High frequency fatigue

Abstract

The object of these researches was to determine the effect of the frequency of alternation of stress on the fatigue limits of various metals. The work was carried out in the Engineering Laboratories, Oxford. Tests were made on rolled, normalised and hardened steel; rolled aluminium; annealed copper and normalised armco iron. The ordinary frequency employed in fatigue tests is 50 periods per second. Jenkin in 1924 carried out tests up to 2000 periods per second and in the research described in this paper tests were made at frequencies up to 20,000 periods per second. In all the higher frequency tests the specimen counted of a bar supported at the nodes, and vibrating freely- Jenkin used an electro-magnetic method to produce the vibrations, but this will not work for very high frequencies and a new method had to be invented. In the experiments now described fluctuations of air pressure acting directly on the specimen were used to make them vibrate and a number of methods were tried before a successful apparatus was evolved. The apparatus consists essentially of the two blowers used to vibrate the specimen. Each blower consisted of a small adjustable resonating chamber, into which air was admitted by a throttle valve in the back, while the front was closed by one face of the specimen. The portion of the specimen was so arranged that as it vibrated to and fro it alternately released the air pressure or allowed it to mount up in the chamber. The strains were calculated on the assumption that the bar vibrated freely and the only measurement necessary was the amplitude of vibration at the centre of the bar. Lord Rayleigh has shown how the strains may be calculated for a long, thin vibrating bar. But using the method of vibrating by air, the bars had to have a moderate with and for the biggest speeds they had to be short, so that Lord Rayleigh's theory was no longer sufficiently accurate. The experiments would have been impossible but for the assistance of Prof. Love, who explained how the known theory could be applied to bars of moderate with, such as could be used in our apparatus.