Wang X. Vehicle Noise and Vibration Refinement

Подождите немного. Документ загружается.

Vehicle vibration measurement and analysis

X. WANG, RMIT University, Australia

Abstract: This chapter introduces the basic concept of vibration

measurement, illustrates the pre-requirements for vibration testing and

environment testing, and provides tips for the installation of vibration

objects. The quotation method for vibration levels and the measurement

method for complex modulus and damping loss factor are also discussed.

The principles of vibration isolation and vibration absorber are briefed.

Two case studies are given to illustrate how the principles and methods

are applied to solve engineering problems.

Key words: vibration transducers, accelerometers, subjective evaluation,

hand sensing, proximity sensors, velocity sensors, charge ampliﬁ er,

power supplier, Young’s modulus, damping loss factor, human vibration

limit, vibration isolation, vibration absorber.

3.1 Introduction

After completing benchmark study and setting vehicle development targets,

vehicle noise and vibration reﬁ nement starts with a vehicle development

process. Vibration is one of the most common customer complaint issues.

Vibration may be caused by faulty engine combustion, wheel–tyre imbal-

ance, prop shaft imbalance, driveline angle induced torsional vibration,

brake shudder, etc. In order to identify the root cause of vibration, a series

of diagnostic tests has to be conducted. Vibration measurement plays a key

role in the vehicle development process. In order to avoid vibration prob-

lems in the early design stages, vibration isolation and absorber design have

to be conducted; in particular, powertrain vibration isolation design has to

be carefully carried out to minimize the structure-borne vibration transmit-

ted from power plant to chassis and body structure. This chapter introduces

the vibration measurement fundamentals and quotation indexes, and illus-

trates design principles and methods of vibration isolation and vibration

absorption for solving engineering problems. This includes the subjective

vibration evaluation method as illustrated below.

3.2 Hand sensing

Humans are sensitive to vibrations. Before any instruments became avail-

able all vibration analysis was done by listening and feeling. This method

Copyrighted Material downloaded from Woodhead Publishing Online

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

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:01:01 AM

IP Address: 129.132.208.2

34 Vehicle noise and vibration reﬁ nement

is still used by those who do not have access to instruments. We have some

built-in sensors in our skin and our ears. These biological transducers

served a survival function and still do. The human ear is amazingly sensitive

to the smallest pressure change. The pressure sensors in our skin can detect

constant pressure, and also oscillating pressure or vibration.

Humans can directly detect vehicle noise, loudness and vibration smooth-

ness by listening and feeling before any electronic instruments are con-

nected. This is a valid sensing method for vibration analysis if certain

precautions are taken, namely calibration and frequency analysis. It is dif-

ﬁ cult to compare one vehicle with vibration ‘readings’ separated by several

days, especially if many similar vehicles are seen during the interval. The

‘measurements’ are also highly subjective. One person’s judgement of

‘rough’ could be another person’s ‘acceptable’. This system of human per-

ception gives an overall vibration reading. The best that can be obtained

with hand sensing of vibration is a crude, overall, subjective vibration mea-

surement that sounds an alarm when mechanical failure is imminent. The

method of hand sensing works satisfactorily in vehicle noise, vibration and

harshness (NVH) departments where some human calibration techniques

are employed by NVH engineers and technicians. First, one NVH engineer/

technician is given responsibility for a speciﬁ c vehicle line program, which

is his or hers and no-one else’s. So the human variable is removed. Second,

this one person checks the vehicle on a daily basis, so there is not a long

interval between ‘measurements’; and third, there are usually identical

vehicles nearby to compare against. Using these methods of calibration the

NVH engineer/technician can be successful in identifying noise and vibra-

tion problems within the vehicle by human hand feeling measurements.

In order to identify the sources of the noise and vibration problems,

frequency analysis is necessary. Humans are equipped with a frequency

analyser. The combination of the human ear and brain is actually a pretty

good spectrum analyser and is extremely sensitive. Human hearing has a

sensitive bandwidth of about 40 Hz or less for people with a good musical

ear. The human ear is sensitive to air vibrations from about 20 Hz to

20,000 Hz. This is also the frequency range of most annoying vibrations of

mechanical equipment. 30 Hz is barely audible for most people but it can

certainly be felt with the pressure sensors in the ﬁ ngertips. A metal object,

such as a coin, can be held between the ﬁ ngertips while probing for vibra-

tions. Evidently there is some ampliﬁ cation from the metal object pressed

against a vibrating surface. Hand sensing works for low-frequency vibra-

tions (less than 100 Hz). For higher frequencies, you should listen to the

tones, as the majority of mechanical vibrations are within the frequency

range of the human ear to detect.

Using the human ear for frequency analysis of vehicle noise is more

effective when the ear can be coupled directly to the vehicle. This means

Copyrighted Material downloaded from Woodhead Publishing Online

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

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:01:01 AM

IP Address: 129.132.208.2

Vehicle vibration measurement and analysis 35

using a stethoscope, a wooden stick or a screwdriver. Any vibration analysis

should always be done ﬁ rst by listening and feeling to give the analyst a

subjective frame of reference.

If no objectionable vibration or noise can be felt or heard, you can

be conﬁ dent that there is probably no vibration problem in terms of

mechanical design and wear. Humans are well calibrated in this respect to

mechanical damage criteria, but this judgement depends on no interfering

background noise.

Human judgement of vibration severity can be quantiﬁ ed by the level of

acceleration:

• 0.001–0.01 g is the threshold of perception.

• 0.1 g is considered unpleasant.

• 0.5 g is considered intolerable by most subjects.

Vibration problems and human sensitivity to them are well correlated in

the absence of background vibration. If vibration is uncomfortable to

people, then it is probably causing serious damage to the vehicle or machine.

As vibration analysts we are looking for vibration. We want to amplify and

take a closer look at it. This is the whole purpose of vibration instruments

– to amplify vibration and display it in a way that we can better

understand.

Instruments increase the signal-to-noise ratio. Suppose you wanted to

examine the vibration from one of three running pumps. You could

approach the pump of interest and place your hand on it, or couple your

ear to its housing with a wooden stick. This will increase the signal of inter-

est above the background noise. Alternatively, you could turn everything

else off and run only the pump of interest – this will reduce the background

noise and again increase the signal-to-noise ratio. By placing a transducer

directly on a vehicle structure of interest, we couple directly to that source

of vibration and separate out the background vibrations. The signal-

to-noise ratio is increased.

To perform any signiﬁ cant vibration analysis the important parameters

to measure are frequency and amplitude. Hands and ears can make both

of these measurements subjectively. The rotational speed is the most fun-

damental number required to do any vibration analysis. A stroboscope with

accurate speed readout, or a tachometer, should be the ﬁ rst instrument to

be purchased. The stroboscope is, however, not reliable for ‘stopping’

motion below 300 rpm.

The methods described above using the hands, ears and a stroboscope/

tachometer (and perhaps a stethoscope, wooden stick or screwdriver) are

not yet outdated. Keep these tools available in your toolbox and use them

frequently to correlate with your instruments as a check.

Copyrighted Material downloaded from Woodhead Publishing Online

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

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:01:01 AM

IP Address: 129.132.208.2

36 Vehicle noise and vibration reﬁ nement

3.3 Basic vibration measurements

Ears, hand and watch/timer can be subjectively used to sense vibrations.

Most of us, however, want a hard number that we can compare to.

Transducers, along with a readout instrument, can provide this number.

3.3.1 Types of vibration transducer

A transducer is a device for converting the mechanical motion of vibration

into an electrical signal. It is commonly called a ‘pickup’. There are gener-

ally three types of vibration transducers: displacement, velocity and accel-

eration. The most common type of displacement transducer is the proximity

probe as shown in Fig. 3.1. It operates on the eddy current principle. It sets

up a high-frequency electric ﬁ eld in the gap between the end of the probe

and the metal surface that is moving. The proximity probe includes an

oscillator–demodulator and a cable.

The proximity probe senses the change in the gap and therefore measures

the relative distance, or displacement, between the probe and the tip of the

surface. The probe measures the gap between 10 and 90 mils. The practical

maximum frequency that proximity probes can sense is about 1500 Hz. The

minimum frequency is 0 Hz – it can measure static displacement.

The velocity transducer is an adaptation of a voice coil in a speaker. It

consists of an internal mass (in the form of a permanent magnet) suspended

Moving

metal shaft

Rigid

support

Gap

Proximity

probe

Eddy

current

3.1 Proximity probe.

Copyrighted Material downloaded from Woodhead Publishing Online

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

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:01:01 AM

IP Address: 129.132.208.2

Vehicle vibration measurement and analysis 37

on springs. Surrounding this mass is a damping ﬂ uid, usually oil. A coil of

wire is attached to the outer case as shown in Fig. 3.2.

In operation, the case is held against and moves with the vibrating object,

while the internal mass remains stationary suspended on the springs. The

relative motion between the permanent magnet and the coil generates a

voltage that is proportional to the velocity of the motion, hence it is a

voltage pickup.

Velocity transducers provide a direct measure of vibration velocity (SI

unit – m s

−1

, although mm s

−1

is commonly used). The main points regarding

velocity transducers are:

• They are often moving coil devices – an electric coil is suspended around

a permanent magnet that is kept static.

• They are designed to have a low natural frequency (typically a few Hz),

chosen to be much lower than the anticipated excitation frequency

range (typically 10–1000 Hz).

• They are usually large and fairly heavy, but robust.

• They cannot be used on lightweight structures due to the mass loading

effect – they change the vibration characteristics of the component

whose vibration they are trying to measure.

• Multi-axis transducers are available. These usually stand on their own

base – often an adjustable tripod is provided to cater for use on sloping

or uneven surfaces.

Springs

Coil

Permanent

magnet

Damping

fluid

Isolating

washer

Vibrating

machine

surface

3.2 Velocity transducer.

Copyrighted Material downloaded from Woodhead Publishing Online

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

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:01:01 AM

IP Address: 129.132.208.2

38 Vehicle noise and vibration reﬁ nement

• They are simple – low-cost voltage ampliﬁ cation is commonly used.

Battery usage is relatively efﬁ cient and the transducer and its signal

conditioning are often housed in a single rugged package.

The most common acceleration transducer is the piezoelectric acceler-

ometer (Fig. 3.3). It consists of a quartz crystal (or barium titanate – a

manufactured quartz) with a mass bolted on top. In operation, the acceler-

ometer case is held against the vibrating object and the mass wants to stay

stationary in space. With the mass stationary and the case moving with the

vibration, the crystal stack gets compressed and relaxed. The piezoelectric

crystals generate a charge output that is going positive and negative as the

crystals are alternately compressed tighter and relaxed about the preload.

The charge output is proportional to force, therefore acceleration according

to Newton’s Second Law; hence the name accelerometer.

The accelerometer is self-generating, but the signal output has such high

impedance that it is not usable by most analysis equipment. The output

impedance of the accelerometer must be matched to the impedance of the

readout instrument for accepting the signal from the accelerometer. The

output signal from the accelerometer must be converted to a low-

impedance signal by special electronics. This electronic circuitry can be

outside or inside the accelerometer. There are two types of accelerometer

that you can explore with the following links:

Springs

Connector

(case ground)

Internal

electronics

Piezoelectric

crystal stack

Isolating

washer

Vibrating

machine

surface

Seismic

mass

Bolt

compressing

crystals

3.3 Piezoelectric accelerometer.

Copyrighted Material downloaded from Woodhead Publishing Online

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

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:01:01 AM

IP Address: 129.132.208.2

Vehicle vibration measurement and analysis 39

• Voltage type

• Charge type.

The charge-mode accelerometer must have a charge ampliﬁ er nearby. This

provides the proper impedance matching. The short piece of cable between

the accelerometer and the charge ampliﬁ er is critical. It must be a low-noise

cable and its length cannot be changed. The accelerometer, cable and

charge ampliﬁ er must be calibrated as a unit and not changed.

The voltage-mode accelerometer has the impedance-matching electron-

ics built inside. It needs no charge ampliﬁ er, and the cable length is not

critical. All it needs is a low-cost power supply to power the internal elec-

tronics. Some readout instrument manufacturers provide built-in power

supplies to power their electronics.

It is important to use charge ampliﬁ ers with charge accelerometers and

power supplies with voltage accelerometers. When the electronics are

built into the accelerometer, this simpliﬁ es their use. The voltage acceler-

ometers are labelled ICP (integrating circuit piezoelectric) by some manu-

facturers. Accelerometers provide a direct measure of vibration acceleration

(SI unit – m s

−2

, or g, where 1 g = 9.81 m s

−2

The main points regarding accelerometers are:

• They are designed to have a natural frequency well above the antici-

pated excitation frequency range.

• They offer a wide frequency range from 0.2 Hz to 10 kHz for better

than 5% linearity.

• They are usually small in size and light in weight (can be less than 1

gram).

• For high-temperature applications the charge mode system should be

used.

• The accelerometer is sensitive to vibration frequencies much higher

than the proximity probe or the velocity pickup.

• A wide dynamic range (160 dB) and good linearity means they can

detect very small vibrations and not be damaged by very large vibrations

in a wide frequency range.

• The accelerometer is the only transducer that can accurately detect

pressure wave vibrations in high frequencies.

Other transducers that are used today for vibration measurement are the

strain gauge, piezo-resistive accelerometer, piezo-ﬁ lm/piezo-patch, etc.

3.3.2 Selection of appropriate accelerometers

The ﬁ rst decision to make is whether to use a high-impedance or a low-

impedance accelerometer. The advantages of the high-impedance acceler-

ometer are:

Copyrighted Material downloaded from Woodhead Publishing Online

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

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:01:01 AM

IP Address: 129.132.208.2

40 Vehicle noise and vibration reﬁ nement

• Wide dynamic range

• Ability to use long connector cables between the accelerometer and the

charge ampliﬁ er.

The main disadvantage of high-impedance accelerometers is their cost.

The advantages of low-impedance accelerometers are:

• Low cost

• Simpler signal conditioning.

The main disadvantages of low-impedance accelerometers are:

• Size and mass (once the electronics are built in)

• Intolerance to being overloaded

• A rather high limit on the lowest frequency at which they may be used.

Subsequent choice of accelerometer will depend on the following factors.

• Sensitivity: The larger transducers have greater sensitivity. Higher-

amplitude motions require a low-sensitivity accelerometer. A low-level

motion should be measured by a high-sensitivity accelerometer.

• Frequency range: The upper limit is determined by the natural fre-

quency of the piezoelectric elements. Generally, smaller transducers

may be used to measure higher frequency vibrations. The lower-

frequency limit is determined by the time constant of the signal condi-

tioning. Only the charge-ampliﬁ ed accelerometer is capable of measuring

down to (almost) 0 Hz.

• Physical size: As a general rule the mass of the accelerometer should

be less than 10% of the mass of the vibrating structure to which it will

be attached. Otherwise the additional mass of the attached accelerom-

eter will alter the vibration characteristics of the structure. This is often

termed the mass loading effect.

– Larger accelerometers have lower natural frequencies and are more

sensitive but cannot be used for measuring high-frequency vibration

or on lightweight panels due to the mass loading imposed by them.

– Smaller accelerometers have higher natural frequencies and will not

mass load lighter panels.

• Mounting accessories: The preferred method of mounting the acceler-

ometer for high-frequency measurements is with a screw-threaded stud.

However, various hard epoxy or cyanoacrylate adhesives are also avail-

able as alternatives. Other mounting methods include magnetic mounts,

beeswax, plasticine and double-sided adhesive tape for lower frequen-

cies (below 2 kHz). All of these, although convenient, will restrict the

upper frequency limit of the measurement. Some accelerometers are

made with their cable terminations on top of the casing to facilitate stud

mounting. The mass of the mounting adds to the mass loading effect.

Copyrighted Material downloaded from Woodhead Publishing Online

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

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:01:01 AM

IP Address: 129.132.208.2

Vehicle vibration measurement and analysis 41

• Cables: Low-impedance accelerometers with built-in ampliﬁ cation can

be used directly even with long lengths of coaxial cable before the fre-

quency response is affected. High-impedance transducers require special

low-noise/high-insulation-resistance cables known as microdot cables,

which are of limited length and also fragile. The cables impose limita-

tions on:

– Temperature exposure

– Linearity

– Transverse sensitivity

– Damping

– Strain sensitivity.

3.3.3 Charge ampliﬁ ers

A good-quality charge ampliﬁ er (such as the Brüel & Kjær Type 2626

shown in Fig. 3.4) will have the following features:

3.4 A typical charge-type ampliﬁ er.

Copyrighted Material downloaded from Woodhead Publishing Online

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

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:01:01 AM

IP Address: 129.132.208.2

42 Vehicle noise and vibration reﬁ nement

• A charge input, with a microdot cable termination on the front or rear

panel

• A facility for the user to enter the transducer sensitivity to three signiﬁ -

cant ﬁ gures (commonly as pC per m/s

)

• A variable ampliﬁ er to allow input signals of amplitude between 0.1 mV

per m/s

and 1 V per m/s

• An indicator that warns the user when the ampliﬁ er is being

overloaded

• A low-frequency limit to suppress low-frequency noise (particularly if

signal integration is to be used)

• Integration circuits to allow the direct measurement of velocity or

displacement

• A low-pass ﬁ lter to remove unwanted signal components prior to

ampliﬁ cation

• Battery and external supply via a mains transformer.

Modern intelligent data acquisition and analysis systems (such as Brüel

& Kjær Pulse) integrate charge ampliﬁ er or voltage power supply units into

the data recording instrumentation in which a transducer database can be

established and maintained by a Brüel & Kjær default transducer database

or other key-in transducer data selection and a factory calibration chart and

details of transducer sensitivity can be downloaded.

3.3.4 Calibration of accelerometers

For a good-quality charge ampliﬁ er the factory calibration chart and sensi-

tivity data can be used along with charge ampliﬁ er gain to calculate a cali-

bration factor (V/m s

−2

In order to get conﬁ dence on the measured data, calibration must be

conducted before measurements are made. A calibrator exciter can be used

to calibrate the entire signal chain from the accelerometer to the display.

By comparing a known exciter vibration (B&K 4294) level of 10 m s

−2

rms

(±3%) at 159 Hz with the voltage measured by the transducer which is

mounted on the top part of the exciter, the calibration factor is determined

(V/m s

−2

With high-impedance accelerometers and good-quality charge ampliﬁ ca-

tion, the calibration factor is independent of the (reasonable) length of the

microdot lead. Therefore, subsequent to initial calibration, a replacement

microdot lead (perhaps of longer length) does not need recalibration of the

transducer chain.

For the intelligent data acquisition and analysis systems (such as B&K

Pulse), a transducer can be directly connected to the front end without any

charge ampliﬁ er or power supply in between.

Copyrighted Material downloaded from Woodhead Publishing Online

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

ETH Zuerich (307-97-768)

Sunday, August 28, 2011 12:01:01 AM

IP Address: 129.132.208.2