Shock and Vibration Bulletin. Bulletin 35 Part 6 of 7.

摘要

Many items of equipment in a variety of vehicles are subjected to both shock and short (nonstationary) bursts of random excitation. It is common practice to analyze these latter events as though they were stationary and measure their severity in terms of spectral density of acceleration. The errors involved and difficulties of doing this are reviewed. Actually, short bursts of random vibration are similar to random shocks, particularly when the failure, if any, is due to the single highest peak (SHP) of the response, rather than to fatigue. The importance of the SHP as a measure of damage is also discussed. When it is the correct damage criterion, we should attempt to control it and not spectral density. Since the SHP is a random variable, we cannot control it using random excitation, but we can with shock tests. For systems with a single degree of freedom, or those whose sensitivity to damage is known to be confined to a narrow frequency band, the solution is simple, once the statistical distribution of the SHP is known. Available experimental and theoretical data are reviewed. Using the prescribed percentile of the SHP distribution, the shock test as severe as the random excitation is derived. In certain cases, when both stationary random and shock tests are prescribed, the former can be eliminated by deriving a shock test whose highest response peaks at each frequency envelop those of both original tests. The Situation with multidegree of freedom systems is shown to be significantly more difficult and not amenable to one method of solution. A technique is introduced for extending the results to these situations. Preliminary results obtained on a analog computer for the two degree of freedom system are presented to illustrate certain trends.