The full aeroacoustic two-port measurement procedure in flow ducts involves the determination of a scattering matrix and a source term. In this paper, the improved error suppression of a recently developed over-determination technique for the source is investigated. By inducing disturbances in the reflection coefficients it is shown that by applying the over-determination the accuracy is improved. For errors introduced via the scattering matrix the results are identical regardless of method applied. From measurements it is shown that random errors in the microphone data are better suppressed using the over-determination method. While this is true for bias errors due to external sound sources as well, the accuracy is unsatisfying when the external source generated sound is more than about 3 dB louder than that of the two-port source.

In the study of in-duct aero-acoustic phenomena two-port analysis based upon in-duct measurements has become an important method for the plane wave region. However already at moderate Mach numbers (0.2 - 0.3) the errors in the results can be hard to suppress. Ways of dealing with this includes the use of over-determination methods and methods of obtaining more accurate wave numbers. In this paper different methods for the determination of the source data and the scattering matrix are evaluated. For the source data in the form of a cross spectrum matrix an over-determination method is introduced. In addition, a method of obtaining the mean Mach number from experimentally determined wave numbers is described.

In order to evaluate the methods measurements are conducted at a mean Mach number of 0.2 for two test cases – an empty duct and a mixer plate inside the duct. The main improvements in the scattering matrix results are achieved by discarding measurements from an abundant set, based upon the measured coherence between excitation and fluctuating pressures. For the source part, it is shown that the error in the magnitude of the source cross spectrum matrix can be significantly suppressed by having additional (> 2) reference microphones on each side of the two-port.

The mean Mach number obtained from an experimentally determined wave number is observed to yield more accurate scattering matrix results in both phase and magnitude, than that based upon flow velocity measurements at one point and an assumed flow profile.

The flow-acoustic interaction in a T-junction can result in both amplification and attenuation of incoming waves. The frequency ranges of amplification are mainly governed by the time scales of convection of hydrodynamic instabilities over the side branch opening. With bias flow, i.e. flow through the side branch the amplification can increase or completely vanish. In this paper this effect is studied for a T-junction of rectangular ducts, with a grazing Mach number of 0.1 and a varying bias inflow. When the bias inflow Mach number changes from 0 to 0.01 the amplification is increased severely. Further increasing the bias inflow has the effect of lower the amplification again towards zero at a bias Mach number of 0.05.

The potential reduction of the aeroacoustic noise generated by an in duct plate, byallowing it to be flexible, is studied experimentally. The test object is a triangular plateinserted at an angle in a circular flow duct. Results are given for the active and passiveacoustic properties. In addition the flow field and the vibrations of the plate arecharacterized. It is found that an appropriately yielding plate reduces the flow generatednoise while keeping the mean flow field unaffected.