Simulation of Couplers by Søren Jønsson, Bin Liu, Lars B. Nielsen, Andreas Schuhmacher Brüel & Kjær AES, Workshop 7, 2003 March 23rd Brüel & Kjær 2003, 1
Agenda Introduction to Couplers Traditional Simulation using LPM FEM/BEM Simulation Comparison of Results Conclusion Brüel & Kjær 2003, 2
Introduction to Couplers Critical properties of Acoustic Couplers Must allow for REPEATABLE measurements Must be possible to CALIBRATE Must be ACCURATE and STABLE Should be STANDARDIZED Design to improve voice quality & intelligibility Design to meet telecom & audio standards Comparison of transducers, headsets, handsets, earphones etc. Brüel & Kjær 2003, 3
Introduction to Couplers Acoustic Couplers overview IEC 318 (ITU-T P.57 Type 1) Ear Simulator IEC 711 (ITU-T P.57 Type 2) Ear Simulator ITU-T P.57 Type 3.1 Concha bottom Simulator ITU-T P.57 Type 3.2 Simplified Pinna Simulator Increasing realism & complexity ITU-T P.57 Type 3.3 Pinna Simulator Brüel & Kjær 2003, 4
30 F requency Sensitivi tyrespon seforty pe4185, re1khz 20 db 10 0-10 OpenEar Sensitivity ClosedE arfreque ResponFrequen se ncy Sensitivity Respon se (Hz) 10-20 0 20 50 1k 2k 5k 970 Ear Simulators and Standards Ear Anatomy and Definitions Volumes Concha 4.3 cm 3 Ear canal 1.0 cm 3 Behind drum 0.65 cm 3 Total ear 6.0 cm 3 Brüel & Kjær 2003, 5
30 F requency Sensitivi tyrespon seforty pe4185, re1khz 20 db 10 0-10 OpenEar Sensitivity ClosedE arfreque ResponFrequen se ncy Sensitivity Respon se (Hz) 10-20 0 20 50 1k 2k 5k 970 Ear Simulators and Standards ITU-T Type 2 - IEC 711 occluded ear simulator - Standardized - accepted as a reference worldwide - Bruel & Kjær Type 4157 (and used in 4158/59/95) - Intented for calibration of insert earphones, sealed and unsealed (hearing aids) in the frequency range from 100-7.000Hz - Simulates the ear canal from app. 10mm behind EEP and in Brüel & Kjær 2003, 6
30 F requency Sensitivi tyrespon seforty pe4185, re1khz 20 db 10 0-10 OpenEar Sensitivity ClosedE arfreque ResponFrequen se ncy Sensitivity Respon se 10-20 0 20 50 1k 2k 5k 970 Ear Simulators and Standards IEC 711 coupler: Requirements in according to standard: Acoustical transfer impedance specified Equivalent volume at 500Hz equal to 1.26cm3 With of principal volume 7.5mm Length of principal volume must produce ½-wavelength resonance at 14kHz, i.e. ~12.4mm (Hz) ~170u ~69u Brüel & Kjær 2003, 7
Ear Simulators and Standards IEC711 coupler Setup for measuring transfer impedance Brüel & Kjær 2003, 8
Ear Simulators and Standards IEC-711 coupler typical measured data with tolerance curves from IEC60711 40 35 30 25 db 20 15 Typical Lower Tolerance Upper Tolerance 10 5 0 100 1000 10000 100000-5 Frequency (Hz) Brüel & Kjær 2003, 9
Traditional Simulation Using LPM Ear-canal impedance: Tube w. ridgid walls analog to an LC transmissionline, i.e: C Z at c M a = V ρc = A = C at 2 ρc a, n = C Z 2 c a,4 + C a,6 + C a,8 Brüel & Kjær 2003, 10
Traditional Simulation Using LPM Ear drum impedance: Two slits terminated by volumes, i.e. RLC-helmholtz resonators tuned to match drum-impedance Narrow slits Z a = 12nl a b 3 + 6 jω 5 ρl ab Brüel & Kjær 2003, 11
Traditional Simulation Using LPM 200 Acoustic impedance of slit2 150 real Imag 101 200 DB( Ir(Ra7) / ( I(Ra7)* I(Ra7) ) ) DB( - Ii(Ra7) / ( I(Ra7)* I(Ra7) ) ) Acoust i c i mpedance of sl i t 1 150 real imag SEL>> 100 100Hz 200Hz 300Hz 400Hz 600Hz 800Hz 2. 0KHz 3. 0KHz 5. 0KHz 7. 0KHz 10KHz 20KHz DB( Ir(Ra5) / ( I(Ra5)* I(Ra5) ) ) DB( - Ii(Ra5) / ( I(Ra5)* I(Ra5) ) ) Fr equency Brüel & Kjær 2003, 12
Traditional Simulation Using LPM IEC-711 coupler LPM simulation Undamped resp. 50 40 20 0-20 -40 100Hz 1. 0KHz 10KHz 20KHz db( V( out *f ) ) Fr equency Brüel & Kjær 2003, 13
Traditional Simulation Using LPM Brüel & Kjær 2003, 14
Traditional Simulation Using LPM IEC-711 coupler LPM simulation Damped response 50 40 Transfer i mpedance of IEC711 coupl er 30 20 10 0-10 100Hz 1. 0KHz 1 0 KHz 2 0 KHz DB(V(out*f)) Fr equency Brüel & Kjær 2003, 15
Transfer impedance simulation (LPM) IEC-711 coupler: Measured data and simulated data using LPM 50 40 30 db 20 Measured data LPM-simulation 10 0 100 1000 10000 100000-10 Frequency (Hz) Brüel & Kjær 2003, 16
Standard FEM Simulation Numerical model solving the Helmholtz equation inside the coupler Models the air inside without visco-thermal effects (loss-free) Allows for computation of acoustic modes (frequency and mode shape) Brüel & Kjær 2003, 17
Transfer impedance simulation (FEM) Frequency response (d0=1um@4.5mm) of full model Pressure (db) ref. @500Hz 50 40 30 20 10 0-10 -20-30 1373 Hz 5052 Hz 12800 Hz 100 1000 10000 100000 Frequency (Hz) Brüel & Kjær 2003, 18
Visco-Thermal fluid Thermal conductivity and viscosity play a role near the boundaries The thermal & viscous mode only exist within a small boundary layer In large spaces, the acoustic mode dominates Acoustic mode Thermal + Viscous + Acoustic mode Boundary Boundary layer Brüel & Kjær 2003, 19
Visco-Thermal FEM Finite Element model based on the narrow gap equation - const. pressure/zero particle velocity in thickness direction - small gap width compared with acoustic wavelength No need to mesh in the thickness direction Includes visco-thermal effects Thin Slit 2D FE Model Brüel & Kjær 2003, 20
Visco-Thermal FEM Test Case 1.11 mm 2.57 mm Thickness: 0.17 mm Brüel & Kjær 2003, 21
2D FE model of Slit hard wall normal velocity excitation open end hard wall hard wall Brüel & Kjær 2003, 22
Simulation on Slit Acoustic input impedance FE simulations vs. lumped-parameter model Brüel & Kjær 2003, 23
Coupler Model Mixed FEM/BEM model (4818 elements & 2337 nodes) Brüel & Kjær 2003, 24
Model Setup Vibrating piston Response point (for transfer impedance simulation) Brüel & Kjær 2003, 25
Coupler Response @ 1000 Hz Sound pressure map Brüel & Kjær 2003, 26
Frequency Response Brüel & Kjær 2003, 27
Conclusion A complete BEM/FEM model of the IEC 711 coupler (B&K Type 4157) taking visco-thermal effects into account was presented. Comparison between a measurement and simulation of the transfer impedance show excellent agreement. A more correct and detailed 3D simulation of the coupler can be done using BEM/FEM compared to traditional LPM. The performance of virtual couplers can be evaluated at the very early design stage using the proposed BEM/FEM technique. Brüel & Kjær 2003, 28
Brüel & Kjær 2003, 29 Any Questions?