The determination of fatigue limits under alternating stress conditions

In the year 1867 A. Wöhler, locomotive superintendent of a railway company in Berlin, exhibited at the Paris Exhibition the results of some experiments on the endurance of metals, and was thereupon engaged by the Prussian Government to carry out the more exhaustive enquiry into this subject with which his name is always associated. The results of his labours were published in 1871, and were highly appreciated, but few additional experiments were made until the subject was again taken up successively by Sir Benjamin Baker, Reynolds and Smith, Rogers, Stanton and bairstow, Eden, Rose and Cunningham, and Prof. Hopkinson. All these experiments are confined either to fatigue bending or to push and pull tests, using only steel or iron, whereas the present ones include a large number of torsion fatigue tests on various metals. Until comparatively recently there was no satisfactory standard of comparison for fatigue tests, the determination of the asymptote or limiting fatigue stress for an infinite number of revolutions from a few irregular test results leading to very uncertain conclusions, so much so that by some it was considered very doubtful whether there were any real fatigue limits, while others adopted as standards of comparison the fatigue stresses which would cause fractures at the millionth repetition. The first problem which had to be investigated was therefore to ascertain the relationship between the intensities of fatigue stresses and the numbers of repetitions of these stresses which would cause fracture; and, should this relationship be found to indicate the existence of a limiting stress for an infinite number of revolutions, or more briefly of a fatigue limit, then the next step would have to be its exact determination.