of Chaboche. A non-proportinality measure is introduced in the damage evolution equation and a modified Coffin-Manson type formula is derived. A novel fatigue life prediction approach based on the critical-plane concept of Brown and Miller is proposed.

Notes: The analytical solution of Sadowsky and Sternberg for the stress field around an biaxial elliptical cavity in an infinite elastic body was carefully examined. For uniaxial tension, it was found that the stress concentration factor varies very slowly along the equator of the cavity, regardless of the shape of the cavity. As was found for two dimensional notch problems, the normalized stress fields around cavities with different shapes are virtually identical within the stress concentration region and differ only slightly beyond that region. This fact permits simple, yet very accurate approximate stress equations to be developed for three dimensional notch problems by modifying the previously obtained solutions for two dimensional notch problems.

Notes: The effects of repeated point contact loading of a toughened ceria-stabilised zirconia ceramic, CeTZP, by softer metallic cones, have been investigated. All tests were conducted on a purpose-designed and built computer-controlled apparatus. The results clearly indicate that zirconia is susceptible to cyclic loading. The evolution of the cyclic deformation and cracking processes have been analysed in terms of contact pressures, applied loading conditions and number of cycles. The ground tips of both quenched, and quenched and tempered, silver steel cones are plastically deformed during the initial loading cycle to produce a conforming planar surface approximately 50-100 μm in diameter. In all cases, plastic deformation in the CeTZP was observed within, and adjacent to, the contact zones. The degree of plastic deformation increased as the number of cycles increased. The expansion associated with the tetragonal-to-martensitic phase transformation in the zirconia caused granular lifting from the surface, at the edge of the contact zone. This resulted in localised spalling at the peripheries of the contact zones, where the tensile stresses are the highest, followed by a radial expansion at higher cycle numbers. This investigation has identified that the rate of evolution of the fatigue damage, in decreasing order of significance, was controlled by: (a) cone hardness, (b) cyclic load amplitude and (c) mean load.

Notes: The mechanism of “475°C embrittlement” in age-hardened ferritic stainless steel, E-Brite and A129-4, is investigated. Experimental results for smooth tensile and notched bending fracture tests are interpreted using a finite element simulation of the stresses at fracture. Yield is characterised by profuse slip band formation. Transgranular fracture initiation is observed at slip band intersections with grain boundaries. Deformation twinning occurs during brittle fracture. Slip bands and deformation twins are identified using lattice rotations measured with electron back-scatter diffraction patterns. Mechanisms for the ductile-to-brittle transition are discussed.

Notes: Abstract— In response to the increasing structural applications in duplex steels for welded structures, fatigue behaviour of a SAF 2304 grade duplex stainless steel was investigated, considering both the base metal and GTAW welded joints. Fatigue curves and fatigue limits under rotary bending fatigue were obtained. The study focused attention on the microstructural features of fatigue crack propagation of the two series of experiments, thereby permitting an evaluation of the tortuous crack path of welded joints and the mechanisms related to threshold microstructural barriers.

Notes: Abstract— Constant amplitude fatigue tests with welded specimens under fully reversed four-point bending as well as under axial loading have shown that vibration stress relieving does not lead to a fatigue life improvement of welded parts when compared to the as-welded state. Thus, a substitution of thermal stress relieving by a vibration treatment is not successful. This was also proved by residual stress measurements in the welded parts studied in this paper.

Notes: Abstract— The use of the hot spot stress approach to the fatigue analysis of welded components is briefly described. Results are presented of fatigue tests on arc welded steel joints (C-Mn and stainless), carried out at Lappeenranta University of Technology between 1980 and 1993, based on the hot spot approach. Based on experimentally-measured hot spot strains, the fatigue capacities of around 100 specimens of C-Mn steel joints, and 80 stainless steel joints, were found to be consistent. The fatigue class FAT 100, or in many cases FAT 112 or higher, can be used as the design hot spot fatigue strength for toe failure of welded joints of moderate thickness, i.e. up to 10 mm. A sharp transition at the fusion zone from the base metal to the weld was clearly shown to be detrimental, leading to a fatigue capacity below average. The log, value of the standard deviation of fatigue life, or the fatigue capacity (Δσ3N), was typically 0.13 within a series of C-Mn joints. Statistical analysis of all test data concerning weld toe failure gave a standard deviation of 0.24. By considering all the specimens in one series, a mean fatigue strength of FAT 148, and a characteristic fatigue strength of FAT 107, were obtained.

Notes: Abstract— This report presents a brief outline of three papers on stress concentration factor formulae, which have been frequently referred to in Japanese technical literature. Most well known is a book entitled “Stress Concentration” [1] while Refs [2,3] have often been quoted for cases of fillet joints. The stress concentration factor covered in this report is the so-called theoretical, elastic stress concentration factor. A discussion on plastic stress concentration factors is not in the scope of this report.

Notes: Abstract— The toughnesses of specimens of heterogeneous multipass weldments with various ratios of weld metal to base metal can be characterized by a single value of the local fracture stress within a narrow scatter band. Relationships were found between the normalized value of this parameter and the lower bound values of fracture toughness parameters such as the fracture load, the energy absorbed, and the crack opening displacement. These facts indicate that using such a toughness parameter, it is possible to assess a heterogeneous multipass weldment by a single value; within a narrow scatter band. Further experiments indicate that this parameter could be measured on specimens simulating the microstructure of the weakest zone where cleavage cracking is initiated.

Notes: A model based on micro-mechanical concepts has been developed for predicting fatigue crack growth in titanium alloy matrix composites. In terms of the model, the crack system is composed of three zones: the crack, the plastic zone and the fibre. Crack tip plasticity is constrained by the fibres and remains so until certain conditions are met. The condition for crack propagation is that fibre constraint is overcome when the stress at the location of the fibre ahead of the crack tip attains a critical level required for debonding. Crack tip plasticity then increases and the crack is able to propagate round the fibre. The debonding stress is calculated using the shear lag model from values of interfacial shear strength and embedded fibre length published in the literature.If the fibres in the crack wake remain unbroken, friction stresses on the crack flanks are generated, as a result of the matrix sliding along the fibres. The friction stresses (known as the bridging effect) shield the crack tip from the remote stress, reducing the crack growth relative to that of the matrix alone. The bridging stress is calculated by adding together the friction stresses, at each fibre row bridging the crack, which are assumed to be a function of crack opening displacement and sliding distance at each row. The friction stresses at each fibre row will increase as the crack propagates further until a critical level for fibre failure is reached. Fibre failure is modelled through Weibull statistics and published experimental results. Fibre failure will reduce the bridging effect and increase the crack propagation rate.Calculated fatigue lives and crack propagation rates are compared with experimental results for three different materials (32% SCS6/Ti-15-3, 32% and 38% SCS6/Ti-6-4) subjected to mode I fatigue loading. The good agreement shown by these comparisons demonstrates the applicability of the model to predict the fatigue damage in Ti-based MMCs.

Notes: This paper presents J-integral solutions for geometries with shallow edge cracks (0.05 ≤a/W≤ 0.1) subjected to combined tension and bending. Tension and bending is applied simultaneously and sequentially. Curves of J versus structural strain are obtained from two-dimensional plane strain finite element analyses using medium to low work hardening power laws to represent material behaviour. Empirical equations are derived from the finite element results to estimate the J-integrals, using previously developed J-estimation schemes for SENT and SENB specimens with only minor modifications. The predictions presented are invariably safe, but due to the complexity of the problems studied and the variability of the loadings considered, the degree of conservatism is found to be high in some cases.

Notes: A study has been made of the effect of short periods of overloading on the environmental hydrogen induced fracture (HIF) life of 0.42% C, 0.87% Cr, 0.2% Mo steel tested in a 0.5 mol/L H2SO4 solution under continuously hydrogen charging conditions. Experimental results showed that when the overloading was applied during the early or middle stage of the test, the HIF life was longer than that obtained at constant stress; however, if the overloading was applied during the later stages, a shortened HIF life was obtained. It is important to note that the processes of HIF (including hydrogen absorption, transportation and accumulation, crack initiation and propagation) depend not only on the electrochemical condition, but also on both stress-strain state and stress history. In view of the above considerations, effects of plasticity induced closure, residual compression stress, dislocation shielding and overload damage, which control HIF life, are discussed.

Notes: When a body subject to contact loads develops a crack, the origin, orientation, and tendency to propagate or not depend on fabrication details and maintenance history as well as the service environment. Research over the past twenty years has led to better characterization of crack behavior in some contacting bodies (railroad rail and wheels). Computational methods in solid mechanics provide the basis for quantitative results, but true understanding also requires application of materials science principles. The two disciplines are being brought to bear in combination; quantitative assessments of potential manufacturing improvements and prediction of effects on performance in service are now within reach.

Notes: The novel experiment developed in our Institute to investigate crack initiation, rapid crack propagation and crack arrest with one specimen, the ring test, was applied to a ferritic HSLA pipeline steel. The maximum crack speed achieved in these experiments was between 230 m/s and 1165 m/s. The fracture toughness at crack arrest, KIa, was determined by a static analysis of this specific test. In all cases, it was found that KIa was much lower than KIc. The values of KIa decrease when the maximum crack speed increases, the results being largely scattered. The fracture toughness at crack arrest is therefore not an intrinsic parameter of the material for a given temperature. Cleavage fracture obtained under these conditions is characterized by the existence of numerous cleavage microcracks, mechanical twins and unbroken ligaments. The decrease in fracture toughness when crack speed increases is related, using the Beremin or the RKR model, to the high strain rates at the tip of a rapid propagating crack. A model which takes into account the effect of unbroken ligaments left in the wake of a propagating crack is developed to account for the large values of KIa which were occasionally measured.

Notes: Creep and cyclic deformation behavior of two lead-free high temperature solder alloys, 95Sn-5Ag and 99Sn-1.0Cu, a high lead alloy 97.SPb-1.SAg-1.0Sn, and an Ag-modified eutectic alloy 62.SSn-36.1Pb-1.4Ag, were studied. Room temperature and high (100°C and 150°C) temperature fatigue tests (with cyclic strain amplitude up to 6.0%) for the four solders were conducted, with the fatigue lives ranging from a few cycles to more than 100,000 cycles. It is shown that among the alloys studied, 62.SSn-36.1Pb-1.4Ag (the modified Sn-Pb eutectic alloy) has the lowest fatigue resistance in term of low cycle fatigue life (strain controlled). The high lead alloy, 97.SPb-1.5Ag-1.0Sn, has the highest strain fatigue resistance in the large strain region (Δ 〉 2.0%). Temperature has a significant effect on alloys 95Sn-5Ag and 99Sn-1.0Cu, but has a negligible effect on the Ag modified Sn-Pb eutectic alloy 62.5Sn-36.1Pb-1.4Ag and 97.5Pb-1.5Ag-1.0Sn. Creep studies show that these alloys generally have a very significant primary creep regime (up to 20%); thus, any realistic constitutive relation has to take such a primary creep phase into consideration. Cyclic deformation of alloy 95Sn-SAg was simulated by using a constitutive relation built upon a 2-cell model, which covers both primary and secondary creep. This model provides a good estimate of the peak stresses (the minimum stress and the maximum stress in each cycle); it agrees with experimental results when the applied cyclic strain is small and/or the applied strain rate is very low.

Notes: Abstract— A series of low cycle fatigue experiments have been conducted on a 42CrMo steel under tension-torsion loading. Thin-walled tube specimens were used. Low cycle fatigue under various loading paths, including circular and square paths, have been investigated.The plastic work criterion for low cycle fatigue failure has previously been generally accepted, but it is difficult to calculate stress and strain for complex loading paths, especially for non-proportional loading. This present study suggests a simple method for the calculation of the stable cyclic stress and strain values based on a Modified Endochronic Constitutive Theory (MECT) that redefines an intrinsic time scale. The loading path effect under non-proportional loading is also considered when evaluating fatigue life.The results show that the plastic work approach using the MECT method in multiaxial fatigue calculations correlates reasonably well the data and is a reflection of loading path dependence.