1. The general validity of the refraction correction described herein depends on the fact that for wavelengths less than the tunnel radius the physical problem is only weakly dependent an source type. The weak dependence is discussed more fully in reference 12. For long wavelengths, however, the source type can have a more significant effect,SC discussed by Mani (ref. 16) et al. That is, the shear layer will affect differently the sound produced by a monopole, dipole, quadrupole, etc. This is illustrated in figure 4 of Amiet's study (ref. 12) in which the far-field sound pressure level for a monopole is compared with that for a dipole with its axis normal to the tunnel centerline. The difference af approximately 314 dB between the two amplitude corrections at zero frequency can be even greater when comparing higher order singularities. Since the source type is in general unknown, derivation of a low frequency shear layer correction from the result for a specific source type, as is done in equation (5-3) of reference 13, may lead to erxors in practical applications. Shear Layer Correction for an Off-Axis
2. Correlation at M = 0.4, f = 10 kHz, E, = 37.5". Test Configuration 1
3. Figure 12 shows a comparison of theory and experiment for the case of an on-axis acoustic source located at X/Ro = 1.33. Shear layer separation distance and far field microphone position are specified in the figure. The excellent agreement, at the selected frequencies and Nach numbers is representativeof the goad agreement at all operating conditions tested in Configuration 1. The only exception occurred at f = 1 kHz (not shown) where the small measured phase difference introduced some scatter into the experimental data points.
4. o 0.1 100 - 0 0.3