Open Collections

Dipole radiation intensity generated by cylindrical struts

University of British Columbia

The purpose of this work is to further establish the viability of the Causality Correlation Technique as a diagnostic tool for treatment of noise problems. Acoustical dipole radiation can be generated by obstructing a subsonic flow with a rigid strut, if the strut exerts fluctuating forces on the fluid. Such forces would be forces of reaction arising from unsteadiness in the local flow and would form a distribution of acoustical dipole sources over the surface of the strut. For the experiments reported herein, the subsonic flow issues from a circular nozzle which is 3.8 x 10⁻² m in diameters. The 'quiet' air jet operates at an exit Mach number of .217. The exit velocity is 72 m/s and is approximately uniform over the exit plane. In the first experiment the cylinder model is stationed at the potential core of the jet; the Reynolds number is 6.3x10⁴ (based on cylinder diameter and exit velocity). Later the cylinder model is stationed at the turbulent transition region of the jet. The 'Dipole Radiation Intensity (DRI)' (see Sec.3.1.2) is a uniquely defined and measurable quantity that is intimately related to the classical dipole strength. The 'spatial distribution' of the DRI can be constructed on a surface using the Causality Correlation Technique (see Siddon¹). The 'DRI distribution' is constructed on the surface of the rigid cylindrical strut. A diagnosis is made of the aerodynamic noise generation mechanism using the said distribution. Circumferential profiles of DRI reveal that: for both laminar and turbulent incident flow, a fluctuating lift force on the cylinder model produces dipole radiation (dumbell shaped radiation directivity) with its major lobes in the cross stream direction; for turbulent incident flow, an additional fluctuating drag force produces additional dipole radiation with its major lobes parallel to the flow. The SPL that is generated by unit length of the cylinder model becomes 10 db more intense when the model is in the turbulent mixing region of the jet compared to that generated when the model is in the potential core. The far field SPL originating from the surface exclusively is predicted from the integrated DEI distribution. For laminar incident flow the predicted SPL is (69.3 ±2.3) db. This may be compared with an overall SPL of (70.1 ±.5) db which was directly measured. However the overall SPL contains jet and wake noise contamination. The level of this contamination is estimated and is then deducted from the overall SPL. The resulting corrected SPL is (68.5) db. For turbulent incident flow the predicted SPL is (71.33 ±2) db compared with an overall SPL of (74.2 ±-5) db from which the corrected SPL was estimated to be (73.6) db. The error in the predicted SPLs arises from a systematic error which is incorporated in the DRI distributions. The main factors contributing to the error in the DRI are: an uncertainty in the sound travel time and limited resolving power of the correlator. Nevertheless the SPLs predicted by the Causality Correlation Technique are in close agreement with the corrected SPLs. Results of the present work support the claim that useful information on noise generation may be obtained quickly by performing measurements at critical points only, on a surface.

Text

2010-03-05

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.

http://hdl.handle.net/2429/21537

Mechanical Engineering

University of British Columbia

Graduate