From the Rolls-Royce experimental archive: a quarter of a million communications from Rolls-Royce, 1906 to 1960's. Documents from the Sir Henry Royce Memorial Foundation (SHRMF).
Procedure and results of shaft fatigue and stress testing.
| Identifier | ExFiles\Box 132\1\ scan0151 | |
| Date | 20th September 1939 guessed | |
| -3- in early tests, 3° in intermediate tests, settling to 2.4° in latest tests. The shaft is therefore swung through a constant amplitude close to the resonance peak, but on the low side of it. (I notice from test logs that in early tests it sometimes took 30 minutes to get the speed adjusted until amplitude remained steady. The corresponding torsional frequency is pretty close to 7250 a minute, or 435,000 an hour. Failure occurs in 2 1/2 to 40 hours. Tests were run in two shifts from 8 a.m. to 11 p.m. so that generally they could get a test done every day. The test rig and operator had to be separated by a fairly sound-proof partition because of the noise. The operator's only job is to keep the amplitude constant. To do this he has to reduce the speed very slightly from shortly after the start of the test. When failure approached he has to reduce the speed quite suddenly, and after he has done this for a few minutes, to fully develop the final failure cracks, he shuts down, and removes the crankshaft. The sudden increase in swing, which calls for rapid decrease of speed, is counted as the point of failure. In only one case out of nineteen shafts was the specimen actually broken through. In only one case out of nineteen shafts was the specimen actually broken through. The calculated stress corresponding to the various amplitudes used varies from +- 9500 lbs. per square inch to +- 16,000. And the life on sound shafts varies from about one million cycles on original shafts to twenty million on final alloy shafts with 6 o'clock oil holes omitted and the pins Tocco hardened. From the results some tentative stress life lines have been drawn on double logarithmic paper. A remarkable thing, from the photos. of the final magnaflux inspection, is the considerable number of similar cracks through the pins occurring on any one shaft, often crossing through the oil-holes at almost exactly 45°. In other shafts torsional cracks at 45° occur in the main bearings with shatter cracks branching from them in all directions. Woodruff keyways with the edges carefully filleted and polished showed less tendency to crack away from the corners, but were a weak spot throughout the tests. Oil had to be pumped through the exciter at all times, and passed through an oilcooler. | ||