Wang X. Vehicle Noise and Vibration Refinement

Подождите немного. Документ загружается.

Noise and vibration reﬁ nement of chassis and suspension 345

but absorbs mechanical energy by transfer into heat. Typical values reached

are up to 25°, in some applications even much more. Note that, as for each

resonating system, the bandwidth and the maximum damping at tuning

frequency are anti-proportional.

Pure rubber bushes (non-hydraulic) provide relatively low damping

values, as a function of the rubber compound and the design (rubber is able

to provide damping under shear load), of typically up to 8–10°. For a

detailed explanation of the physics of rubber for static and dynamic loads,

see Vibracoustic (2002) and Walter (2009), and for the inﬂ uence of the

clamping of bushes upon the stiffness in real mounting cases see Walter

(2009).

Isolators can also be described by means of a quadrupole. This is a

mechanical translation from the electronic domain, in which voltage and

current may be identiﬁ ed with force and velocity. This is not a very common

approach, but it enables one to correctly describe all cross-couplings,

including the crossover from the receiver back to the source. Impedance

misalignment can be optimized more easily, reﬂ ected and absorbed ener-

gies get more demonstrative, and also active mounts can be described. For

details see Sell (2005).

14.6.3 Stiffness and damping measurements

Static and dynamic component stiffness are measured in different ways.

Static stiffness c (Fig. 14.11(a)) is measured with slow sawtooth-like signals

between force limits representing normal usage of the full vehicle (or also

maximum durability loads). The time history of applied force and resulting

mount deﬂ ection is recorded and displayed in a four-quadrant graph. The

hysteresis of the curves is a measure of the component damping. A single

value for the stiffness can be obtained from the slope of the hysteresis curve

at a working point (with or without preload).

Dynamic stiffness k (Fig. 14.11(b)) is measured with sinusoidal stepped

sweeps of constant amplitude. The stiffness at a speciﬁ c frequency is the

quotient of the maximum force and maximum deﬂ ection, the phase angle

between both quantities being the loss angle δ. Stiffness and damping values

are plotted against frequency in the ﬁ gure.

For data compatibility, it is important to deﬁ ne the speed of measure-

ment as well as the amplitude of both the static and the dynamic measure-

ment. Common values for the measurement speed are 10–30 mm/min for

the static data with application-speciﬁ c maximum force values. The ampli-

tude of the sinusoidal dynamic measurement can vary, but for ‘pure’ Road

NVH load cases it may go down to e.g. ±0.025 mm, while spanning frequen-

cies up to 500 Hz, establishing high requirements to the measurement

rig hardware. Note that a mount will be measured softer with increasing

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

346 Vehicle noise and vibration reﬁ nement

amplitudes, which is valid for static as well as for dynamic load cases. For

the amplitude-related reversible and non-reversible effects on the stiffness

(Payne effect, Mullins effect), see Walter (2009).

14.6.4 Dynamic isolation capability

All bushings and mounts are stiffer for dynamic loads than for static loads

(the higher the frequency, the stiffer). Here we have to differentiate

between properties of the compound only and properties of the complete

component. While usually a bush supplier will not discuss pure compound

data with a vehicle manufacturer (which would be data from pure rubber

test probes with common, standardized geometry), all data available for the

vehicle manufacturer is established on the actual design of the rubber–

metal bushes. To still enable a differentiation between compound inﬂ uence

and design inﬂ uences, different values can be established from the static

and dynamic stiffness measurements. In most cases, an estimation of the

dynamic isolation capability of the compound can be obtained from the

ratio of the dynamic stiffness at 100 Hz to the dynamic stiffness at 1 Hz.

This ‘dynamic restiffening ratio’ R

is given by

()

100

(14.1)

and is inﬂ uenced mainly by the compound, while the bush design, i.e. the

ratio of shear loads to compression, largely inﬂ uences the difference

between static and dynamic response. The latter is important in evaluating

the compromise between vehicle dynamics load cases and NVH properties.

So, to quantify the mount behaviour between (quasi-)static load cases with

larger amplitudes and dynamic loads of small amplitudes, we establish a

dynamic-to-static stiffness ratio R

with

3000 1400

1200

1000

800

600

400

200

–3000

2000

–2000

1000

–1000

–4 –3 –2 –1 0

Force (N)

Solid line: stiffness (N/mm)

Dashed line: phase angle (°)

Displacement (mm) Frequency (Hz)

(a) (b)

1234 0 100 200300400 500

14.11 (a) Static and (b) dynamic axial stiffness data. Measured data

for a damper top mount. The resulting static stiffness around the

origin – 430 N/mm – is marked in the dynamic curve at 0 Hz.

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 347

()

100Hz

(14.2)

Note that this latter metric leads to higher values for the stiffening factor,

but might be the preferred one for a vehicle manufacturer, compared to a

bush manufacturer. Typical values are between 1 and 2. To optimize a bush

for low R

makes sense only for the case of a low damping compound. At

locations which require a high damping compound, R

may be much higher,

but the advantage then lies in the damping of oscillations rather than in a

soft decoupling.

Note that the use of ‘c’ for the static and ‘k’ for the dynamic stiffness is

in accordance with VDA 675 460 and 675 480. See there for more details.

Different standards may use different symbols, e.g. ‘k

’ for the static

stiffness.

14.6.5 Shore hardness

The frequently used shore value for the speciﬁ cation of the hardness of the

rubber compound has only restricted meaning with regards to NVH. The

shore hardness cannot replace the correct establishment of the static and

dynamic stiffness of the component. The higher the shore value is, the

higher the dynamic restiffening tends to be. This shows the limitation in

simply modifying bushes for a higher load case by increasing the shore

hardness of the rubber compound. Depending on the application, usually

a shore value of 50–55 should not be exceeded in order to maintain proper

dynamic isolation capability.

14.6.6 Tuning options in geometry and material

For tuning the properties into the desired direction, different options are

available. Either the integration of speciﬁ c features into the mount will

improve the performance, or a review of the basic design (when iterating

from an existing mount to a similar application) can help. For the latter

case, it is important to see the basic requirements for the geometry, to have

a sufﬁ cient bonded area to ensure durability, and to have as few restrictions

as possible to maximize shear loads against compression. So for an existing

design, a marginal change of the areas that are bonded to the inner or outer

core may deliver some reﬁ nement. This may be an option, for example, if

original durability loads and/or peak loads can be detected as overstrict,

then the bonding change potentially reduces either translatoric stiffness or

rotational (cardanic and torsional) stiffness, while keeping the radial prop-

erties. Other options require one to integrate speciﬁ c features or compo-

nents into the mount. For example, the reduced cardanic and torsional

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

348 Vehicle noise and vibration reﬁ nement

stiffness mentioned above can also be obtained by introducing additional

curved inner tubes in a radial mount, transforming compression into shear

load.

Very different load cases in the different directions require different rates

in all directions. This can be obtained either by orienting the mount accord-

ingly (the radial directions are usually much stiffer than the axial one) or

by splitting the radial stiffnesses either by voids in the rubber corpus or by

introducing outside clips or inside inserts to the mount. Snubbers or bump

stops can be internal or external, providing an enhanced tuning range

between soft initial stiffness and increasing progression over deﬂ ection.

Quite new are mounts which are manufactured using different compounds

for speciﬁ c parts of the mount.

14.7 Future trends

14.7.1 Future targets

The future evolution of road-induced NVH targets may be dependent on

the vehicle class. While luxury products still tend to shift the comfort bor-

derline further and further (or lift the compromise between the attributes

to higher and higher levels), in the medium and small vehicle segments

often stagnation can be experienced when comparing new vehicles with

their predecessors. Cost pressures and ﬂ exibility requirements with regard

to modularity may restrict the potential for improvement. Especially for

tyres, it is expected that safety and fuel economy aspects will dominate

development in the near future above the other attributes, both aspects

complicating reﬁ nement improvements.

14.7.2 Robustness and efﬁ ciency

Today, a vehicle platform for a mass product also has to support the deriva-

tion of niche products with relatively low sales volumes but high scheduled

market penetration. So, apart from mass products, which are balanced to

show the performance of all attributes on a relatively high level, the concept

also has to support the tuning of a vehicle towards a sport sedan (tuned

towards vehicle dynamics), a light off-road vehicle (wheel travel, off-road

loads, etc.), a SUV or medium van (high loads), etc. So it makes no sense

to use a suspension concept that can be developed only for one speciﬁ c

application; instead we need a robust tuning bandwidth for all required

vehicle derivatives while keeping the reﬁ nement level high in terms of

kinematic and elastokinematic parameters. Also, the capability of suspen-

sions to work, for example, with an isolated subframe as well as a ﬁ xed

bolted one can be relevant, and compatibility with different springs (e.g.

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 349

steel coil spring versus air spring) and dampers (standard dampers versus

self-levelling), with all-wheel-drive, or with add-on features such as active

steering, active roll control, etc., is important in enhancing the spectrum of

applications. This ﬂ exibility in most cases will be limited automatically by

the amount of ‘overdesign’ that will be accepted for the standard, i.e.

volume application.

14.7.3 Consequences of light weight

For airborne noise attenuation as well as for structural noise transfer, the

mass of the affected components or systems is an important parameter. So,

as a result of overall vehicle weight reduction, the mass law forces an increase

of the transferred airborne noise from tyre patch to the vehicle interior,

unless countermeasures are taken. Since, as always, the primary line of attack

has to be to the main path, which is a very individual one, no general rules can

be given here. In order not to lose the weight advantages with these counter-

measures, it has to be carefully evaluated whether the original isolation

should be reconstituted or whether a secondary means (e.g. interior absorp-

tion) is more efﬁ cient. For structure-borne noise transfer similar relations are

valid. In most chassis locations, pure inertia supports better isolation, espe-

cially in the environment of bushings. So it is again a matter of individual

tuning to achieve comparable structure-borne noise isolation with weight-

reduced components. Modal alignment here plays a major role in adjusting

modiﬁ ed system components to match the vehicle environment.

14.7.4 Future features

Controlled dampers and active dampers

Dampers with continuously controlled damping by means of switchable or

analogue-controllable valves have slowly spread from the luxury class to

mass products, having a moderate potential to combine a smooth ride and

a few noise improvements with good suspension control in the – mostly

only few – required moments. The next generation of dampers, which are

really active, i.e. able to control wheel motion by expanding and compress-

ing the damper actively, are currently still in a prototype phase. They show

much potential to improve ride comfort but face issues of power consump-

tion, controlling algorithms and cost, so it is expected that they will slowly

enter the luxury vehicle class only.

Active mounts and active absorbers

In contrast to the above ride-oriented technologies for dampers, the appli-

cation of active mounts or active absorbers is aimed directly at noise and/

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

350 Vehicle noise and vibration reﬁ nement

or vibration improvements with regard to nominal performance as well as

robustness. Active mounts, positioned in the load path of the mount, have

been found to be difﬁ cult to implement due to the high loads that have to

be transferred through them. In contrast, active absorbers in the vicinity of

the input location can be tuned to minimize the transferred vibrational

energy and can be expected to be used more as they improve construction

efﬁ ciency and integration into the vehicle infrastructure. Depending on the

sensors and algorithm, the target function could be the minimization of

road-induced vibration, tyre ﬁ rst-order imbalance, or the fundamental

engine orders at an engine mount. See Vibracoustic (2002) and Hofmann

(2002).

14.8 References

Arkenbosch, M., Mom, G. and Nieuwland, J. (1992), Das Auto und sein Fahrwerk

(‘Steinbuch’). Stuttgart, Motorbuch-Verlag.

Causemann, P. (1999), Kraftfahrzeugstossdämpfer: Funktionen, Bauarten,

Anwendungen. Landsberg/Lech, Verlag Moderne Industrie.

Gent, A.N. and Walter, J.D. (2005), University of Akron, The Pneumatic Tire.

Washington DC, NHTSA.

Hofmann, M. (2002), Trelleborg GmbH, Antivibration Systems: Fundamentals,

Designs, Applications. Landsberg/Lech, Verlag Moderne Industrie.

Kruse, A. (2002), ‘Characterizing and reducing structural noises of vehicle shock

absorber systems’. ZF Sachs AG, SAE paper 2002-01-1234.

Sandberg, U. and Ejsmont, J. A. (2002), Tyre/Road Noise Reference Book. Kisa,

Sweden, Informex.

Sell, H. (2005), ‘Charakterisierung des dynamischen Verhaltens von elastischen

Bauteilen im Einbauzustand’. Weinheim, Germany, Vibracoustic GmbH.

Torra i Fernández, E. (2004), ‘Tyre air cavity acoustic modes and their inﬂ uence on

the automobile interior noise’. Licentiate Thesis KTH, Stockholm.

Vibracoustic GmbH & Co KG (2002), Schwingungstechnik für Automobile.

Weinheim, Germany.

Walter, H. (2009), ‘Untersuchung der Steiﬁ gkeit von Fahrwerklagern bei idealen

und realen Einspannbedingungen’. Diploma thesis, Rheinische Fachhochschule,

Köln, Germany.

Zandbergen, P. (2008), ‘Design principles for multi-link rear suspensions’. Ford

Motor Company, 6. Tag des Fahrwerks 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

351

Body structure noise and vibration reﬁ nement

G. M. GOETCHIUS, Material Sciences Corporation, USA

Abstract: The body structure of an automobile is a critical part in

determining the overall levels of noise and vibration which reach the

vehicle occupants. It is the body structure which separates, and therefore

insulates, the occupants from the many and varied sources of noise

generated while a vehicle is in operation. As a result, the body structure

must be carefully engineered to block incoming noise and vibrational

energy, and to minimize the effects of whatever energy does ultimately

get into the body structure. This chapter focuses on the basic

mechanisms which affect both structure-borne and airborne noise

transmission in the body, and offers speciﬁ c design practices and generic

targets for many of the attributes which are critical to achieving overall

noise and vibration goals of the vehicle.

Key words: noise, vibration, body, structure, stiffness, panel.

15.1 Introduction

All noise and vibration which reaches the occupants of an automobile must,

by deﬁ nition, travel through the body structure. As a result, the body struc-

ture plays a critical role in determining the overall levels and the quality of

the noise inside the car. How well the body structure blocks incoming

energy and how the various resonances of the body structure (there are

many) affect that energy as it travels through the body structure are critical

issues and will be the focus of this chapter.

This chapter separates the body into three main areas:

1. Global body structure

2. Body attachments

3. Body panels.

For each of these three main areas, best practice design guidelines and

generic targets are offered, along with the description of the physical mech-

anisms which govern their behavior. The intent of the chapter is to provide

the reader with a basic understanding not only of the ‘what’ and ‘why’ of

automotive body NVH, but also of the ‘how’.

Since NVH is a highly complex ﬁ eld, and since speciﬁ c design practices

will always depend on the unique situation of every car design, the design

practices proposed here should be thought of as generalizations. Nonetheless,

Copyrighted Material downloaded from Woodhead Publishing Online

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

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:09:03 AM

IP Address: 129.132.208.2

352 Vehicle noise and vibration reﬁ nement

they form a solid starting point from which a strong foundation can be

constructed and upon which future reﬁ nement can be accomplished.

15.2 Basic principles

The basic principles which govern the behavior of body structure NVH

must be understood before establishing design guidelines and generic

targets. Therefore, this chapter will discuss the fundamental principles of

the source–path–receiver model, the importance of separating structure-

borne noise and airborne noise, the two main ways in which NVH energy

is attenuated (energy impedance and energy dissipation), and ﬁ nally, the

importance of a systemic or ‘holistic’ approach to applying all design

strategies.

15.2.1 Source–path–receiver model

No discussion of noise and vibration would be complete without an intro-

duction to the ‘source–path–receiver’ concept. All noise and vibration phe-

nomena can be broken down into their constituent parts: source(s), path(s)

and receiver(s). In an automobile, there are many sources which are often

acting simultaneously on the body structure. Furthermore, each individual

source can have many paths through which energy can reach the vehicle’s

occupants. Finally, the ‘receivers’ (the human occupants) receive and

perceive multiple noise and vibration signals through different parts of

their anatomy. Figure 15.1 shows a typical set of sources and paths for

powertrain-induced noise and vibration.

Note that the ﬁ rst ‘path division’ is between structure-borne and airborne

noise. This distinction is important since the mechanisms by which noise

and vibration enter the body structure are signiﬁ cantly different between

structure-borne and airborne noise. Also notice that there are multiple

sources and multiple paths, so the resulting energy ﬂ ow is extremely

complex.

The ‘golden rule’ of NVH attenuation with respect to the source–path–

receiver model is to reduce the energy at the source ﬁ rst. This seems logical,

but its importance cannot be overstated. Since most sources (engine, tires,

wind, etc.) have many paths through which energy can penetrate the body,

if the energy is not controlled or stopped at the source, one must ‘chase’

the energy through the many paths into the body. The effort to treat the

paths becomes more difﬁ cult, less efﬁ cient and ultimately more costly than

treating the source.

Having said that, it is often the case that the source is a ﬁ xed quantity

and cannot be changed, and in this case, one has no choice but to work on

the paths. Since the body structure is, in fact, a set of many paths for energy

Copyrighted Material downloaded from Woodhead Publishing Online

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

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:09:03 AM

IP Address: 129.132.208.2

Body structure noise and vibration reﬁ nement 353

to pass through, much of the focus in this chapter will be on addressing path

issues with respect to the body structure.

15.2.2 Structure-borne and airborne noise

It is important to deﬁ ne what is meant by the terms ‘structure-borne’ and

‘airborne’ noise, since the governing physics and the actual body compo-

nents involved in these two areas are markedly different. Structure-borne

noise is deﬁ ned as noise which reaches the car interior by virtue of mechani-

cal vibration transmitted through attachment points of the powertrain,

chassis and other vibration sources. Airborne noise is deﬁ ned as noise

which reaches the car interior by virtue of sound pressure waves impinging

on the various surfaces and panels of the body. It can be seen, then, that

the deﬁ nition of structure-borne and airborne revolves around how the

energy gets into the body structure.

Structure-borne noise is generally limited to the frequency range of

20–600 Hz but can vary depending on the speciﬁ cs of each vehicle. Airborne

noise is generally limited to the frequency range of 400–10,000 Hz but can

also vary depending on the vehicle. Figures 15.2 and 15.3 illustrate the basic

mechanisms for structure-borne and airborne noise. Figure 15.4 shows the

Powertrain noise and vibration

Structure-borne

Source Path Receiver

Engine

mounts

Driveline

Engine

vibration

Occupant

Interior cabin

† Trim panels

† Seats

† Headliner

† Carpet

† Dashmat

Body panels

† Dash

† Front floor

† Rear floor

† Wheelwells

† Roof

† Cowl plenum

† Quarter panels

† Door panels

Body attachments

1. Front mount

2. Rear mount

3. LCA front

4. LCA rear

5. Strut front

6. Strut rear

7. Trans mount

8. UCA front

9. UCA rear

Body panels

† Dash

† Front floor

† Rear floor

† Wheelwells

† Roof

† Cowl plenum

† Quarter panels

† Door panels

Engine

box space

Exterior

space

Intake

noise

Exhaust

noise

Airborne

15.1 Source/path/receiver example.

Copyrighted Material downloaded from Woodhead Publishing Online

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

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:09:03 AM

IP Address: 129.132.208.2

354 Vehicle noise and vibration reﬁ nement

relative contribution of structure-borne and airborne noise contribution for

most automotive applications over the entire automotive frequency range.

15.2.3 Basic vibration attenuation strategies

Body structure NVH control can be broken down into two fundamental

strategies:

Vibrating

source

15.2 Structure-borne noise: basic mechanism.

Vibrating

source

15.3 Airborne noise: basic mechanism.

Structure-borne

100

500

Contribution (%)

Log frequency (Hz)

5000

Airborne

15.4 Structure-borne noise and airborne noise contributions to total

noise.

Copyrighted Material downloaded from Woodhead Publishing Online

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

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:09:03 AM

IP Address: 129.132.208.2