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).
Finish, manufacturing process, and failure analysis of Phantom II springs.
| Identifier | ExFiles\Box 16\7\ Scan212 | |
| Date | 8th December 1929 | |
| Copy. To Wor.{Arthur Wormald - General Works Manager} from EY. 8th December, 1929. RL/M26.11.29. In regard to Phantom 2 springs, as explained at our conference in London, I do not consider we are doing anything in the way of improved finish, which is not essential. Whilst the greater length of the rear spring from our point of view should reduce cost, if considered on a weight basis alone, the fact that we have so far had to produce forged eyes at each end instead of one eye only, and that a rolled one, has materially increased the cost. We are rolling the bar flat as suggested, but this in itself would not relieve us from the necessity to grind from a 'bedding' point of view, as the bedding operation would, under such, circumstances, cause local puckering due to scale and skin effects. Firths confirmed that if we had flat rolled bar it would be necessary to grind for the above reasons as this was their experience. In addition, grinding is necessary to prevent torn places produced in rolling from becoming corrosion pits, which would in turn, produce fatigue failure. Plating is necessary to preserve the plate (after it has been cleared of pits and scales) against corrosion produced by wet weather conditions. There is ample evidence to show that spring failures are mainly due to the reduction of the resistance to fatigue, about localised corroded portions of the plate, and almost every fatigue crack can be clearly seen to have started at a small corrosion pit in the plate. Even in special springs which were polished all over, erratic life was displayed. I therefore went to examine the springs personally, and found they were all breaking across a spot where a paper label was pasted on. I removed the label to discover the plates were severely corroded beneath them and failures had occurred starting from small pits, sometimes having three distinct cracks, all running separately and developing independently. After I removed the labels, the same batch of springs went up from thirty hours to seventy and eighty hours before failure occurred, a result we have never achieved even with American Springs, whose best life was forty hours. In the last pamphlet published by the Department of Industrial Research, they state emphatically that unground plates are essentially erratic in spring practice, and the only remedy for erratic behaviour is to grind the plates all over. EY. | ||