ALBERT

Notes: Stress intensity factors are determined for internally and externally cracked, pressurized thick cylinders with partial autofrettage (less than 100% overstrain). The solutions are based on a superposition of existing solutions which does not involve any loss of accuracy. Implications of the stress intensity factor results for the safe-life design of pressure vessels subjected to fatigue are discussed.

Notes: Fatigue crack propagation tests were conducted under conditions of equibiaxial, uniaxial and shear loading by using a cruciform specimen in a servo hydraulic testing machine. The effect of non-singular stress cycling on the fatigue crack propagation rate was examined based on the observation of crack opening behavior. The crack propagation rate was significantly influenced by the non-singular stress parallel to the crack when it was correlated to the stress intensity range. The crack closure behavior was greatly affected by the non-singular stress. The crack propagation rate was uniquely correlated to the effective range of the stress intensity factor except for the case of completely reversed shear loading where significant plasticity was detected. The crack opening displacement range was concluded to be a parameter controlling the crack propagation rate for all the stress conditions examined in the present experiments. Some discussion is made on the effect of material anisotropy on fatigue crack propagation in a biaxial stress field.

Notes: The paper presents a comprehensive review, supplemented by original data, of the engineering fatigue behaviour of copper. Variations in manufacturing route and softening treatments are shown to have little effect on the fatigue of annealed copper but the high cycle fatigue strength is increased by cold work. The high strain fatigue behaviour is defined in terms of the plastic strain range and the cyclic stress-strain characteristics are documented. Fatigue behaviour in bending and torsion is defined by data and related to that in tension by simple design rules.Notches are found to reduce the laboratory measured fatigue strength of copper by ∼ 30% and the effect of surface finish, surface distortion and surface residual stress is defined in the literature. Fatigue crack growth is defined in terms of stress intensity factor range ΔK by an upperbound law and, together with the conditions for non-growth (ΔK0), shown to relate to the equivalent conditions for steels via the ratio of the respective elastic moduli.The effect of environment on the fatigue of copper has received scant attention in the literature, such results as exist suggesting little if any reduction in strength to be brought about by gaseous or aqueous environments. The most dramatic change is the improvement of about an order of magnitude which results when tests in vacuum are compared with equivalent tests in air. Results of fatigue tests on copper in ammoniacal environments are conspicuously absent from the literature.As the test temperature is reduced below room temperature there is a predictable increase in high cycle fatigue strength, a reduction in fatigue strength occurring above room temperature. High strain fatigue test results presented in terms of plastic strain range appear insensitive to temperature although at very low strain rates and high temperatures a reduction in fatigue strength occurs. A linear life fraction cumulative damage creep-fatigue law appears sometimes to be non-conservative but much more testing is needed to evaluate fatigue damage summation laws generally for copper.Numerical data are given in support of all the aspects of the engineering fatigue behaviour reviewed in the paper.

Notes: The birth and growth of short cracks is analysed from an elastic-plastic fracture mechanics viewpoint. Low to high cumulative damage tests from the low stress to high strain regime indicate that there is no crack initiation period in the metallurgical sense and that cracks grow from the first cycle, but at a slow rate. The initiation phase terminates when one crack starts to dominate and accelerates to failure, its initial size being given by 〈displayedItem type="mathematics" xml:id="mu1" numbered="no"〉〈mediaResource alt="image" href="urn:x-wiley:8756758X:FFE263:FFE_263_mu1"/〉 for the medium carbon steel tested here, of grain size 56 μm.

Notes: Fractographic analyses have been used to explain the cyclic crack growth behaviour of A533-B, Ducol W30, a C-Mn steel and type 304 stainless steel in simulated light water reactor environments at ambient temperature. Fractographic observations have offered an explanation for anomalous crack growth behaviour and have also indicated where micro structural or environmental variables dominate in producing certain fracture modes and crack growth rates. An understanding of the operative corrosion fatigue mechanisms has been formulated through these fractographic analyses. Environmental crack growth in the ferritic steels has been described by a model involving both anodic dissolution and hydrogen embrittlement. Conditions where only one of these mechanisms would dominate have been identified and limits to their effect postulated. A crystallographic mode of failure observed in the austenitic type 304 stainless steel has also been explained by a selective dissolution process.

Notes: Abstract —Low-cycle fatigue properties were investigated on four carbon steels and five low alloy steels specified in JIS (Japanese Industrial Standard) for machine structural use, which are the most commonly used in Japan. Several different heats from each of several representative manufacturers were sampled so as to represent the average fatigue characteristics of current materials. The cyclic deformation behaviour of material was denned by comparing the monotonie yield stress on the extrapolated tensile work hardening curve with the cyclic yield stress in the cyclic stress-strain curve determined by incremental step test. The normalized ferrite-pearlitic steels cyclically hardened, while the quench-tempered martensitic were cyclically stable or softened. The S–N relations derived from the strain-controlled low-cycle tests were compared with the results obtained by load-controlled high-cycle tests. The extrapolated S–N curves based on the cyclic stress-strain curve predicted the fatigue strength in the high-cycle range to be stronger for cyclic-hardening steels, but weaker for cyclic-softening steels. The predicted S–N curves for stable steels coincided with the high cycle test data. The fatigue limit had a proportional relationship with cyclic yield stress, slightly depending on the cyclic deformation behaviour. On the other hand, the cyclic yield stress was found to exhibit a very good linear correlation with the monotonie tensile strength, independent of cyclic deformation behaviour. This explains the empirical law that the fatigue limit is approximately proportional to the tensile strength.

Notes: This paper puts forward a new method for analysing the behaviour of very short fatigue cracks. A probability function is introduced into the definition of the growth threshold, which rationalises the scatter in experimental data produced using an aluminium bronze alloy. This probability function can be visualised in terms of the microstructure of the material.It is shown that, in this material as in mild steels, fatigue crack initiation is not the critical stage. Initiation occurs relatively easily, but the cracks so formed may grow to only a few grain diameters in length before being arrested; thus it is the behaviour of cracks of this length which is critical in determining the fatigue strength of the material.These observations, when combined with the probability functions, allow estimation of the probability of failure of a component or structure in service with greater confidence than the methods used at present.

Notes: The high temperature low cycle fatigue properties of modified 9Cr-1Mo ferritic steel in a hot forged and a hot rolled condition have been evaluated. The hot forged material exhibited inferior fatigue properties as compared to the finer grained hot rolled material. Analysis of the data indicates that a larger grain size adversely affects the initiation stage but has little effect on the propagation stage. A steeper slope on the Coffin-Manson plot results when the number of cycles to initiation is reduced.

Notes: The overload induced fatigue crack propagation behavior of several aluminum and steel alloys was examined as a function of the baseline stress intensity factor range (δKb). In order to gain a clearer understanding of the parameters which influence the cyclic delay phenomenon, under both plane strain and plane stress conditions, tests were conducted at δKb values ranging from the near threshold regime to high δK levels approaching fast fracture. Large amounts of overload induced cyclic delay (˜100,000 cycles) were observed at both high and low δK levels (provided the plastic zone size/thickness ratio and plastic zone size/grain size ratio approached unity, respectively) with significantly less delay occurring at intermediate δK values. All alloys examined exhibited this type of delay behavior which can be described by a “U-shaped” plot. The delay phenomenon at high δKb levels under plane stress conditions was attributed to increased crack closure associated with large tensile displacements in the wake of the advancing crack. At low δKb levels increasing cyclic delay was attributed to an increased effective overload ratio as δK approached δKth.

Notes: Changes in the crack growth behavior of 7075-T651 aluminum specimens which had been exposed to temperatures between 121 and 179°C were evaluated. Specimens were fatigue tested at room temperature under flight-by-flight loading conditions. Results from these tests were then compared with data from the as-received material. Exposure to 141°C or more produced a definite increase in specimen life, apparently due to a decrease in crack growth rate. In contrast, preliminary experiments using constant amplitude loading did not show noticeable effects of these thermal treatments on fatigue lifetime. The thermal treatments were found to cause overaging of the metastable precipitate microstructure, which was thought to be responsible for the changes in fatigue response.

Notes: The initiation stage and short crack behaviour in torsional fatigue of a 0.4% C steel was investigated by a replication technique.The fatigue cracks initiated and propagated in the ferrite phase which is located at the prior austenite grain boundaries in the form of long allotriomorphs. At this stage of crack development it is proposed that crack growth rate depends on the extent and intensity of plasticity at the tip of the crack. Crack growth per cycle is correspondingly proportional to the strength of the slip band. The ferrite-pearlite boundaries are strong barriers to crack propagation, which is manifested by a deceleration of growth and possible arrest. On raising the stress level the previously non-propagating cracks may continue to grow by branching or joining with other cracks in the ferrite phase. This process is repeated until the stress fields of one or more dominant cracks attain a critical value to sustain continued growth that leads to failure.

Notes: CT type specimens containing two layers of 309L and 308L cladding stainless steels welded to A508 carbon steel and 316 stainless steel were specially devised to test the influence of R ratio and environment on the crack propagation rate behaviour of cladding materials at 300°C. Large effects are shown. The crack growth rate under vacuum can be smaller by more than one order of magnitude as compared to air. Large differences are also shown on the fracture surfaces, where it is observed than vacuum promotes the formation of large crystallographic {111} facets. The effect of environment is briefly discussed on the basis of existing gas adsorption models.

Notes: The monitoring of fatigue crack initiation mechanisms and short fatigue crack growth usually involves interrupting otherwise continuous load cycling for examination of the test-piece. This permits attainment of high resolution over a large area of surface where the fatigue cracks could initiate. However, breaks in the fatigue cycling can affect the fatigue endurance of the test-piece and in the present study of Ti-6A1-4V this has been shown to depend on the duration of the rests, the number of cycles between rests, the type of examination employed and the hydrogen content of the material. The results of the work are discussed in terms of the possible influence of the monitoring technique on the fatigue process.

Notes: Abstract— Fatigue crack tip opening displacements and strains in the material very close to the crack tip have been determined from measured displacements for cracks grown in both vacuum and humid air environments. The environment alters both the relations between crack tip opening displacement and crack tip strain, and the effect of cyclic stress intensity on these factors. Results of dynamic observation of intermittent crack growth are correlated with fractographic evidence. The relationships between crack tip parameters are used in a previously developed mathematical model. The effect of wet air on fatigue crack growth is found to be a reduction in crack tip plasticity.

Notes: Abstract— Unexpected arrest of long fatigue cracks was observed in mild steel single edge notch three-point bend specimens tested under constant amplitude loading. Arrest was associated with a low, but still positive, slope of the crack length against stress intensity factor curve, and can be explained using the R -curve concept for fatigue-crack growth. At a stress ratio of 0.1, the fatigue threshold was 6.6, 7.3 or 8.0 MN/m3/2 depending on the definition of threshold used. This has obvious implications for both the development of a standard test method for the fatigue threshold and the application of data to practical problems.

Notes: Fatigue crack growth rates have been determined on standard specimens containing long cracks (∼5–10mm) and on specimens containing two-dimensional short cracks (∼0.10–0.50mm). Large differences have been observed indicating that at a given stress intensity factor short cracks propagate much faster than long cracks. Mouth opening displacement measurements for both specimen geometries have shown that the crack closure effect is largely responsible for the observed effect. These results are used to rationalize the behaviour of short cracks initiated from natural sites which were either graphite nodules or microshrinkage pores. The three-dimensional aspect of these natural small cracks is analysed and discussed in detail.

Notes: For ultra-high strength steels and aluminium alloys, a fatigue crack could initiate from a notch tip under cyclic compressive load. The threshold value for fatigue crack initiation under compressive load can be as great as four times that under tensile load. The crack grew at a decreasing rate until eventually it stopped growing altogether under cyclic compressive load with a maximum length of 0.2-0.5 mm. If the minimum compressive load was near zero, i.e. compression to zero load cycling, the threshold value was near that under tensile loading and the compressive fatigue crack could continue to grow; however, the crack growth rate under compression to zero load fatigue was 10–100 times less than that under the tensile fatigue loading.

Notes: An investigation has been carried out on austenitic stainless steel 0Kh16N15M3B under normal conditions and also to neutron irradiation of 6.8 × 1016nm-2s-1 (E 〉 0.1 MeV) intensity. Thin-walled torsion cylindrical specimens were tested in strain-controlled fully reversed loading mode at 923 K. Various ranges of strain, pre-loading fluences and half-cycle hold times (1, 5 and 30min) were applied. Neutron irradiation was found to result in hardening of the steel, stimulating cyclic stress relaxation and a reduction in cyclic life. When acting together, neutron irradiation and static loads cause a more significant reduction in the number of cycles to failure than if summed up as independent factors. Application of a kinetic failure criterion based on a damage parameter enables an estimation to be made of the limiting state of the steel under high-temperature cyclic loading with hold periods.

Notes: Fatigue crack propagation tests have been made on 150 mm wide panels of 1.6 mm thick 7475-T761 clad aluminium alloy sheet with and without adhesively bonded patches of pre-formed carbon fibre reinforced plastic (CFRP). The test frequency was 10 Hz, the minimum stress: maximum stress ratio, R, was 0.1 and the peak applied fatigue stress was 60 MPa.The tests were undertaken to assess the possibility of preventing the growth of fatigue cracks, or reducing their rate of growth, by the application of CFRP patches to one face only of pre-cracked aluminium alloy sheet. The variables examined included the type of patch and adhesive; the size, shape and thickness of the patch; and the effect of chamfering the edges of the patch and the removal of the cladding prior to patching.Results indicated that correctly designed and bonded CFRP patches substantially decreased the subsequent crack growth rate. The size and thickness of the patch had significant effects upon the reduction of fatigue crack growth rate whereas the shape of the patch, chamfering and the removal of the cladding prior to patching had little influence.

Notes: An experimental investigation is reported of the possible effects of adiabatic shear bands, caused by projectile impact, on the fatigue strength of a titanium alloy (Ti-6%A1-4%V). No significant reduction in fatigue strength due to the presence of adiabatic shear bands was found, nor did the fatigue cracks initiate from the bands. The results are relevant to the problem of foreign object impact damage to compressor blades in gas turbine aeroengines.

Notes: In order to study creep-fatigue interactions under multiaxial stress states, both push-pull and reversed torsion low-cycle fatigue tests were carried out using an austenitic stainless steel, SUS 304, at 923 K in air. From the tests, it is concluded that the hold-times introduced at the peak strain reduce low-cycle fatigue lives in the push-pull mode, but in the torsional mode they were not so harmful. This difference in the hold-time effect is discussed from considerations of crack formation and propagation and the stress amplitude applied to the specimen.Both maximum principal strain range, Δε1, and the von Mises' equivalent strain range, Δεeq, provide a nearly adequate comparison base for the assessment of biaxial low-cycle fatigue lives in tests without strain hold-time but are inadequate for hold-time tests. An equivalent stress range, Δσ*, which includes the effect of the stress parallel to the fatigue crack and which was previously proposed by the authors for no hold-time tests, is applied to the hold-time tests in the biaxial stress state. It is found that Δσ* is a good parameter for the correlation of biaxial low-cycle fatigue data in both no hold-time and hold-time tests.

Notes: Small fatigue crack growth behaviour in a low alloy steel was investigated under two stress step multiple loading in which the secondary stress was below the fatigue limit. Crack growth rates were presented in terms of a stress intensity factor and compared with data obtained under constant amplitude loading. In the higher ΔK region, crack growth rates increased monotonously with increasing ΔK even though the stress level was below the fatigue limit, and tended to be lower than those for constant amplitude loading. In the lower ΔK region, cracks showed a complicated behaviour, that is, an initial high growth rate was observed followed by an arrest or a drop to a minimum value and then a gradual increase. The average crack growth rates per cycle at both primary and secondary stresses in each block were approximately consistent with the da/dn-ΔK relation for constant amplitude loading. A suggestion for the prediction of crack growth life is given.

Notes: Abstract— Tests for fatigue crack initiation were carried out on two different aluminium alloys. Results and analysis of initiation under constant amplitude loading are presented; elastic and elastoplastic analyses are applied. Initiation under programmed block loading is investigated and damage accumulation is discussed. Tests were performed on two notch root radii:5 and 0.5mm. The electric potential method was used to detect fatigue crack initiation. Three point bending tests on smooth specimens were carried out to follow the evolution of damage during the crack initiation phase.

Notes: Abstract— High strain fatigue properties of AISI 316, AISI 316N and Sandvik 253MA have been investigated at 600°C. The two latter alloys, which contain significant amounts of N, exhibit a higher resistance to fatigue than 316. This effect is accompanied by a planar dislocation slip mode in 316N and 253MA as opposed to a wavy slip mode in 316. The results provide strong evidence that N improves fatigue strength in austenitic stainless steels, by inhibiting cross-slip of screw dislocations.

Notes: Abstract— A phenomenological model of cumulative creep damage combining simulated grain boundary cavitation with internal redistribution of stress is developed and matched to constant load tensile creep data for an AISI 316 stainless steel tested at 550°C. The model is shown to predict the creep life of the material when it is subjected to single step changes in load provided the strain rates subsequent to the change are imposed in the model. It is inferred that this supports current suggestions that cavitation failure may be strain controlled.

Notes: Recently a controversy has developed over whether crystallographic crack growth near threshold in Ni-base superalloys occurs along {111} slip planes or {100} planes at room temperature. In this work on Nimonic API crack propagation is shown to occur on both {100} and {111} planes. The most common facet plane is {111} and this is the only orientation observed at the lowest stress intensities, but at higher stress intensities occasional {100} facets are also produced. This behaviour is compared with similar results in aluminium alloys.

Notes: A crack nucleation model is constructed based on the random slip of the dislocations within the fatal persistent slip band (PSB). From such a model, we derive the notch-peak probability function. The crack is considered to nucleate when a critical notch depth is attained. This depth is constant for applied strains corresponding to the plateau of the cyclic stress- strain curve because the stress is constant. The PSBs behave in such a way that the more concentrated their localized slip the fewer the cycles required to attain the critical notch depth. Using the experimental relation between the applied strain and the magnitude of the slip offsets (reported in Part I) in conjunction with observations of the notch distributions along the gage surface, we have obtained the relation between the applied strain and the cycles for crack nucleation. A log-log plot of this relation shows a slightly different slope from that of the Coffin-Manson life data, previously reported. However, by allowing for cycles spent in Stage II propagation, good agreement is obtained.

Notes: The statistical nature of fatigue crack growth rate has been examined for G41400 steel exposed to different heat treatments and for different loading and environmental conditions. Techniques for determining the intrinsic value of the crack propagation rate are discussed. The intrinsic scatter due to the material has been evaluated and considered in terms of its effect on fatigue life prediction, both normal and log normal distributions of fatigue crack growth data being considered. It appears that the intrinsic material scatter in crack propagation rate is closely related to the scatter in total life.

Notes: A d.c. potential drop formula has been established for monitoring crack growth in circular notched members subjected to torsion. Changes in crack depth of 20 μm can be detected in cracks growing up to 2 mm in depth.