Measured data not available. If a measured data base is not available, then for Procedure I - Functional Shock, and Procedure V - Crash Hazard Shock Test, employ the applicable SRS spectrum from Figure 516.6-8 as the test spectrum for each axis, provided Te of the test shock time history falls between the values in the accompanying Table (516.6-I). This spectrum approximates that of the perfect terminalpeak sawtooth pulse. It is highly recommended that the test be performed with a waveform that is composed of either (1) a superposition of damped sinusoids with selected properties at a finite number of designated frequencies or (2) a superposition of amplitude modulated sine waves with selected properties at a finite number of designated frequencies, such that this waveform has an SRS that approximates the SRS on Figure 516.6-8 where the duration of this waveform is a maximum of Te provided in Table 516.6-I. In reality, any complex test transient is suitable if it equals or exceeds this spectrum requirement over the frequency range of 5 to 2000 Hz, and meets the duration requirement. Use of the classical terminal-peak sawtooth pulse and the classical trapezoidal pulse is the least permissible test alternative in the case of no data being available (see paragraph 2.3.2c). In cases in which there is a vibration requirement for the materiel in addition to a shock requirement it may be possible to perform the vibration test in lieu of the shock test in the tailoring procedure. An example of this form of tailoring is contained in Procedure I - Functional Shock. Figure 516.6-9 provides two ASD curves to be used for comparison with specified ASD test environments to determine if random vibration is of sufficient severity to be used in lieu of measured or specified shock levels. The SRS for stationary random environments developed from these ASD curves, envelopes the appropriate SRS spectra on Figure 516.6-8. For some empirical justification of this, see Annex C of this method