Shigeru Kitagawa

Study on Fatigue of Metals by Transmission X-Ray Diffraction (I)

As one of the X-ray studies of fatigued metals, the author adopted the transmission X-ray diffraction method in this study. Specimens of pure aluminum poly-crystals which were completely annealed were used, and they were fatigued by cyclic tension and compression strainings with a small fatigue test machine which was designed by the author. The test pieces were of the size of about 0.5mm in thickness, 2-5mm in length and 6mm in width at the smallest section. For most of those fatigue tests, the strain amplitudes were considerably large, which could not be measured, then the fatigue lives were about 104 cycles. The surface of the fatigued specimen was observed at times with an optical microscope, which could be directed and adjusted in three dimensional bearings. It was noticeable that most of the fatigue cracks occurred at grain boundaries. In process of the fatigue test, the transmission X-ray diffractions of that specimen were exposed on a Laue camera. The X-ray source was of cobalt target, and the electron beam was 30kV and 8mA. A slit of 0.5mm in diameter was used for the X-ray beam. The asterisms of the diffraction patterns were measured statistically, and it was confirmed that the asterisms increased with increase in the number of strain cycles. It seemed that there were correlations between the asterisms and the fatigue damage, though no clear conclusion could be deduced. The fatigued specimens were further thinned by electrolytically polishing until about 0.1mm in thickness. In the transmission X-ray diffraction of the thinned specimen, it was clarified that the diffracted spots were isolated to fine spots which were elongated; and at the same time,

X-Ray Investigation of Fatigue Damage of Metals with Actual Service Loadings

It is well known that X-ray diffraction breadth varies with stress cycles. Let the variation of breadth be designated by y=b/B-1, where b is breadth after any number of stress cycles, and B is its initial value. In order to analyze the fatigue process, the equation for the breadth variation was introduced by the authors. The equation is constructed with two parts. One expresses a quickly increasing part of breadth in early fatigue cycles, and corresponds with workhardening process. If the variation is denoted y1, the equation is as follows where k1 and m1 are constants, n is the number of stress cycles, σ1 is designated as σ1=σ-0.7σw, σis the stress amplitude and σw is the endurance limit. The other one expresses the slowly increasing part of that over the whole fatigue life, and it corresponds with fatigue damage process. If the value is denoted y2, the equation is as follows where k, A, D and m are constants. S-N curve can be represented with the following equation under constant amplitude loadings.The constants A, D and m are the same as those mentioned above, and can be determined from the fatiguc tests. The variation of breadth y is represented with sum of y1 and y2.The tested material was 0.10%C steel which was annealed at 700°C for 2hr. For the X-ray apparatus CoKα line was used. The distance between the specimen and the film was 70mm.In the present experiment, when the variation of breadth increased until y=0.19, the material fractured in the case not only of constant amplitude test but also of sinusoidally varying amplitude test. The measurement values and calculated curves agreed well with each other.

Fatigue Tests of Metals under Repeated Bending Stresses with Varying Mean Bending Stresses

Many investigations of fatigue of metals under actual stresses have been carried out recently. Most of them have been done in the case where only the amplitudes of repeated stresses are varying with time. In the case of actual stress spectra of machine members, however, not only the amplitudes of repeated stresses but also the mean stresses fluctuate, and extremely complicated stress patterns are repeated.As one of the investigations of those phenomena, the fatigue tests of the low-carbon steels have been carried out by the authors in the case where the repeated bending stress amplitudes are constant and the mean bending stresses fluctuate cosinusoidaly with time.In order to analyze the varying stress spectra, and to discuss the strength and lives of metal members under those fluctuating stresses, it is necessary to measure the values of actual stresses induced on the test piece, For this purpose, the bending moments applied on the test pieces have been measured from the elastic strain of the driving shaft or of the jaw gripping of the test piece with the electrical resistance strain meter, the electromagnetic oscillograph, and the vacuum tube volt meter.Besides in order to examine the patterns of the repeated stresses, they have been monitored with the synchroscope, then it is ascertained that those patterns are exactly sinusoidal, and it is also ascertained with electromagnetic oscillograph that the mean stresses fluctuate surely cosinusoidaly with time.As the results of the tests, it has been made clear that, in the case of the repeated bending stresses with the varying mean bending stresses, the endurance limits and the fatigue lives of the repeated bending stresses are considerably reduced in comparison with those in the case of the repeated bending stresses with the constant mean bending stresses.