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).
Valve spring design, harmonics, surging, and performance comparisons.
| Identifier | ExFiles\Box 56\2\ Scan077 | |
| Date | 2nd May 1929 guessed | |
| SPRING. Apparently Donkin feels that any practical cam, however designed, is bound to be represented by a harmonic series in which there will be a number of important components in the higher orders of harmonics. He therefore finds that in practice if he makes the spring stiffer than theoretically desired, and also allows one or two coils to close each end when full open to form a checking effect on the surges, he will stop the surging. This is not mere theory as over 20 makes of cars have been fitted with modified valve springs by the Cleveland Wire Spring Co. with good results, and some of them, for instance Chrysler, have certainly given us reason to worry in the silence they have achieved. Our spring frequencies work out at 12,400 cycles per minute for E-70442 valve spring, and 14,450 cycles per minute for E-70443 the tappet spring, both of which are in the lower half of the range of 10,000-30,000 cycles per minute given by Donkin. Donkin strikes me as very practical because it seems as though a revised cam form to remove or reduce the 11th harmonic say, would be so close to the original cam, that both of them would be within the limits of manufacturing error. I will send you samples of any revised springs which may seem to offer improvement. At the same time our present valve lifts, designed on a different theory of constant accelerations, may prove to have some startlingly large harmonics very different from Fig. 14 of Jehle's paper. Spring Position. All the above only deals with flutter and with max. engine speed, of course. The noise of closing seems to be a different question. Here we have been until recently very insistent on getting parallel ends on the springs, until we found that your springs which are much quieter than ours, are not as square on the ends. Neither do your springs appear to have coils more evenly spaced than ours. Attempting an analysis, we find that almost any spring can be twisted on the engine into one quiet position, and would be all right if it stayed there, but unfortunately after a few hundred miles it starts to creep round and goes on creeping from them onwards. On one car we have locked the lower ends of the springs into a plate which will prevent them rotating, and we hope in this way to prevent the recurrence of noise after a few hundred miles. -continued- | ||