Diagnosing Interior Noise due to Exterior Flows in STAR-CCM+ Phil Shorter, CD-adapco
Overview Problem of interest Analysis process Modeling direct field acoustic radiation from a panel Direct fields for individual modes Direct fields due to random vibration induced by flow Modeling the random reverberant response in a cavity Summary
Overview Problem of interest Analysis process Modeling direct field acoustic radiation from a panel Direct fields for individual modes Direct fields due to random vibration induced by flow Modeling the random reverberant response in a cavity Summary
Typical sources for interior noise Airborne : 300 Hz to 10 khz Structure-borne : 0 to 800 Hz Wind-noise : 50 to 10 khz Antennas/Racks Rear turbulence Mirror/Greenhouse Wipers HVAC Front End Underbody Many sources of interior noise are caused by exterior flow
Simplified example Cross-Section AA Exterior flow 40 m/s Panel A A Interior acoustic cavity Experimental setup M. Smith et al Validation tests for flow induced excitation and noise radiation from a car window, Proc. 33rd AIAA Aeroacoustics Conference. Sound package foam/fiber
Overview Problem of interest Analysis process Modeling direct field acoustic radiation from a panel Direct fields for individual modes Direct fields due to random vibration induced by flow Modeling the random reverberant response in a cavity Summary
Analysis process 1. Model flow 2. Model direct field 3. Model reverberant field Sff Unsteady transient analysis performed in STAR-CCM+ Pressure time history data exported across surface of panel Vibro-Acoustic model used to predict direct field radiation into acoustic space when random fluctuating pressure applied across exterior surface of panel Vibro-Acoustic model used to predict reverberant response in cavity for a given input power in the direct field
Step # 1 : modeling the flow 1. Model flow 2. Model direct field 3. Model reverberant field Sff Unsteady transient analysis performed in STAR-CCM+ For details see: M. Smith et al Validation tests for flow induced excitation and noise radiation from a car window, Proc. 33rd AIAA Aeroacoustics Conference 2012. P. Bremner, Vibroacoustic Source Mechanisms under Aeroacoustic Loads Proc. 33rd AIAA Aeroacoustics Conference 2012. Vibro-Acoustic model used to predict direct field radiation into acoustic space when random fluctuating pressure applied across exterior surface of panel Vibro-Acoustic model used to predict reverberant response in cavity for a given input power in the direct field
Overview Problem of interest Analysis process Modeling direct field acoustic radiation from a panel Direct fields for individual modes Direct fields due to random vibration induced by flow Modeling the random reverberant response in a cavity Summary
Step # 2 : modeling panel direct field 1. Model flow 2. Model direct field 3. Model reverberant field Sff Unsteady transient analysis performed in STAR-CCM+ Pressure time history data exported across surface of panel Vibro-Acoustic model used to predict direct field radiation into acoustic space when random fluctuating pressure applied across exterior surface of panel Vibro-Acoustic model used to predict reverberant response in cavity for a given input power in the direct field
Modeling a glass panel Choice of Vibro-Acoustic method (FE, BEM, SEA etc.) depends on wavelengths of interest and size of system Small glass panel (0.4 x 0.2 x 5e-3 m) has approx. 45 modes below 10 khz This example therefore uses a deterministic (analytical) representation of the panel and its modes Mode 1 : ~380 Hz Mode 27 : ~3.9 khz Mode 40: ~8.9 khz Frequency (Hz)
Modal direct fields at 1 khz Normalize mode shape to have maximum velocity of 0.1 mm/s, look at radiated sound at 1kHz when mode shape radiates in a baffle (in this example, acoustic radiation calculated using boundary integral) Abs(Re{P}) db re:2e-5 Mode 1 : ~380 Hz Mode 27 : ~3.9 khz Mode 40: ~8.9 khz Different modes have very different radiation efficiencies (below coincidence)
Modal direct fields Mode 1 : ~380 Hz 500 Hz 1 khz 5 khz Mode 27 : ~3.9 khz Mode 40: ~8.9 khz Directivity of radiated field from a given mode shape changes with frequency
Overview Problem of interest Analysis process Modeling direct field acoustic radiation from a panel Direct fields for individual modes Direct fields due to random vibration induced by flow Modeling the random reverberant response in a cavity Summary
Random modal forces due to flow (S ff ) Fluctuating surface pressure from STAR-CCM+ can be transformed to the frequency domain and averaged over overlapping segments to give modal cross-spectral force matrices : S ff (f) FSP(x,y,t) F (t) Re{S ff (1kHz)} (x,y) <S ff (f)>
Random modal responses In a random vibration analysis, the cross-spectral modal response (Sqq) is related to the cross-spectral modal forces (Sff) by Modal dynamic stiffness matrix
Direct field response (<S pp >) Total direct field pressure response within the cavity can be found from S qq and the modal direct fields calculated previously. S pp (db re:4e-10 Pa 2 /Hz) (40dB dynamic range) (60dB dynamic range) (80dB dynamic range) 500 Hz 1 khz 5 khz Direct field radiation from flow induced random vibration
Look at spatial variation along a line A2 A1 (40dB dynamic range) (60dB dynamic range) (80dB dynamic range) 500 Hz 1 khz 5 khz
Direct field pressure at 500 Hz Typical drivers ear location A1 A2 Evanescent ( sloshing ) in near-field, free-field propagation outside near field
Overview Problem of interest Analysis process Modeling direct field acoustic radiation from a panel Direct fields for individual modes Direct fields due to random vibration induced by flow Modeling the random reverberant response in a cavity Summary
Step # 3 : modeling reverberant field 1. Model flow 2. Model direct field 3. Model reverberant field Sff Unsteady transient analysis performed in STAR-CCM+ Pressure time history data exported across surface of panel Vibro-Acoustic model used to predict direct field radiation into acoustic space when random fluctuating pressure applied across exterior surface of panel Vibro-Acoustic model used to predict reverberant response in cavity for a given input power in the direct field
SEA model Reverberant response within the cavity involves short wavelength system level response (function of sound package distribution within the cavity). FE/BEM typically frequency limited for many applications and so Statistical Energy Analysis (SEA) commonly used. P in E (reverberant energy) P coupling P dissipated SEA is based on a set of power balance equations for the reverberant field (expressions developed for power input, power dissipation and power transmitted to adjacent subsystems)
Direct and reverberant fields at 500 Hz Total field Direct field Reverberant field Decreasing absorption Increasing absorption (Ensemble average) response in the reverberant field is spatially uniform. Levels depend on sound package within vehicle. For typical sound package configurations the direct and reverberant field contributions can often both be important at the drivers ear location.
Overview Problem of interest Analysis process Modeling direct field acoustic radiation from a panel Direct fields for individual modes Direct fields due to random vibration induced by flow Modeling the random reverberant response in a cavity Summary
Summary Prediction of interior noise due to exterior flows of significant interest in many applications ( Aero-Vibro-Acoustics ) Aero-Vibro-Acoustic analysis requires STAR-CCM+ model of exterior fluctuating surface pressures and vibro-acoustic models of interior noise Simple numerical example presented in this paper (glass panel in wall of wind tunnel radiating into an acoustic cavity) Vibro-Acoustic analysis performed: Hybrid (deterministic+sea) modeling approach used Direct fields calculated for individual modes Direct fields calculated for flow induced random vibration Reverberant field (and total cavity response) calculated using SEA Example highlights that both direct and reverberant field contributions may be important and may therefore need to be included in vibroacoustic analysis of interior windnoise