Measured Fatigue Life Research Articles

Static Fatigue Behavior of Polycrystalline Beryllia

The delayed failure/static fatigue behavior of a polycrystalline BeO ceramic was examined by measuring the time to failure as a function of the static stress level. Using this technique, curves were generated over a temperature range of 8° to 50°C. The test environment consisted of a fluorocarbon fluid containing small quantities of water. The measured fatigue life was a strong function of both stress and temperature. The data also provided evidence of a fatigue limit which was insensitive to the temperature. These results were consistent with data generated from direct V–KI measurements. The fatigue data were used to examine the applicability of a generalized slow crack growth model relating the subcritical crack velocity to the stress intensity factor. Initially, agreement between the experimental and predicted behavior could be obtained only by assuming that the activation volume for the stress‐assisted crack growth reaction was negligible. This result implies that the major stress dependency was associated with the preexponential term in the crack growth equation. However, an alternate and more reasonable description of the experimental data was obtained by assuming that the activation volume exhibited a linear dependence upon the mechanical stress.

FATIGUE OF POLYCRYSTALLINE COPPER AT CONSTANT AND VARIABLE PLASTIC STRAIN AMPLITUDES

Fatigue behavior of polycrystalline copper of two different grain sizes was studied using constant and randomly varying plastic strain amplitude tests. The different regimes in the ess-curves of large-grained samples (d˜ 2.0 mm) were found to be reflected also in the measured fatigue lives in both test modes. The variable amplitude fatigue lives of the small-grained samples (d˜ 0.03 mm) seemed to follow the Coffin–Manson linear law. The damage accumulation rate in variable amplitude tests was found to be lower than in the corresponding constant amplitude tests, e.g. when expressed as cumulative plastic strains to failure. The fatigue life predictions based on the Palmgren–Miner rule, rain-flow analysis and the constant amplitude life data measured in this work were also conservative.

The effects of large vibration amplitudes on the dynamic strain response of a clamped-clamped beam with consideration of fatigue life

The effects of large vibration amplitudes on the dynamic strain response, near to the fundamental resonance, of a clamped-clamped, thin beam is examined. Complementary theoretical and experimental studies showed that the harmonic distortion of the induced total strain, due to sinusoidal excitation at mid-span, is mainly due to the axial strain component. A limited set of fatigue experiments showed the considerable decrease in fatigue life, which occurs due to non-linear vibration, compared to that of a cantilevered beam of the same material. The statistical approach to the analysis of non-linear vibration induced by random loading is examined theoretically and experimentally, good correlation being achieved between predicted and measured fatigue lives.

An Investigation on Formulation of Cyclic Local Strain Range and Estimation of Fatigue Life in Notched Plate

Many studies have been made on application of results of fatigue tests on small smooth specimens to evaluation of the fatigue life in structural members based on the estimated strain behavior in stress-concentrated region. Most of the estimation formulae for strain behavior so far proposed have been derived from the analyses under monotonically increasing load. In order to evaluate the fatigue life reasonably, however, it is more desirable to know cyclic local strain range rather than monotonic strain behavior.In the present study, therefore, it was tried to establish a simple method to estimate the local strain range under cyclic loading. Namely, elastic-plastic FEM analyses under cyclic loading were conducted on notched plates, and an investigation was made into effects of applied stress condition, stress concentration factor and material properties on the local strain range. As a result, the formulae to estimate shakedown limit and cyclic local strain range were derived.A comparison was made between some results of fatigue tests on notched plates so far published and the fatigue life estimated using the formulae proposed in this study. Although the scatter band in the relation between estimated and measured fatigue lives is rather wide, it is concluded that the method proposed can be used to estimate well on the whole the fatigue lives in the wide range of 1025 × 106 cycles in notched steel plates of 4180 kgf/mm2 strength levels.

Predicting Catastrophic OD Initiated Fatigue Failure of Thick-Walled Cylinders Using Low Cycle Fatigue Data

ABSTRACTA procedure is presented to predict fatigue failure of thick-walled cylinders containing discontinuities at the OD. Both crack initiation life and crack growth are considered. The elastic-plastic strains at an OD discontinuity are estimated using an elastic analysis and stress concentration factors. The strain estimates are then used in conjunction with low cycle fatigue data to determine the initiation life. Crack growth life is estimated by integration of a power law relationship. The results obtained by using this analysis method compared to measured fatigue life data for several OD initiated failures in thick-walled cylinders agrees to within about 10 percent.

Measured Fatigue Lives for a Resonant Cantilever with Random and Pure Tone Excitations

A cantilever beam of 2024-T3 aluminum sheet (3.25 in.×1.00 in. × 0.063 in.) was shaken mechanically at its clamped end, the force being applied parallel to the thickness dimension. A 0.125-in.-diam hole near the clamp afforded points of strain concentration. Both pure tone and broadly filtered random forces were used to excite resonant response in the fundamental mode, at about 185 cps. The rms surface strain in a cross section containing the hole ε was monitored and held constant. Life N to first visible appearance of a fatigue crack was measured; lifetimes ranged from about 1×105 to 6×105 cycles. The pure-tone data agree very closely with previous measurements on a riveted cantilever excited by sound. In comparison with available fatigue data for reversed bending in a specimen with small concentration factor, these data yield a fatigue strain concentration factor of 1.6, in good agreement with theory. The measured ratio of rms pure-tone strain to rms random strain yielding the same life-time ranged from 1.5 to 1.6 (3.6 to 4.3 db). The ratio calculated from assumptions of (a) Rayleigh distribution of peak strains, (b) noninteracting accumulation of damage, and (c) Nε8.25 = const, is 1.4 (3.1 db). (Research sponsored by Aeronautical Systems Division, Air Force Systems Command, U. S. Air Force.)

XCVIII. The growth of fatigue cracks

Summary The qualitative features of the growth of a fatigue crack are discussed. A model is proposed whichh as these features and from which quantitative predictions can be made. These are in satisfactory agreement with experimental measurements. It is concluded that cracks may be formed at an early stage in fatigue and that the growth of a crack may occupy a large part of the measured fatigue life.