Showing papers on "Vibration fatigue published in 1979"

Acoustic harmonic generation due to fatigue damage in high‐strength aluminum

Abstract: It is shown that acoustic second harmonic generation is a useful tool for studying surface microcrack development during fatigue of a high‐strength aluminum alloy. A fundamental (5 MHz) surface acoustic wave (SAW) was transmitted across the gauge section of flexural fatigue specimens of Al 7075‐T6. The second harmonic amplitude was determined after several increments of fatigue, as a function of external load and the amplitude of the fundamental. It was found that the second harmonic signal is at a maximum close to zero external load and increases with progressing fatigue. Harmonic generation, attributable to microcracking at the surface, has been observed as early as 10–20% of the expended fatigue life. A simple analysis to obtain a coefficient of harmonic‐generation efficiency versus applied surface stress is described. This analysis considers the effect of changes in attenuation of the fundamental and harmonic waves associated with degree of surface microcrack opening as a function of surface stress. I...

Fatigue performance of a bituminous road mix under realistic test conditions

Abstract: A study whose purpose was the verification of Miner's rule for estimating the cumulative damage resulting from the phenomenon of fatigue is reported. A repeated-bending apparatus driven by a minicomputer, which was devised to generate and control stress or strain waves of variable amplitudes, is described. The fatigue behavior of a bituminous mix subjected to nine different loading patterns (simple, random, and block) was determined. The influence of rest periods of different lengths was studied for these cases. It is concluded that (a) to the extent that the spectrum of load amplitudes is known, a prediction method derived from Miner's law is applicable with an acceptable accuracy for random sequences that include no rest periods and (b) rest periods markedly increased fatigue life for the three loading patterns considered. These initial conclusions were used to derive a generalized form of Miner's law for loading conditions in which both stress amplitudes and the duration of rest periods are variable. This generalized law was verified by simulating actual conditions of traffic loading in fatigue tests. /Author/

Fatigue-Limit Effect on Variable-Amplitude Fatigue of Stiffeners

Abstract: The effect of the constant-amplitude fatigue limit on the variable-amplitude fatigue life was determined experimentally. Tensile specimens with transverse stiffeners welded either automatically or manually were tested in such a way that, from one set of tests to another, an increasing number of cycles fell below the fatigue limit. Data were developed from 500,000 cycles — 60,000,000 cycles of loading. The root-mean-cube stress range was found to be a valuable transfer function between constant and variable-amplitude fatigue life data when its value exceeded the fatigue limit. When it was lower, the modification suggested by Yamada and the first writer—based on the assumption that those cycles in the histogram that fall below the fatigue limit do not drive the crack—led to good predictions of variable amplitude fatigue life.

Network Alternatives to Reduce Turbine-Generator Shaft Stresses

Abstract: Network alternatives to minimize shaft mechanical fatigue due to switching three-phase faults in close proximity of a 1025 MVA turbine-generator connected to a strong 345 kV grid are investigated. The comparative analysis is based on estimated fatigue curves for the major shaft sections, relating maximum alternating torque to fatigue incurred per incident of torsional oscillation. Two half-sized step-up transformers connected to alternate buses, coupled by a bus tie breaker are shown to be effective in reducing such fatigue.

Fatigue Analysis of Offshore Structures

Abstract: As the offshore industry moves into deeper water, the dynamic behavior of structures becomes a very important parameter in the overall design procedures. In particular, the cyclic nature of wave loads has a significant effect on the fatigue life of the structure. This paper presents a procedure for fatigue analysis where the dynamic response of the structure is analyzed through a spectral approach. The sea waves which constitute the forcing function acting on the structure are represented as energy spectra; the response is obtained in spectral terms and is subsequently interpreted according to probabilistic concepts. As part of this study we have considered the characteristics of the distribution of the response. In the usual approach to fatigue analysis one assumes that the peaks of the response follow a Rayleigh distribution. This assumption is valid if the Cartwright-Longuet-Higgins' measure of bandwidth (epsilon) is equal to zero. However, in practical fatigue applications, this idealization is rarely, if ever, encountered. Thus, in order to better characterize narrow-banded phenomena, we have investigated descriptions which do not assume epsilon = 0. An investigation of caisson designs indicates that fatigue using the more accurate response distributions can differ appreciably from those obtained with the Rayleigh distribution.more » 10 references, 5 figures, 2 tables.« less

Test simulation of fighter aircraft maneuver load spectra

Abstract: This paper studies the use of power spectral density (PSD) and related random data analysis techniques in the test laboratory to characterize low frequency loading spectra output response, measured from a specimen, for the purpose of comparing test machine responses. In recent years digital techniques have been developed that simulate aircraft maneuver spectrum loading in the time domain. In this study one of these techniques is used to generate the input load spectrum signal to an electrohydraulic closed-loop test machine for fatigue testing of graphite/ epoxy to titanium step lap bonded joints at various frequencies and PSD shapes. The output signal was measured directly from the specimen by recording on tape the dynamic strain gage signal. This signal then was processed to determine the PSD response and root mean square (RMS) time history. These data were compared with the recorded strain amplitude time history and the input load amplitude, frequency, and PSD information. A qualitative correlation was found between variations in specimen fatigue life, output signal waveform and amplitude, and variations in PSD. PSD is a measure of loading energy in the time domain and can be influenced both by load amplitude and waveform variations. Thus, the way a fatigue test machine processes an input load and frequency signal can affect the measured fatigue life of certain materials.

A Low-Cycle Fatigue Test for Pipe-Type Power Transmission Cables

Abstract: An apparatus for performing low-cycle flexural fatigue tests on pipe-type power transmission cables with wrapped-tape dielectric is described.

Method of evaluating thermal fatigue resistance of heat-resisting alloys

Abstract: A method based on the strain-kinetic criterion of long-term low-cycle failure is proposed for evaluating the thermal fatigue strength of heat-resistant alloys. The criterion describes the conditions for the critical states of quasi-static and fatigue failure for constant-stress-amplitude, constant-strain-amplitude, and intermediate loadings. It is shown that linear treatment of strain-kinetic approach is suitable for fatigue evaluation, and that the processes of cyclic strain occurring under thermal fatigue loading are unsteady and characterized by accumulation of considerable unidirectional deformations. Base information may be obtained from long-term static tensile tests and low-cycle fatigue tests. Evaluation of thermal fatigue strength must take the component of quasi-static damage into account.

Experimental Random Fatigue in Elastic Range - First Order Models.

Abstract: : An experiment program, based on 8 probabilistic parameters and 3 experiment designs, was conducted for 24 specimens for the fatigue life of materials in elastic range under random vibrations. The magnitude levels of all probabilistic parameters for the response of every specimen were computed from its spectral moments and were coded. By regressing the log of fatigue life on coded probabilistic parameters the first order models to predict fatigue life were obtained. (Author)

Constructing the fatigue curves of metals on the basis of an energy criterion

Abstract: 1. Since the amount of energy irreversibly dissipated in a metal to fractute is not constant when there is a variable stress amplitude, the construction of energy criteria of fatigue should be based on experiment data. 2. In connection with the change in the fracture mechanism in the region of the transition from low-to high-cycle fatigue, there is also a change in the parameters of the relationship between static and cyclic energy characteristics. This change must be considered in developing methods of calculation. 3. The fatigue curves calculated from well-known energy criteria are for most materials displaced from the experimental curves. 4. Based on the above-discussed energy relations, a function was proposed to describe fatigue curves with sufficient accuracy relative to empirical results.

Effect of energy dissipation on structural fatigue in case of random vibrations

Abstract: We examine a case when the strength of a structure is determined by the stress s(t) at themost endangered place, on the assumption that s(t) is a narrow-band random process. This assumption is correct if one resonance form of vibrations is excited. The problem of predicting structural fatigue, on condition that s(t) is a narrow-band process, was examined by Miles [i]~ His results were refined and generalized by Bolotin [2].