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).
Technical discussion on valve-spring stress, surge, and the factors leading to fatigue failure.
| Identifier | ExFiles\Box 56\2\ Scan063 | |
| Date | 1st January 1929 | |
| Contd. position for an equivalent static compression by an amount which may be as great as or greater than a half patch of the particular coil. At an instant one free wave later, the conditions are exactly reversed. Thus, while the stress range statically may amount to 15,900 lb. per sq.in., the stress range in vibratory operation may vary between zero and solid stress, or even greater than the latter, depending upon the amount of impact due to coil clash. To be a little more specific about this stress range, the stresses on the spring whose natural period was 10,750 free vibrations per minute ran as follows: Stress, Lb. per Sq. In. Spring compressed solid 69,000 With valve open 42,200 With valve closed 25,300 When this spring is vibrating severely in its own period, the coils which are closed and clashing will be stressed 69,000 lb.sq.in., or even more, depending upon the impact value of the coil clash. At the same instant, a coil at the opposite end of the spring will be opened up a certain amount, due to the wave motion. If the latter coil is opened up an amount which will make its pitch at that instant equal to its pitch when the spring is in a free or totally unloaded condition, then the stress range will be 69,000 lb. per sq.in. or more. This, however, is a very extreme case. It is especially important to remember that the rapidity with which this stress range is passed through is of an order much greater than camshaft speed. For example, in the case of the spring mentioned, which has a natural period of 10,750 free vibrations per min., the entire abnormally high stress-range will pass from the maximum to the minimum value in 0.00558 sec., or in the time for the wave to assume a position exactly out of phase with its former position. From the foregoing discussion it can be seen that the ef- -3- January, 1929 fect of surge upon a valve-spring is to bring about: (1) Existence of a maximum stress of a higher value than that calculated by the conventional formula. (2) Occurrence of a higher stress-range in the end coils than that calculated on a static basis. (3) A state whereby the stress-range is traversed with great rapidity. All investigations along the line of metal fatigue tend to show that fatigue life is a function of the maximum stress existing, the stress range through which the material is worked, and the total number of oscillations to which the material is subjected. Accordingly, the abnormal stress conditions and the abnormal rapidity with which the stress range is crossed, due to the surge, tend to accelerate greatly the fatigue of the spring. Four Contentions Advanced These notes suffice for us to advance four major contentions relative to the general subject of valve-spring surge. (1) Valve-spring surge is a product of a resonant condition between the rate of the forced vibrations, due to the camshaft, and the natural frequency of the spring. (2) Surge cannot be eliminated in the present type of spring, as the tendency to vibrate is an inherent characteristic of a spring. A solution is offered, however, by designing springs for high frequency so that they will not vibrate until they are subjected to rates of forced vibration in excess of speeds above the range of normal maximum camshaft speeds. (3) Surge is affected by the design of the valve-gear. A tendency to magnify spring surge may exist if one or more parts of the valve-gear are also in resonance with the spring. (4) Surge is a prolific source of valve-spring failure, due to the stress conditions and the rapidity of the stress cycle set up by the surge. | ||