Simulation of noise treatments in aircraft

5th European HyperWorks Technology Conference 111107

Peter Davidsson

Simulation of noise treatments in aircraft

Creo Dynamics AB

• Started in January 2010

• Office in Linköping and Lund

• Key persons with background from A2 Acoustics and Saab Aerospace

• 13 employees

• Extensive experience in Aerospace Acoustics

• SME

• Link between the research community and industry

• Multidisciplinary acoustic challenges: acoustics, structural dynamics, fluid mechanics and composites

Competences • Vibro-Acoustic FEM

• Propeller Noise

• ECS Noise

• Noise & Vibration Measurements and Analysis

• Active Noise Control

• Tuned Vibration Absorbers

• Acoustic Liners

Creo Dynamics – Strategi

Creo Dynamics – Kompetenser

Akustik

Creo Dynamics utvecklar både aktiva och passiva lösningar för att förbättra ljud och vibrationsegenskaper hos produkter

Aero-/Termodynamik

Experter inom CFD och termodynamiska beräkningar

Strukturdynamik

Experter inom såväl strukturdynamik som vibro-akustik och akustisk utmattning

Kompositer

Design och analys av produkter i kompositmaterial

6

Aerospace acoustics A 400M Saab 2000 and Gripen

Controller

Cabin acoustics

First modes of a cabin structure

Noise box – trim panels

System design Pre study

Measurement

Correlation

Noise box – trim panels

Panel

Porous material – structural domain

Porous material – fluid domain

Acoustic cavity

Biot’s formulation for porous material

Generic Car; Rear Side Window Buffeting

Microphone location

Buffeting 110 Db, 22Hz

A2Z-PS-08-028 2011-11-14

11

Aerospace acoustics - A400M

Integrated optimization

• Increasing interest in turboprop and open rotor powered aircraft

• Active Noise and Control Systems development starts when aircraft structure and cabin interior design is fixed

• Far from optimum due to severe constraints, e.g. for actuator locations and attachments

Integrated optimization

• Identified need for joined optimization of both structure, cabin interior and noise control system

• Hyperworks products very suitable for integrated optimization

• Radioss

• Hyperstudy

• Optistruct


Monitor microphones

Control microphones Actuators

External pressure field

act

act

monextmon TF Fpp ][


• Aim:

– Minimize noise level in monitor microphones

– Limited to low frequency tonal noise

Primary field Total field

act

act

monextmon TF Fpp ][


• Aim:

– Minimize noise level in monitor microphones

• Means:

– Passive noise control system

– Active noise control system

• Design

– Finite element simulations for evaluation of system properties

– The actuator and sensor location determines the system performance


• Simulations can be used for studying:

– Potential in different treatments

– Size of the system

– Mounting conditions

– Sensitivity in modifications in the structure and acoustic cavity

– Sensitivity to changes of external pressure field

– …

Simulation procedure

• Finite element model generation

• External pressure field

• Preloading

• Primary field

• Dynamic condensation

• Optimization – Actuator properties

– Actuator and control sensor location

Finite element model generation

p

d

d

0

0

F

0

F

F

FRFFRFFRF

FRFFRFFRF

FRFFRFFRF

s

d

d

d

ext

s

ext

d

pppspd

spsssd

dpdsdd

FdKM 2

FFRFFDd 1

Actuator dofs and monitor dofs

Monitor

TVA, shaker

Helmholtz, loudspeaker

External pressure field

• Shortcut (no CFD)

• BEM including flow

• Mapped to FEM

0

F

F

FRFFRFFRF

FRFFRFFRF

FRFFRFFRF

p

d

dext

s

ext

d

pppspd

spsssd

dpdsdd

ext

ext

s

ext

d

Preloading due to cabin pressure

Primary respons, structure

2.BPF 1.BPF

Primary respons, acoustic cavity

d

dd

ext

d

d

pppspd

spsssd

dpdsdd

ext

ext

s

ext

d

s

d

FFRFd

0

0

F

FRFFRFFRF

FRFFRFFRF

FRFFRFFRF

p

d

d

p

d

d

How do we get the actuator force?

The equation system can now be written:

2.BPF 1.BPF


• Available treatments

Actuator Sensor Actuator force

Passive Dynamic vibration absorbers

F from local displacement

Helmholtz resonator

Q from local acoustic pressure

Active Shaker/ Piezo actuator

Accelerometer F from ANC system

Microphone

Loudspeaker, Active panel

Microphone Q from ANC system

Still, how do we get the actuator force?

Passive, Tuned vibration absorbers

• Change the dynamic stiffness, do not absorb vibration

m

)1(~

iLFkk

)(sx

)(dx

m

dF

)(2 smountd xmF

mmount

sds

n

nd xhxkF

1

22

2~

dsd xmxxk 2~

sdd xxkF ~

Force in the spring: Equiv. dyn mass Force can be written

Passive, Tuned vibration absorbers

• Change the dynamic stiffness, do not absorb vibration

m

)1(~

iLFkk

)(sx

)(dx

m

dF

)(2 smountd xmF

mmount

d

ddd

ext

dd FFRFdd

dddddddd

d

d h dIdHF

ext

ddddddd

d

d dHFRFIHF1

For the actuator dofs: The TVA force Force can be written

A system of size equal to the number of included TVA’s needs to be solved for each configuration

Active Noise Control system Actuator forces

IFFQeeTT

– Acoustic pressure in the control microphones

– The primary response in the control microphones from the external

pressure field

– The actuator forces

– The frequency response functions between the force actuators and

control microphones

– Determine the control effort (leak factor in the LMS-algorithm)

- Determine the influence of each microphone on the cost function

Object to minimize the function

e

TT

act QpTFITFQTFF ][][][1

act

act

ctrle FTFpe ][

ep

e

actF

TF

Q

A2Z-PS-08-028 2011-11-14

31

Actuator forces

• The passive system reduce the local vibration/pressure level, based on the local properties.

• The ANC system determines the driver signals for the actuators, i.e. the actuator forces, based on the SPL in the control sensors.

– In the cabin, the control microphones are placed at the trim panels.

A2Z-PS-08-028 2011-11-14

32

Analysis procedure

• The aim is to find the optimal configuration of actuators and control microphones in order to minimize the noise inside the cabin.

• The primary response and FRF’s are derived once for each frequency line

• The best configuration is then searched for in an optimization procedure

– Finding the “optimal” force

Optimization procedures

• Different procedures may be used to optimize a noise controlling installation: – Maximum amplitude. The tuned vibration absorbers are

placed in the positions where the maximum displacement occurs in the baseline analysis.

– Sequential maximum amplitude. The first tuned vibration absorber is placed where the maximum vibration amplitude occurs in the baseline analysis. The system is then re-analyzed and the next damper is placed where the vibration now has its maximum. This is repeated until all DVA-positions are determined.

– Simulated annealing • Both active and passive systems

34

Optimization procedure (Simulated annealing)

• Present configuration

• All available locations

• Monte Carlo simulations

– The new configurations are derived by randomly choosing a number of devices (actuators and microphones) from the present configuration and randomly choosing a number of devices from all possible positions.

– Annealing factor

– Only solving a system with size equal to the number of actuators

Reduction in monitor nodes 1.BPF

PassiveTVA’s

ANC, Shakers

Primary field

Reduction in monitor nodes 2.BPF

PassiveTVA’s

ANC, Shakers

Primary field

System size

PassiveTVA’s

ANC Shakers and microphones

Results example

• Potential in noise reduction Actuator Sensor Potential

1.BPF 2.BPF 3.BPF

Passive Dynamic vibration absorbers

8-12 4-7 0

Helmholtz resonator

8-12 4-7 0

Active Shaker Accel. 10-15 5-8 1-3

Microphone 15-25 8-12 2-4

Loudspeaker Microphone 15-25 8-12 3-6

Thank you

• Contacts:

– Peter Davidsson

[email protected]

– Gustav Kristiansson (VD)

[email protected]

mailto:[email protected]

mailto:[email protected]