Wang X. Vehicle Noise and Vibration Refinement

Подождите немного. Документ загружается.

Vehicle interior noise reﬁ nement – cabin sound package design 315

mine and requires a certain experience. Furthermore, some of the para-

meters required in the application of SEA are difﬁ cult to obtain in

certain situations and require some speciﬁ c testing measurements.

13.7 Target setting and deployment on

vehicle subsystems

The design of a vehicle’s acoustic sound package can be divided into three

main tasks. The ﬁ rst step is to deﬁ ne the global vehicle noise targets, the

second is deployment of the subsystem’s targets, and the ﬁ nal task is the

design of the sound package for the given targets by an optimisation process.

It is important to underline that the third step cannot be completed prop-

erly without the previous two steps of target setting activities, as a good

sound package solution requires a good balance of the component’s targets.

13.7.1 Vehicle noise targets

The ﬁ rst requirement for designing an acoustic package is knowledge of

the global acoustic targets of the vehicle (Baret et al., 2003). It is important

that the targets should be well focused on the customer’s requests, which

are not easy for acoustic performance, as noise perception is strongly sub-

jective. For this reason, vehicle makers usually conduct surveys among

different customers and collect their judgements from testing different

vehicles in different running conditions. The difﬁ cult part is establishing

objective and measurable targets that are well correlated with the subjec-

tive judgements. This is obtained by measuring different acoustic data (such

as sound pressure level, loudness, articulation index, sharpness, etc.) on the

same vehicles and in the same test conditions as the subjective tests, and

trying to ﬁ nd which measurement results correlate with the subjective

judgements. This allows for deﬁ nition of the vehicle noise targets under

different running conditions.

13.7.2 Deployment at subsystem level

After setting up the previous global vehicle targets the second step is to

cascade these global targets into sub-targets at levels of subsystems (com-

ponents). This is important as the design process is much simpler if a given

acoustic target is deﬁ ned for each of the different subcomponents, which is

often a requirement as some subcomponents can be designed or co-designed

by different suppliers. As an example, the overall vehicle’s acoustic target

in a given condition is ﬁ rst divided into global structure-borne and airborne

noise performance targets. If we consider the airborne noise target we can

then assign a target for the sound isolation and sound absorption of each

Copyrighted Material downloaded from Woodhead Publishing Online

Delivered by http://woodhead.metapress.com

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:07:54 AM

IP Address: 129.132.208.2

316 Vehicle noise and vibration reﬁ nement

component of the cabin (front ﬂ oor, rear ﬂ oor, baggage compartment ﬂ oor,

roof, ﬁ rewall, etc.). This target deﬁ nition is often obtained using the avail-

able simulation methodologies which allow one to ﬁ nd a good balance

between the sub-targets, avoiding the risks of setting up too high or too low

requirements to speciﬁ c components that would result in ﬁ nal design solu-

tions being not well optimised in terms of cost and weight. For the same

reason, intensive benchmarking measurements and tests on existing vehi-

cles are very common in this phase.

13.7.3 Target achievement and optimisation process of the

sound package design

Once the acoustic targets have been deﬁ ned, design of the sound package

can start. The appropriate solutions and materials for both sound isolation

and sound absorption have to be selected according to the given targets:

speciﬁ c experimental tests and numerical calculations allow one to verify

the achievement of the acoustic requirements. A good design process must

be focused on the optimisation of design solutions in order to ﬁ nd the best

compromise in terms of acoustic performance, weight and cost. The cost

must include not only the rough material cost but also the cost, time, prob-

lems and risks related to their application in the vehicle’s production

facilities.

13.8 Conclusions

In this chapter the relevant topics related to the vehicle’s sound package

design have been underlined. After a short introduction describing the

sources and main paths of noise in a vehicle, attention has been paid to the

two main solutions adopted for noise reduction, which are sound isolation

and sound absorption. The chapter then described the material character-

istics and measuring facilities in this ﬁ eld of application, and ﬁ nally some

of the solutions adopted in vehicle sound package design, including the

target setting criteria and a short description of the available simulation

methodologies for interior noise.

13.9 References

Baret C., Nierop G. and Vicari C. (2003), New ways to improve vibro-acoustic

interactions between powertrain and car body, Graz, Austria, 2nd Styrian Noise

Vibration and Harshness Congress

Beranek L. (1960), Noise Reduction, New York, McGraw-Hill

Beranek L. and Ver I. (1992), Noise and Vibration Control Engineering, Principles

and Applications, New York, Wiley-Interscience

Copyrighted Material downloaded from Woodhead Publishing Online

Delivered by http://woodhead.metapress.com

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:07:54 AM

IP Address: 129.132.208.2

Vehicle interior noise reﬁ nement – cabin sound package design 317

Danti M., Nierop G. and Vigé D. (2001), FE simulation of structure-borne road

noise in a passenger car, Lake Como, Italy, NAFEMS World Congress

Danti M., Nierop G. and Vigé D. (2005), A tool for the simulation and optimization

of the damping material treatment of a car body, Traverse City, MI, SAE NVH

Conference

Fuccelli A. (2006), Metodologia numerico – sperimentale per l’analisi del rumore

di rotolamento di un pneumatico trasmesso per via aerea, MSc Thesis, Torino,

Politecnico di Torino, Italy

George A. (1990), Automobile aerodynamic noise, SAE paper 900315

Lyon R.H. (1995), Statistical Energy Analysis of Dynamical Systems: Theory and

Application, Cambridge, MA, MIT Press

Maquet T. (2005), Étude de l’inﬂ uence du matériau et de la géométrie du passage

de roue sur la propagation acoustique du bruit de roulement, MSc Thesis,

Compiègne, Université de Technologie de Compiègne, France

Mezzomo F.F. (2003), Determination of the acoustic transparency in laminated

glasses, MSc Thesis, Torino, Politecnico di Torino, Italy

Roger M. (1996), Fundamentals of Acoustics, von Karman Institute for Fluid

Dynamics, Lecture Series 1996–04

Vigé D. (1999), Modelli predittivi di campi aeroacustici per applicazioni automo-

bilistiche, MSc Thesis, Torino, Politecnico di Torino, Italy

Vigé D., Bonardo E., Chiesa L., Vanolo M. and Vicari C. (2008), An hybrid experi-

mental–numerical methodology for the prediction of the external rolling noise in

running conditions, Graz, Austria, Styrian Noise, Vibration and Harshness

Congress

Copyrighted Material downloaded from Woodhead Publishing Online

Delivered by http://woodhead.metapress.com

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:07:54 AM

IP Address: 129.132.208.2

318

Noise and vibration reﬁ nement of chassis

and suspension

B. REFF, Ford-Werke GmbH, Germany

Abstract: Today, people spend more time in their cars, as average trafﬁ c

speeds are reduced while distances travelled increase. So noise and

vibration performance gain in importance, while the car should still offer

an active driving experience in case the road allows. This chapter deals

with the inﬂ uence of the suspension type and the adjustable suspension

parameters upon the achievable reﬁ nement. Special insight into the

acoustic properties of tyres, suspension components and isolators is

given, including from the viewpoint of tuning, weight restrictions and

other expectations.

Key words: road NVH, road-induced noise, comfort, impact, tyres,

isolation.

14.1 Introduction

Nowadays, many people drive longer distances in their cars, partly for

private purposes, partly for commuting to and from work. With average

speeds tending to decrease, people are spending more time in their cars, so

comfort in the vehicle is becoming more important. People want to be

transported as comfortably as possible with little tiredness, while the vehicle

should still offer an active and agile driving experience in situations when

trafﬁ c allows it.

The focus of this chapter is on noise and vibration optimization of the

chassis. This area contains all the audible responses that are due to the

vehicle rolling on the road, and also the respective vibrations, as perceived

by the driver and passengers. The lower frequencies of these local vibra-

tions overlap with the low-frequency motions of the full vehicle on undulat-

ing roads, which are known as ride comfort. This overlap is in the range of

the so-called secondary ride, which spans frequencies from 3 to 20 Hz. The

resulting responses can be airborne or structure-borne, as well as the excita-

tions. The area of chassis reﬁ nement is also called ‘Road-induced NVH’

where NVH stands for noise, vibration and harshness (with reference to

the exciting source, in contrast to, e.g., ‘powertrain NVH’). In this chapter,

‘road-induced NVH’ and ‘chassis reﬁ nement’ are used synonymously.

Copyrighted Material downloaded from Woodhead Publishing Online

Delivered by http://woodhead.metapress.com

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:07:21 AM

IP Address: 129.132.208.2

Noise and vibration reﬁ nement of chassis and suspension 319

All items that are relevant for the attenuation of other structural or air-

borne excitations (e.g. powertrain noise) are also relevant for road-induced

NVH. Also, all foundations of the body treatment for increased isolation

as well as for radiation control and sound attenuation or absorption are

also valid for road-induced NVH. Hence, the reader is asked to refer to the

respective chapters in this book. Many relevant topics can be mentioned,

but most topics must be kept very short, so that the references given at the

end of the chapter should be used to go into more detail.

14.2 Road-induced noise, vibration and harshness

(NVH) basic requirements and targets

14.2.1 Load cases: cruising versus impact

Typically a cruising condition is the vehicle excitation due to the road

surface texture and the tyre–road interaction, and is regarded as time-

invariant. An impact condition is due to single or repeated events. Cruising

data is mostly analysed in the frequency domain by means of a stationary

frequency analysis. A constant-bandwidth analysis by means of a fast

Fourier transform (FFT), usually as an autopower spectral density (PSD),

gives detailed information about the frequency content, e.g. resonances,

and allows analysis of the root cause by comparing auditory or tactile

response with chassis accelerations, forces, etc., or by means of more

sophisticated approaches like transfer path analysis (TPA); see also Section

14.2.3. TPA is based on the assumption of linearity around a working point

of all relevant systems and components, which can be expected as long as

the road excitation levels are not too high. Impact data can be analysed in

the time domain as well as in the frequency domain. The time domain

analysis can take into account non-linearities in the suspension due to larger

chassis deﬂ ections, because all non-linear relationships can be displayed

correctly (e.g. force over deﬂ ection in a chassis bush).

In order to additionally obtain frequency domain information, which is

very helpful in tracing back for resonances, it is not sufﬁ cient to simply

apply a kind of FFT on a time record of a microphone or acceleration

responses. In order to enhance the poor frequency resolution of a short

time history of an impact response, sophisticated techniques are required

to separate front and rear suspension responses as well as to average away

undesired uncorrelated noises by triggered averaging of several measure-

ment runs over one type of impact. Due to the time-variant and ﬁ nite

character of the signal, the analysed spectra are energy spectral densities

(ESDs).

Figure 14.1 shows examples of the time history and the spectra for cruis-

ing and impact load cases of a C-car vehicle. Note that in order to show the

Copyrighted Material downloaded from Woodhead Publishing Online

Delivered by http://woodhead.metapress.com

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:07:21 AM

IP Address: 129.132.208.2

320 Vehicle noise and vibration reﬁ nement

10 100

SP PSD (dB(A))

–1

–2

–3

–4

0 0.2 0.4 0.6

Time (s)

0.8 1

Sound pressure (Pa)

–5

–10

–15

–20

–25

0 0.1 0.2 0.3 0.4

Time (s)

0.5 0.6 0.7 0.8

Sound pressure (Pa)

Frequency (Hz)

(a)

(b)

(c)

(d)

1000

10 100

SP ESD (dB(A))

Frequency (Hz)

1000

14.1 Time history and frequency responses for cruising and impact cases: (a)

time history, cruising; (b) frequency response, cruising; (c) time history, impact;

(d) frequency response, impact. PSD = autopower spectral density; ESD =

energy spectral density.

stochastic characteristics of the cruising excitation, the time history of the

signal spans only one second, which is, of course, much too short for a

statistically relevant analysis. Similarly for impact, the response for only

one impact event is shown – one local peak for the front suspension hitting

the impact, and one for the rear suspension. The frequency content shown

is an average of several of these events. With regard to the spectral content,

both excitation types resemble one another in the frequency analysis but

have different spectral balance. In addition, due to the non-linearities of

the vehicle that are especially relevant for the impact case, the relative

amplitudes of different peaks will be different between both excitations,

and slight frequency shifts can also occur.

14.2.2 Road types and spectra

Public roads show a large span of excitation levels and spectra (Fig. 14.2).

Textural levels vary from smooth asphalt surfaces with subjective domi-

nance of tyre airborne noise through coarse aggregate chipping surfaces,

which are subjectively dominated by low-frequency structure-borne noise,

Copyrighted Material downloaded from Woodhead Publishing Online

Delivered by http://woodhead.metapress.com

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:07:21 AM

IP Address: 129.132.208.2

Noise and vibration reﬁ nement of chassis and suspension 321

up to rough surfaces with bumps, holes and undulations, which are often a

combination of a constant cruising load case with arbitrarily repeated single

events.

The amplitude of road excitation spectra (Fig. 14.3) usually decreases

with frequency, but the slope can differ between the surfaces, and also some

wavenumber ranges may be emphasized by, e.g., road manufacturing issues,

or pavement structures such as Belgian Blocks.

For setting spectral targets for interior road noise and cascaded quanti-

ties, it is important to know the properties of the reference surface, which

should have a smooth spectrum. Only if this is given can one identify

peaks in the response with vehicle root causes, rather than seeing a quite

undeﬁ ned mixture of road excitation peaks and vehicle responses. Figure

14.4 shows a surface which is not optimal for development purposes,

because it creates a high vehicle response around 90 Hz due to its uneven

spectrum.

(a) (b) (c)

14.2 Examples of (a) smooth, (b) coarse and (c) rough surfaces.

Smooth and coarse: macrotextural view (20 × 13 cm); rough: inclined

view on area approximately 1 × 1.5 m.

–5

–6

–7

–8

–9

–10

110

n (m

–1

)

Smooth

Coarse

Rough

Displacement PSD (m

)

100

14.3 Spectra of smooth, coarse and rough surfaces.

Copyrighted Material downloaded from Woodhead Publishing Online

Delivered by http://woodhead.metapress.com

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:07:21 AM

IP Address: 129.132.208.2

322 Vehicle noise and vibration reﬁ nement

14.2.3 Target quantities for full vehicle and target cascading

On the full vehicle level, aural as well as tactile responses can be targeted.

The aural responses are sensed by microphones or acoustic heads, the

tactile responses by accelerometers, e.g. at seat track, steering wheel, gear

lever, pedal box, etc. The quantities can be frequency spectra, band levels,

overall levels or characteristic values of the time history of the responses.

Spectra can have an equally spaced, narrowband x-axis, which enables

tracing back resonances and combination with narrowband transfer func-

tions, or can have fractions of an octave (typically three or 12, rarely 96) as

the x-axis. A 1/3-octave analysis results in better correlation to the subjec-

tive spectral balance, while a 1/12-octave achieves almost the frequency

resolution of a narrowband FFT for lower frequencies but still shows lower

‘noise’ in higher frequency ranges. Lower frequencies are dominated by

structure-borne noise, higher frequencies by tyre airborne noise; the bor-

derline is between 400 and 800 Hz, depending on speed and road. All of

these considerations are summarized in Fig. 14.5 and Table 14.1.

Many applications exist which calculate complex quantities from the

measured physical values for sound pressure or acceleration. These complex

quantities are able to better correlate to subjective perception, but usually

are hardly cascadable, especially when reduced to two or even one dimen-

sion. Examples are psychoacoustic loudness, articulation index (AI), and

values describing human tactile perception.

–7

–6

–5

–4

–3

–2

–1

100010010

Frequency (Hz)

Displacement PSD (m²/Hz)

–10

–9

–8

–7

–6

–5

–4

Vehicle response (MIC)

Exc. profile for 30 km/h

Sound pressure PSD (Pa²/Hz)

14.4 Comparison of road excitation spectrum and resulting interior

noise. Note that the area around 80 Hz is emphasized not due to the

vehicle, but due to the road.

Copyrighted Material downloaded from Woodhead Publishing Online

Delivered by http://woodhead.metapress.com

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:07:21 AM

IP Address: 129.132.208.2

Noise and vibration reﬁ nement of chassis and suspension 323

14.3 Foundations of road-induced noise, vibration

and harshness

Three fundamentals are deﬁ ned for the vehicle potential of road-induced

NVH performance. The ﬁ rst is the suspension concept, to be chosen in the

framework of vehicle class, performance targets, manufacturing restric-

tions, existing vehicle platform requirements, and speciﬁ c desired perfor-

mance proﬁ les (e.g. handling versus comfort dominated). This deﬁ nes the

range of the ﬁ nal reﬁ nement performance. Of course, the reﬁ nement ranges

of different suspension types overlap. Secondly, within a speciﬁ c suspension

Interior noise,

road excited content

Body system:

airborne tyre

noise transfer

Tyre and rim:

airborne noise

radiation

Airborne

Structure-borne

Structure input

Road to tyre

Tyre and rim:

structural noise

transfer

Chassis systems:

structural noise

transfer

Body system:

structural noise

transfer to acoustic

14.5 Vehicle systems cascade.

Table 14.1 Vehicle targets and cascaded targets (generic)

System Structure-borne

noise

Airborne (tyre)

noise

Vibrations

Full vehicle Interior noise

spectrum for

coarse road

Interior noise

spectrum

for coarse

road

Vibration levels/spectra

at comfort points

Body system NTF (Pa/N) NTF (Pa/Pa) VTF (m/s

/N)

Chassis system VTF (N/N)* n.a. VTF (N/N)*

Tyre–rim system Spindle force for

coarse road

Nearﬁ eld

noise (Pa)

Spindle force for

coarse road

* Different units possible, depending on approach.

Copyrighted Material downloaded from Woodhead Publishing Online

Delivered by http://woodhead.metapress.com

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:07:21 AM

IP Address: 129.132.208.2

324 Vehicle noise and vibration reﬁ nement

concept, the achieved kinematics can vary quite a lot. As an example, the

damper position can be very different. The lower attachment position of

the damper and its inclination inﬂ uence the damper ratio, the ratio inﬂ u-

ences the required upper attachment mount stiffness, and this together with

the mount design and the attachment stiffness of the vehicle body deﬁ nes

the degree of noise isolation at the damper path. Thirdly, only the tuning

of the components enables the greatest compromise to be achieved between

a concept and its kinematics. Such a compromise may partly compensate

for weaknesses of one of the other foundations.

We start with some details of the concepts and the kinematics. The

component potential is handled in Sections 14.4–14.6.

14.3.1 Suspension concepts

In the ﬁ rst instance, the list of functional requirements to a suspension is

quite complex: see Table 14.2 for a rear suspension (it is similar for a front

suspension, plus some additional steering-related items).

Table 14.2 Key requirements for rear suspensions

Factor Requirements

Camber, castor, toe Toe and camber low for tyre wear. Toe adjustable with

minimal effect to camber (enabled by low castor

angle).

Bump camber (Negative) camber angle should increase with bump

travel.

Roll centre Roll centre not too high for straight-ahead stability.

Longitudinal wheel

recession

Wheel should move rearward at bump travel for good

impact harshness.

Longitudinal

compliance

In addition the suspension should be compliant in the

longitudinal direction.

Aligning torque

compliance

Suspension should be stiff around the wheel z-axis to

provide understeer for higher pneumatic trails

(robust for varying tyre widths).

Lateral compliance

steer

Toe-in under lateral forces is required for cornering

stability, especially for RWD vehicles.

Longitudinal

compliance steer

Toe-in under braking forces is required for brake in

turn stability.

Camber compliance High camber stiffness or lateral stiffness is required for

straight-ahead stability, steering feel and high agility.

Castor compliance Castor change under braking or lateral forces to be

minimized to increase stability and minimize shake

and brake judder.

Anti-lift, anti-squat High anti-lift and anti-squat are required to minimize

vehicle pitch at braking and acceleration.

Source: Zandbergen (2008).

Copyrighted Material downloaded from Woodhead Publishing Online

Delivered by http://woodhead.metapress.com

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:07:21 AM

IP Address: 129.132.208.2