Duct Mach Number Research Articles

Assessment of the air flowrate measurement in altitude engine tests by the national measurement standards system

The air flowrate in the engine inlet is an important parameter used to calculate common performance metrics such as the thrust and specific fuel consumption of a gas turbine engine. In the Korea Aerospace Research Institute (KARI) altitude engine test facility, air flowrates are calculated by measuring the static pressure, total pressure and temperature in an engine inlet duct. In the present study, in order to verify the air flowrate measurements at the engine inlet duct of the KARI altitude engine test facility, inlet flow measurement devices which are a total pressure rake, total temperature rake, and boundary layer rake were tested in the national measurement standards system for high pressure gas flow of the Korea Research Institute of Standards and Science (KRISS). Sonic nozzles, calibrated for flows rate up to 10000 m3/h and pressure range up to 50 bar (with an uncertainty = 0.18 %), were used as reference meters. We compared the air flowrates obtained by the area-weighted average duct Mach number in the inlet flow measurement devices of the KARI engine inlet duct to the reference flowrates of sonic nozzles at KRISS up to a Mach number of 0.15.

Flow Modeling of Pseudoshocks in Backpressured Ducts

Scramjet operation in the lower hypersonic regime between Mach 4 and 7 is characterized by what is called dual-mode combustion. In this situation, the backpressure generated by heat release in the combustor separates the wall boundary layer upstream of fuel injection, forming a highly complex flow known as a pseudoshock. The pseudoshock phenomenon has been studied extensively in the experimental literature, and its structure has been found to be affected by the duct Mach number, the extent of the wall boundary layer, and the streamwise area distribution of the duct. This paper presents a quasi-one-dimensional analysis technique for calculating the flow through a pseudoshock. It is quite general, and it can be used for ducts with any supersonic inflow, cross-sectional shape, and streamwise area distribution. Comparison of the results of this analysis with experimental data indicates that it provides a good estimate of both the interaction length and the pressure distribution for -type pseudoshocks that occur in ducts with Mach numbers above two.

CFD Validation of Incompressible Cross-Flow Discharge Coefficients

Validation against air systems problems is required to enable Computational Fluid Dynamics (CFD) codes to be confidently used in the design of turbine cooling air systems. CFD calculations of orifice cross-flow discharge coefficients (Cd) have been compared with measurements by Rohde et al [1]. Simulations have been carried out for cases with a low main duct Mach number (Md ~ 0.25) using incompressible flow modelling. Comparisons have been made of cross-flow discharge coefficients for a range of pressure-head ratios and Mach numbers. Results at a main duct Mach number of 0.07 were obtained using the standard k-ε.: turbulence model which gave agreement to better than 5% for absolute values of pressure-head ratios and discharge coefficients. The trends in the data for pressure-head ratios and Mach numbers were also reproduced. At a higher main duct Mach number of 0.25, the Mach number in the vicinity of the orifice reached 0.8. As expected this rendered incompressible flow modelling unsuitable, resulting in inaccurate determinations of orifice pressure-drops. Work is already in progress to simulate high Mach number cases using a more suitable compressible flow model. The results obtained so far give confidence that CFD will become a valuable tool for evaluating air system losses in novel configurations.

Computation of Noise Radiation from Fan Inlet and Aft Ducts

This paper examines turbofan noise radiation characteristics of both the inlet and the aft exhaust ducts using solutions obtained from a finite element code. The far-field principal lobe peak angle variations with mode cutoff ratio, duct Mach number, and external flow Mach number are examined. The predicted far-field principal lobe peak angle variations compare very well with the available analytical results. It is found that the principal lobe peak angle of the inlet radiation is insensitive to the duct and external flow Mach numbers. In the case of the aft duct, an increase in duct Mach number moves the principal lobe away from the aft duct axis, whereas an increase in external flow Mach number moves the principal lobe toward the aft axis. The solutions (for both inlet and exhaust duct radiation) also show that the far-field radiation principal lobe peak angle is only a function of cutoff ratio and nearly independent of the mode number. The ability of the code to handle ducts with acoustic treatments is also demonstrated.

Evaluation of Optimized Multisectioned Acoustic Liners

A critical examination is presented of the use of optimized axially segmented acoustic liners to increase the attenuation of a liner. New calculations show that segmenting is most efficient at high frequencies with relatively long duct lengths where the attenuation is low for both uniform and segmented liners. Statistical considerations indicate little advantage in using optimized liners with more than two segments, while the bandwidth of an optimized two-segment liner is shown to be nearly equal to that of a uniform liner. Multielement liner calculations show a large degradation in performance, due to changes in assumed input model structure. Finally, in order to substantiate previous and future analytical results, in-house (finite difference) and contractor (mode matching) programs are used to generate theoretical attenuations for a number of liner configurations for liners in a rectangular duct with no mean flow. Overall, the use of optimized multisectioned liners (sometimes called phased liners) fails to offer sufficient advantage over a uniform liner to warrant their use, except in lowfrequency, single-mode application. AdB F f* H* / M m p pA x x* y y* 6 Nomenclature speed of sound sound attenuation function frequency channel height length of duct Mach number transverse mode number dimensionless Fourier coefficient of pressure, p(x,y),p*/pA amplitude of pressure fluctuation or p JcJ 2 dimensionless axial coordinate, x* /H* axial coordiante dimensionless transverse coordinate, y*/H* transverse coordinate fractional impedance variation specific acoustic impedance dimensionless frequency/*//*/cJ or co*//*/27rcJ

Noise Generated by Boundary-Layer Interaction with Perforated Acoustic Liners

A problem that occurs in the application of perforated plate acoustic duct liners is the noise generated by the. turbulent boundary-layer flow over the holes in the liner surface. This flow not only generates noise but also thickens the boundary layer. To observe the noise generation, a series of tests has been run at the McDonnell Douglas Aerophysics Laboratory. These tests demonstrated that the sound pressure level generated at a liner surface can be as large as 158 dB for a duct Mach number of 0.4. The liner self-noise as measured in one liner panel was found to be affected by changes in the impedance of other liner panels. The tests showed that liner self-noise can be an important consideration in liner design. ELF-NOISE caused by flow over solid and porous surfaces, and in particular noise generated by airflows over perforated plate acoustic liners has been known to exist for some time.1'2'3 Such flows not only generate noise but also increase boundary-layer thicknesses and reduce thrust from aircraft engine ducts. It is possible that self-noise in an engine duct with a large amount of acoustic lining could be high enough so that an increase in lining area would actually increase the duct noise level. Recently, tests were conducted at the Douglas Aircraft Company to measure the impedance of acoustic liners using the two-microphone method. A specially designed siren was used to generate the high amplitude sound waves needed for the tests. It was a surprise to find that the self-noise amplitude could be of the same order as the high amplitude sound introduced into the duct from the siren. The self-noise waveform was periodic, with a frequency much different from the liner resonant frequency. Because selfnoise appears to be a basic problem connected with the use of perforated plate liners, a separate study of liner self-noise was carried out. The phenomenon of self-noise reminds the authors of two related problems in unsteady aeroacoustics, that of the edge tone,4'5 and that of vortex shedding behind a twodimensional body.6 In both problems an unsteady and periodic motion is generated by airflow over a rigid surface. Furthermore, the acoustic wavelength is much larger than a characteristic dimension of the surface so that near the surface the flowfield is primarily a hydrodynamic or a pseudosound7 field. That is, pressure fluctuations in the near field are of the order pu2 rather than pew, where p is the fluid density, c is the speed of sound, and u is the magnitude of the fluid velocity at a typical location in the flowfield. In both the edge tone and the self-noise problems a of air impinges on a rigid surface; in the vortex shedding problem the jet is really the entire freestream, which comes to rest on the forward part of the bluff body. In each case hydrodynamic instability is responsible for the unsteady motion, and a portion of the stream energy is radiated as acoustic energy. This portion is quite small if the freestream Mach number Uao/c is small compared to one, but it grows rapidly with Mach number.

Measurements of lining impedance in a duct with flow

The two-microphone technique has been used to obtain data from which values of acoustic impedance have been calculated for perforated-sheet linings. The magnitude and argument of the impedance values have been plotted as a function of frequency, wave amplitude, and duct Mach number for five different lining geometries. These results are compared to theoretical values of lining impedance obtained from a set of equations based on past work [E.J. Rice and C.E. Feiler, NASA TN D-6178 (1971); E.J. Rice, AIAA Paper 73–995 (1973); W.A. Bell, B.R. Daniel, and B.T. Zinn, AIAA Paper 74–61 (1974); A.S. Hersh and T. Rogers, AIAA Paper 75–495 (1975)]. Although the agreement between experiment and theory is good in general, it is not good in detail because the complicated flow field near the orifices is not fully understood nor completely modeled by the theory. Suggestions for possible improvements to the impedance model are discussed.