Liquid piston engines by Gupta, Aman Narayan, Sunny Sharma, Shubham

144 Liquid Piston Engines

as bouncing motion dominates piston dynamics. The duration of sliding
motion of piston along liner was observed to increase with increase in
load and speed conditions which is in agreement with previous available
literature.

This work discusses a two dimensional model of piston secondary
motion. Various dynamic parameters of system were calculated using con-
cept of mobility. These parameters were used to simulate piston secondary
motion and hence validate the model. Effects of load and speed were also
investigated on piston secondary motion. Sliding motion of piston along
liner was observed to increase with increase in load and speed which is in
agreement with previous data available.

Liquid Piston Engines. Aman Gupta, Shubham Sharma, and Sunny Narayan.
© 2017 Scrivener Publishing LLC. Published 2017 by John Wiley & Sons, Inc.

5

NVH Features of Engines

5.1 Background

Diesel engines constitute a major source of power for various ships, buses,
trains as well as road machinery. About one-fifth of total energy consump-
tion in U.S.A. goes towards operating these engines [1] and demand for
these engines is fast growing compared with gasoline engine [2]. Sales
of diesel based engines reached a peak during the 1980s in U.S.A. due to
major oil crises as shown in Figure 5.1 [1].

Various projections at that time had predicted that an increase of about
20% in sales would be achieved at the end of decade [3]. But variations in
the costs of diesel, falling petrol prices and problems in operations of diesel
engines led to a fall in their sales [4, 5].

Petrol engines use spark ignition for initiation of fuel reactions as com-
pared with diesel engines which are based on compression ignition of fuel-
air mixture. Diesel engines operate at higher compression ratio as compared
with petrol engines allowing more useful work in cycle. Combustion in these
engines can be made to occur away from walls, thus helping in overall heat
reduction. In addition there are throttling as well as various pumping losses

145

146 Liquid Piston Engines

US sales of diesel vehicles
500,000

400,000

300,000

200,000 Autos
100,000

0 Light trucks

1978
1980
1982
1984
1986
1988
1990
1992
1994
1996

Figure 5.1 Sales trend of diesel engine based automobiles in U.S.A.

in petrol engines, which is the reason for their lesser efficiency as compared
with diesel engines. Overall fuel efficiency of a diesel engine may be over
40% in medium engines and over 50% for larger engines used in marine
propulsions [6]. These factors have renewed interest of automobile compa-
nies towards diesel engines. Data about diesel engine automotive sales in
Europe have indicated that about a quarter of new automobiles are pow-
ered by diesel engines [7, 8]. In France, diesel engines accounted for about
half of automotive engine sales [9]. In Japan, the number of diesel engine
car sales have tripled in the past [10]. Several commercial vehicle suppliers
have started to manufacture their own Diesel Engines for installation in
engines. Table 1.1 shows the representation of data about U.S. market share
of various diesel engines supplied by automotive manufacturers.

Recently several modern technologies like common rail direct injection
system (CRDI), exhaust gas recirculation (EGR) and turbo charging are
examples of some of modern technologies being introduced for develop-
ment of diesel engines in modern times [12]. Other methods used include
pre-mixed charge compression ignition (PCCI) and homogenous charge
compression ignitions (HCCI) systems [13–15]. However higher period
of pre-mixed combustion in these technologies leads to higher noise emis-
sions. Hence various merits of using a diesel engine may be lost over noise,
vibration and harness performance benchmarks.

5.2 Acoustics Overview of Internal
Combustion Engine

Vehicle noise and vibrations can have a bad effect on overall perfor-
mance of automobiles. These aspects also form important benchmarks for

NVH Features of Engines 147

Table 5.1 Supply of diesel engines by various manufacturer, Year-2013.

Automotive make Engine make Market share
Hino
Freightliner Hino 100%
Cummins 62.3%
International Detroit Diesel 37.0%
Volvo Mercedes Benz 0.7%
Western Star Cummins 7.2%
Mack Navistar 92.8%
Peterbilt Cummins 13.6%
Volvo 86.4%
Cummins 21.2%
Detroit Diesel 78.8%
Cummins 6.0%
Mack 94.0%
Cummins 65.2%
PACCAR 34.8%

Mount Engine Exhaust system
Driveline
Intake Clutch
system
Transmission

Powerplant

Figure 5.2 Power train system.

customers perception of vehicle performance. They are measures of com-
fort levels and vehicle reliability. In automotive terms noise, vibration and
harness (NVH) is used to denote the unwanted sound and vibrations in
an automobile [16]. NVH is a term used for branch of engineering related
to vehicle refinement in terms of sound and vibration performance expe-
rienced by occupants. A typical vehicle system consists of several systems
which include chassis system, power train system, HVAC system and
electronics system [17].

148 Liquid Piston Engines

Vibrations Vehicle vibrations Squeak and rattle
and sound

Noise

Engine Tire Body Miscell-
stiffness aneous

Mount Drive train Vehicle Component Interior Exterior Passenger
medes resonance noise noise noise

Engine Tire road Wind Miscellaneous
Suspension noise

Engine Intake/ Body structure
system exhaust borne

Body
air borne

Figure 5.3 Noise and vibration sources in engine.

A schematic arrangement of a typical powertrain system is shown in
Figure 5.3 [18]. The powertrain system includes engine, transmission sys-
tem, clutch, driving system, intake and exhaust systems.

The engine system is a major source of vibrations which may be clas-
sified as external or internal one. The internal sources are due to variable
pressure on piston head and inertia of moving parts. The external sources
refer to vibrations due to unbalanced moments and variable engine torque
which results in vibrations of whole engine block. The noise sources in
engine consists of mechanical, combustion and aerodynamic noise [19].
Mechanical noise which is proportional to engine speed is due to inertial
effects of relative motion of parts under air pressure or inertial force that
results in impact noise and vibrations.

This noise includes noise due to piston motion, bearing noise, cam noise,
oil pump noise, timing belt and chain noise as well as structural noise of
cover [20]. Aerodynamic noise includes intake noise, exhaust noise and fan
motion noise. Combustion noise is generated due to impulsive pressure
wave due to combustion process resulting in vibrations of engine block due
to impacts on cylinder wall and head [21]. The vibrations due to transmis-
sions and driveline also contribute towards powertrain noise levels. There
are also other noise sources due to squeak and rattle of engine body sys-
tem. The noise experienced by the passengers inside the vehicle not only
depends upon various sources but also upon body structure and acoustic
transfer function. Various noise and vibration sources have frequency
range values. Wind and road tire noise lies in medium frequency ranges.

NVH Features of Engines 149

During the decade of 1970s, increasing attention towards various noise
control led to more attention being paid towards the acoustic performance
of diesel engines. Priede analyzed the relationship between in cylinder pres-
sure development and radiated noise from engine [21]. Kamal has done
finite element analysis of individual engine structure for dynamic analysis
of engines [22]. Later on the basis of noise transfer paths, main bearing of
connecting rod as a key design factor for controlling the indirect combus-
tion noise [23]. In modern days multidisciplinary analysis approaches are
being utilized for NVH performance evaluation of engines. Some of these
include modal analysis, Finite element analysis (FEA), Boundary element
method (BEM), Statistical energy analysis (SEA), lumped mass approach
and transfer path analysis (TPA).

Evaluation of noise performance of engines needs to be carried out both
objectively as well as subjectively [24]. Each of these methods have specific
frequency range over which it is most suited to. e.g. FEA is more suited low
frequency ranges, TPA analysis is more suited in medium frequency ranges
and SEA is suited for high ranges.

5.3 Imperial Formulation to Determine
Noise Emitted from Engine

Figure 5.4 shows plots of in cylinder pressure spectra for two types of
engines [25]. A difference of about 20 dB is seen at 1 KHz frequency.

Based on mathematical relationships, Anderton has developed math-
ematical models to quantify combustion noise according to type of engine

210

Sound pressure level dB 200

190 12.2 litre

180 D.I. diesel
130 mm. bore

170

1.6 litre

160 gasoline

91 mm. bore

150

140

130
100 1000
Frequency-Hz

Figure 5.4 In cylinder pressure spectrum.

150 Liquid Piston Engines

[26]. It involves the concept of mechanical impedance Z(f) between force
applied at top of piston F(f) and average mean square root velocity v(f) of
engine block. i.e.

V 2(f ) (5.1)
Z(f )

F(f )

The average surface velocity v(f) may be expressed in terms of in cylin-
der pressure (p) and cylinder bore (B) as:

V(f ) p B2 (5.2)
4Z( f )

Further the radiated acoustic power (W) from a surface may be expressed
in terms of radiation efficiency ( ) and radiated surface area (S) in form of:

W( f ) CSV( f ) p2 (5.3)
C

Combining the two relationships we have :

W SC p B2 (5.4)
4Z( f )

The intensity of radiated noise I(f) can be written as:

I( f ) C p B2 (5.5)
4Z( f )

In order to minimize the effects of speed on Anderton analyzed various
in cylinder pressure spectra operating various engines under different con-
ditions and found that the trends of plots were almost a straight line in
frequency ranges 0.8 KHz–3 KHz range. The slope of pressure spectrum in
this range was defined as combustion noise index. Using further analysis it
was shown that in cylinder pressure (p) may be expressed as :

p2( f ) ~ N z (5.6)

f Antilog (3N)

NVH Features of Engines 151

where N is engine RPM
From the above relationships we have:

I( f )~ Nz Antilog (3N) cS p B2 (5.7)

f 4Z( f )

I( f ) N z B4 (5.8)
f z( f )2
S

Overall Intensity IO can be expressed by integration over given fre-
quency range (f1, f2) values as:

IO SN z B4 f2 f z Z( f )2 (5.9)
f1

Various empirical relationships have been developed at ISVR, University
of Southampton for prediction of noise in terms of sound pressure levels
for different types of engines. Some of these include :

SPL = 30N.A. Direct Injection Diesel engines log(n) + 50 log(B)+106 (5.10)

SPL = 40Turbocharged Diesel engines log(n) + 50 log(B)–135 (5.11)

SPL = 43Indirect injection Diesel engines log(n) +60 log(B)–176 (5.12)

SPLPetrol engines = 50 log(n) + 60 log(B)-203 (5.13)

As compared with diesel engines, a gasoline engine has higher opera-
tional speeds, smaller bore and smaller reciprocating mass. Consequently
a gasoline engine has lower in cylinder pressure and hence lower sound
pressure levels of radiated noise as compared with diesel engines as seen
from Figure 5.5.

5.4 Engine Noise Sources

Typical noise sources in a combustion engine are plotted in Figure 5.6 [24].
Combustion noise from engine depends upon the speed of combustion pro-
cess taking place in combustion chamber, 50% mass fraction burnt (CA50),

152 Liquid Piston Engines

1m noise level (dBA) 110
100 Diesel engines

90
Petrol engines

80

70

60 2000 3000 4000 5000 6000
1000 RPM

Figure 5.5 Variations of sound pressure levels with engine speed.

7
18
29
3
4 10

5

6
11

Figure 5.6 Schematic representation of various sources of noise (1: Valve train, 2: Chain
drive, 3–4: Accessory noise, 5: Piston slap, 6: Bearing noise, 7: Cover noise, 8: Intake noise,
9: Exhaust noise, 10: Combustion noise, 11: Oil pan noise).

peak in cylinder pressure developed and its position in crank angle domain
as well as pressure derivatives with respect to crank angle. The intensity of
combustion noise is proportional to square of cylinder pressure and it also
depends upon engine speed, load and injection delay period. This noise
can be further classified as direct combustion noise and indirect combus-
tion noise [27]. Direct combustion noise is directly radiated from engine
structure and its transfer function can be calculated by doing an explosion
in combustion chamber keeping other engine parts stationary so mechani-
cal noise does not interfere with radiated noise. Indirect combustion noise
in engines is portion of noise that is transferred to structure from combus-
tion chamber. Mechanical noise which is proportional to engine speed is
due to piston motion, bearing operations, timing belt operation, pump and
valve operations. This type of noise can be estimated by running engine
under motored condition. Flow noise depends upon turbulence, pressure
flow and friction during flow. The effect of tailpipe and radiations of muf-
fler are primary source of exhaust noise.

NVH Features of Engines 153

Table 5.2 Frequency ranges of various noise sources.

Noise source Approximate Effecting factor
frequency range
Combustion Noise In cylinder pressure
Piston Slap 500–8000 Hz Speed, Piston Design
Valve Operation 2000–8000 Hz Valve Type, Engine speed
Fan Noise 500–2000 Hz Speed, Number of Blades
Intake Flow Noise 200–2000 Hz Turbulence
Exhaust Flow Noise 50–5000 Hz Turbulence
Injection Pump 50–5000 Hz Pump features

operation 2000 Hz
Gear Noise
Accessory Belt-Chain 4000 Hz Speed, Number of teeth
3000 Hz
Noise Engine speed, Misalignment,
Number of teeth

Table 5.3 Noise analysis from a V6 engine.

Part dB sound pressure levels

Engine Block 78.7
Cylinder Head 76
Crank Case 79
Engine Base 78
Intake manifold 77
Cam Cover 78
Front Cover 77
Exhaust manifold 74
Oil pan 73

The frequency ranges of various noise sources are enlisted in Table 1.1.
The range of frequency not only depends upon engine load and speed but
also on configuration of engine. Hence identification and estimation of
specific frequency must be done by testing. By comparison of fundamental
frequency and harmonics of individual noise sources, contribution of each
source can be estimated as seen from Table 5.3 for a V6 engine tested in an
anechoic chamber.

154 Liquid Piston Engines

5.5 Noise Source Identification Techniques

There are several techniques that can be used to identify various sources of
noise in internal combustion engines [28]. Some of these include shielding
techniques, surface vibration method and acoustic intensity technique. Of
these methods the lead covering method is the most expensive one as well
as the time consuming. Other methods are time consuming and need a lot
of calculations. These techniques are discussed further in the coming part
of this work.

a. Lead covering method: it is one of the most reliable methods of acous-
tic source identification for engines. This method consists of noise emission
measurement from engine using selective covering of parts of engine with
high transmission loss material which is usually lead. The noise increase
is then noted by removing lead cover from component. The procedure is
repeated for one by one for all components. Figure 5.7 shows results of
such a test done on a 6 cylinder naturally aspirated diesel engine [29].

Total sound power level of this engine was found to be 114 dBA with
valve cover, muffler, front gear cover and oil pan cover contributing about
21%, 10%, 8% and 7% respectively.

b. Surface vibration method: The A weighted sound power level of engine
(Lw[A]) is given in terms of acoustic impedance ( c), surface velocity (u),
radiation efficiency ( ) and surface area (S) by [29]:

Lw[A] = 10 log( c) + 10 log(S) + 10 log(σ) + 10 log (u)
(5.14)

1 m SPL dBA 106 SPL after uncovering
104 None
102 Above Front Left Front gear
100 cover
Valve cover
98 Oil pan
96
94 Muffler
92
90 Right

Average

Micro phone position

Figure 5.7 Noise analysis using lead cover method.

NVH Features of Engines 155

Sound pressure level (dB SPL) 130 (estimated)
120 100 phon
110 80
100 60
40
90 20
80 (threshold)
70
60 100 1000 10k 100k
50
40
30
20
10

0
–10

10

Figure 5.8 Equal loudness contours (grey) (from ISO 226:2003 revision) original ISO
standard shown (dark grey) for 40 phons.

Radiation efficiency is the ability of surface vibrations to convert into air
borne noise. This is also related to critical frequency of component which
may be defined as frequency at which wavelength of vibrations in structure
matches with wavelength of radiated vibrations. At frequencies lower than
critical one, the radiation efficiency is less than unity and vice versa. The
dominant range of critical frequency for components of a diesel engine lies
in range 400–800 Hz. The radiation efficiency can be estimated by con-
sidering engine as a radiating rigid sphere. This efficiency rises at the rate
of 40 d B/decade at frequencies lesser than critical frequency. The value of
critical frequency occurs when kr ≈ 4, where k is wave number of sound
waves and r is radius of imaginary sphere that has same volume as that of
engine.

The measurement of surface vibrations is carried out best by use of accel-
erometers mounted on the engine surface. Positioning of accelerometers
must be carefully done, as surface vibrations vary with wall thickness and
proper balancing between less and strong sensitive measurement points is
necessary. The surface velocity can be calculated by first converting accel-
eration data into frequency domain using Fast Fourier Transformation
(FFT) and then carrying out integration.

Figure 5.8 shows the results of experiments of various components
towards Sound pressure level (SPL) obtained by surface velocity method
[28]. It can be seen that contributions of valve cover, muffler shell, gear
cover and oil pan cover is 29%, 20%, 4% and 15% respectively.

156 Liquid Piston Engines

112 1 Valve cover
2 Muffler
Sound power level Lw/dBA 3 Front gear cover
110 4 Oil pan

108 5 Others

106

104 1 23 5
102
100 4

0 20 40 60 80 100

Ratio to total acoustic power/%

Figure 5.9 Noise analysis using vibrational analysis method.

P HC M
+G

W C~ – E

Figure 5.10 Engine noise model. Combustion noise
Cylinder pressure

Wiener
filter

Noise signal Residual noise

Figure 5.11 Application of wiener filter for estimation of combustion noise.

c. Use of spectro filters [30]: Diesel engines produce a complex level of
noise levels of which combustion based noise is of major interest. The resid-
ual noise produced by various other sources is known as mechanical noise.
Once in cylinder pressure signal is known these two sources can be sepa-
rated using Wiener spectro-filters. These filters extract noise sources that
are correlated with in cylinder pressure signals hence providing an estima-
tion of combustion noise.

The spectro-filter also called Wiener filter is a single input single output
system whose impulse response is denoted by H(t). The input and output of
system is denoted by P(t) and C(t) respectively corrupted by component
M(t). The model is described by following equations:

C(t) = P(t) H(t) (5.15)

NVH Features of Engines 157

P1 H1 C1 M
P2 H2 C2
D
Figure 5.12 Dual cylinder engine noise model.
(5.16)
G(t) = M(t) + C(t)

The spectro-filter H(t) can be estimated from following equations:

W( f ) sPD ( f ) (5.17)
sPP ( f )

W(t) = FT 1[W(f)] (5.18)

In these equations SPP(f) denotes auto spectrum of P(t) whereas SPD(f)
denotes cross spectrum of P(t) and D(t). Convolution of input P(t) with
W(t) gives an estimate of C(t). i.e.

C(t) = P(t) W(t) (5.19)

M(t) = D(t) C(t) (5.20)

In case of a mono cylinder engine C(t) denotes combustion noise, P(t)
denotes in cylinder pressure, M(t) denotes mechanical noise, D(t) denotes
total noise emissions and H(t) denotes transfer function between in
cylinder pressure and noise emissions.

In case of dual cylinder engine, Figure 5.12 depicts the corresponding
engine noise model, which is a multiple input–single output (MISO) sys-
tem. The combustion noise C(t) is now the sum of the two combustion
noises produced by each cylinder. i.e.

C(t) = C1(t) + C2(t) (5.21)

5.6 Summary

According to an estimate in 2006, the automotive industry had turnover
of about 1 Trillion U.S. $ per year with an annual growth of about 6%[24].

158 Liquid Piston Engines

Attributes such as durability and serviceability require a vehicle to be in
service for a certain period of time. Costs of vibration and noise control are
usually very high. e.g. the costs of warranty for brake was about 1 Billion
U.S.$. per year during year 2005. There are a wide range of non-intrusive
methods that can be used to evaluate NVH performance of combustion
engines. In this work results from various experiments done on a water
cooled diesel engine have been discussed which can be used to device
strategies for control noise emissions from engine as well as its condition
monitoring.

References and Bibliography

1. Davis, S.C., Transportation Energy Data Book: Edition 18, Report No. ORNL-
6941, 1998.

2. De Cicco, J., and Mark, J., “Meeting the Energy and Climate Challenge for
Transportation in the United States,” Energy Policy, Great Britain: Elsevier.
Vol. 26, No. 5, pp. 395–412, 1998.

3. John, A., Department of Transportation (DOT) Briefing Book on the
United States Motor Vehicle Industry and Market, Version 1, Volpe National
Transportation Systems Center, Cambridge, 1991.

4. Sperling, D., New Transportation Fuels: A Strategic Approach to Technological
Change. Berkeley, Calif.:University of California Press, 1988.

5. Cronk, S., Building the E-Motive Industry: Essays and Conversations about
Strategies for Creating and Electric Vehicle Industry. Warrendale, Penn.:
Society of Automotive Engineers, 1995.

6. Evangelo, Rakopoulos, “Experimental study of combustion noise radiation
during transient turbocharged diesel engine operation”, Energy, Volume 36,
Issue 8, pp. 495–499, 2011.

7. Walsh, Michael, P., “Global Trends in Diesel Emissions Control – A 1997
Update”, SAE Paper No. 970179, 1997.

8. Krieger, R.B., Stewart, R., Pinson, K., Gallopoulos, N., Hilden, Monroe, D.,
Rask, Solomon and Zima, “Diesel Engines: One Option to Power Future
Personal Transportation Vehicles,” Proceedings of the Diesel Engine Emissions
Reduction Workshop, La Jolla, July 28–31, Washington, 1997.

9. Wang, M., Stork, K., Vyas, A., Mintz, M., Singh, and Johnson, “Assessment of
PNGV Fuels Infrastructure, Phase 1 Report: Additional Capital Needs and
Fuel-Cycle Energy and Emissions Impacts’’, 1997.

10. Walsh, Michael, P., “Global Trends in Diesel Emissions Control – A 1998
Update.” SAE paper No. 980186, 1998.

11. Sonya, G., Labelle, A., Special report TD Economics, U.S. auto sales basking in
their comeback glow, 2014.

NVH Features of Engines 159

12. Kondo, M., Kimura, S., Ηirano, I., Uraki, Y., Maeda, R., ‘’Development of
noise reduction technologies for a small direct-injection diesel engine’’, JSAE
Review, Vol 21, pp. 327–33, 2001.

13. Torregrosa, AJ., Broatch, A., Novella, R., Monico, LF., “Suitability analysis
of advanced diesel combustion concepts for emissions and noise control’’,
Energy, Vol 36, pp. 825–38, 2011.

14. Shi, Y., Qiao, X., Ni, J., Zheng, Y., Ye, N., “Study on the combustion and emis-
sion characteristics of a diesel engine with multi-injection modes based on
experimental investigation and computational fluid dynamics modelling’’,
Proceedings of intuition of Mechanical Engineers, Journal of Automobile
Engineering, Vol 224, pp. 1161–76, 2010.

15. Mohamad, S., Qatu, Mohamed, Abdelhamid, K., Pang, J., Sheng, G., “Overview
of automotive noise and vibration”, Int. J. of Vehicle Noise and Vibration, Vol.
5, No. 1, pp. 1–35, 2009.

16. Genuit, K., “The sound quality of vehicle interior noise-a challenge for NVH
engineers’’, Int. J. of Vehicle Noise and Vibration, Vol. 1, pp. 58–68, 2004.

17. Warring, R.H., Handbook of noise and vibration control, Trade and Technical
press, Modern, Surry, U.K, 1985.

18. Anderton, D., Baker, J., “Influence of operating cycle on noise of diesel
engines’’, SAE paper no. 730241, 1973.

19. Carlucci, P., Ficarrela, A., Laforgia, D., “Study of the influence of the injec-
tion parameters on combustion noise in a common-rail diesel engine using
ANOVA and neural networks’’ SAE paper no. 2001-01-2011, 2001.

20. Pischinger, F., Schmillen, K., Leipold, FW., “A new measuring method for
the direct determination of diesel engine combustion noise’’SAE paper no.
790267, 1979.

21. Priede, T., “Problems and developments in automotive engine noise research”,
SAE paper no 7902, 1979.

22. Hickling, R., Kamal, M., Engine Noise – Excitation, Vibration and Radiation,
New York, London – Plenum Press, 1982.

23. D’Anna, T., Govindswamy, K., “Aspects of Shift Quality With Emphasis on
Powertrain Integration and Vehicle Sensitivity”, SAE N&V Conference 2005,
Transverse City, MI, 2005. SAE-Paper 2005-01-2303, 2005.

24. Sheng, G., Vehicle Noise Sound Vibration and Sound Quality, SAE interna-
tional, 2012.

25. Anderton, D., Noise source identification techniques, ISVR course notes,
2003.

26. Anderton, D., “Relation between Combustion System ad Noise’’, SAE paper
no. 790270, 1979.

27. Russell, M., Haworth, R., “Combustion noise from high speed direct injection
diesel engines’’, SAE paper no. 850973, 1985.

28. Yuehui, Liu., “Engine noise source identification with different methods”,
Transactions of Tianjin University, Vol 8, issue 3, pp. 174–177, 2002.

160 Liquid Piston Engines

29. Grover., Lalor, “A review of low noise diesel engine design at I.S.V.R.’’, Journal
of Sound and Vibration, Volume 28, Issue 3, 8 June 1973, Pages 403–428,
1973.

30. Pruvost, L., Leclere, Q., Parizet, E., “Diesel engine combustion and mechani-
cal noise operation using an improved specto filters’’, Mechanical Systems and
signal processing 23, pp. 2072–2087, 2009.

31. http://www.fev.com/fileadmin/user_upload/Media/Spectrum/en/spec-
trum21.pdf

32. http://luis.lemoyne.free.fr/Indicating_Product%20Overview_2009.pdf.

Liquid Piston Engines. Aman Gupta, Shubham Sharma, and Sunny Narayan.
© 2017 Scrivener Publishing LLC. Published 2017 by John Wiley & Sons, Inc.

6

Diagnosis Methodology
for Diesel Engines

6.1 Introduction

Various events occur in combustion engines cycle that can lead to increased
emissions, fuel consumptions and potential damage to engine. A fuzzy
based pattern recognition has been used for monitoring of various injection
events [1]. Such diagnosis includes analysis of injection pressures and pat-
terns using pressure transducers [2]. Misfires in engines have been analyzed
using variations in engine angular speed [3]. Rizzione [4] and Connolly [5]
have proposed a torque based algorithm for detection of misfire. Various
energy based models have been studied by Tinuat et al [6]. A wavelet based
method has been proposed for localization of engine misfires [7].

Engine knock caused due to spontaneous ignition of mixture during
combustion leading to chamber resonance is another major event that
needs proper diagnosis in diesel engines. A joint time frequency method
has been studied in [8] to detect knock process. Suitable knocking index has
been defined using band pass filtered cylinder pressure signals [9]. Wavelet
analysis [10] and Fourier analysis [11] has also been used for detection

161

162 Liquid Piston Engines

of knock. Features of various events can be extracted from vibration sig-
nals in time domain [12], frequency-domain [13] and time-frequency
domain [14]. The frequency-domain analysis is generally considered the
most adequate signal-processing tool for the non-stationary diesel engine
vibrations. In this part of work some signal processing methods of diagno-
sis have been analyzed with focus on power spectral density function and
time-frequency function analysis.

6.2 Power Spectral Density Function

Power density function (PSD) of a random process provides the frequency
composition of data in terms of spectral density of its mean square value
[15]. The mean square value of a time sample in frequency range , +
can be obtained by passing sample through a band pass filter with sharp
cutoff frequency features and computing the average of squared output
from filter. The average square value will approach a mean square value as
T . i.e.

T 2 (t )dt

2( , ) Lim x x
0
(6.1)
T

6.3 Time Frequency Analysis

Fourier transformation of a function f(t) is given by:

t (6.2)

f ( ) f (t)e j t

0

This analysis is useful as long as frequency content of signals do not
vary with time. Hence time-frequency analysis or wavelet analysis is more
suitable for analysis of noise and vibration signals emitted from engine
[16]. Time frequency analysis is suitable for noise component having slow
frequency changes such as those generated during engine ramp down
whereas wavelet analysis is more suited for those signals having fast fre-
quency changes such as those generated during rattle [17]. In the time
frequency analysis the signal is windowed into small intervals and then
Fourier transformation is taken for each interval [18]. Length of window
can be used to change the resolution of output. A shorter window has

Diagnosis Methodology for Diesel Engines 163

high time resolution, but poor frequency resolution and vice versa. High
time resolution at higher frequencies of wavelet transformation makes
it possible to resolve short consecutive events. The short time frequency
analysis is based on expansion of signal into a set of weighted frequency
modulated Gaussian functions. It is given by:

T )e j tdt (6.3)

STFT( , f ) x(t)h * (t

0

where x(t) is input signal & h(t ) is window function.
A Wigner Ville function has following quadratic time frequency distri-

bution given as [19]:

T

STFT( , f ) x t x * t e j tdt (6.4)
02 2

6.4 Wavelet Analysis

Wavelet analysis map a signal on time frequency plane and are sensitive
towards transient signals. One major drawback of time frequency signals
processing methods is that they produce ripples hence making it difficult
extract valuable information [20]. In wavelet analysis frequency resolution
is better at low frequencies and time resolution is better at higher frequen-
cies. Hence Wavelet analysis results are more accurate [21]. For wavelet
transforms the signal is projected onto a family of zero mean functions
known as wavelets. These have high time resolution and have no cross-
term interference. The power spectral Similar to the short-time-frequency
analysis, wavelet transform is a linear time-frequency transformation.

The squared wavelet transform is called a scalogram. A single scalogram
can easily cover audible frequency range with a time resolution of approxi-
mately 0.1 ms for the high-frequency components [22]. This makes the sca-
logram suitable for such various signals like squeak and rattle noise in an
automobile interior, for which a wide range of frequency analysis is needed.

Mathematically a complex wavelet transform is defined for a function
f(t) as [23]:

CWT(a, b) f (t) 1 (t b) dx (6.5)
aa

164 Liquid Piston Engines

where
ψ(t): Mother wavelet
f(t): Analyzed signal
a: Scaling factor
b: Shifting factor
CWT(a, b): wavelet coefficients

Mother Wavelet function (t) must satisfy following conditions:

a. This function has zero average and decays exponentially to
zero. i.e.

(t)dt 0 (6.6)

b. This function and its Fourier transformation must satisfy
admissibility condition. i.e.

ˆ(t )2 (6.7)
0

|f |

Both dilation as well as translation parameters in CWT are
subjected to variations which makes the use of this meth-
odology more complex. Discretization of signals helps to
reduce this problem to certain extent. The CWT of a signal
discrete signal Xm is defined in terms of sampling time Δt
and sample data points m, n as:

CWT N1 * (m n) t (6.8)
Xj
Xm

m0

where t = m Δt, b = n Δt, m & n varies from 0, 1, 2 …
N 1, N

6.5 Conclusion

This part of the work investigated the effect of changing injection param-
eters on noise emissions and combustion pressure. Noise emitted from
engine depends on the quantity of fuel injected inside cylinder. Various

Diagnosis Methodology for Diesel Engines 165

ranges of noise signals sources were identified. Time–Frequency analysis
showed the onset of various events of engine cycle. Based on the identifica-
tion of various frequency bands it is possible to filter the signals in order to
extract more information about combustion and mechanical based noise
events for detailed analysis which is discussed in a later part of this work.

References and Bibliography

1. Sharkey, A. J. C., Chandroth, O., and Sharkey, N., “A multi-net system for
the fault diagnosis of a diesel engine”, Neural Computing & Applications, 9(2),
pp. 152–160, 2000.

2. Payri, F., et al., “Injection diagnosis through common-rail pressure mea-
surement”. Proceedings of the Institution of Mechanical Engineers, Part D:
Journal of Automobile Engineering, 220(3), pp. 347–357, 2000.

3. Azzoni, P. M., et al., 1996, “Misfire Detection in a High-Performance Engine
by the Principal Component Analysis Approach”, in SAE International
Congress & Exposition. Detroit, MI, 1996.

4. Rizzoni, G., “Estimate of indicated torque from crankshaft speed fluctuations:
A model for the dynamics of the IC engine”. IEEE transactions on vehicular
technology, 38(3), pp. 168–179, 1989.

5. Connolly, F. T., and Rizzoni, G., “Real time estimation of engine torque for the
detection of engine misfires”, Journal of Dynamic Systems, Measurement, and
Control, 116, pp. 675, 1994.

6. Tinaut, F. V., et al., “Misfire and compression fault detection through the energy
model”, Mechanical Systems and Signal Processing, 21(3), pp. 1521–1535, 2007.

7. Chang, J., Kim, K., and Min, K., “Detection of misfire and knock in spark
ignition engines by wavelet transform of engine block vibration signals”,
Measurement Science and Technology, Vol 13, pp. 1108, 2002.

8. Samimy, B., and G. Rizzoni, “Mechanical signature analysis using time-fre-
quency signal processing: application to internal combustion engine knock
detection”, Proceedings of the IEEE, 84(9), pp. 1330–1343, 1996.

9. Borg, J., “Wavelet-based knock detection with fuzzy logic”, IEEE International
Conference on Computational Intelligence for Measurement Systems and
Applications. Giardini Naxos, Italy, pp. 26–31, 2005.

10. Zhang, Z., and Tomita, E., “Knocking detection using wavelet instantaneous
correlation method”, Japan Society of Automotive Engineers (JSAE) review,
Vol 23, pp. 443–449, 2002.

11. Lee, J., “A new knock-detection method using cylinder pressure, block vibra-
tion and sound pressure signals from an SI engine”, SAE International Fuels &
Lubricants Meeting & Exposition, M. I, 1998.

12. Ftoutou, E., Chouchane, M., and Besbès, N., “Internal combustion engine
valve clearance fault classification using multivariate analysis of variance and
discriminate analysis”, T. I. Meas. Control, 2011.

166 Liquid Piston Engines

13. Carlucci, A. P., Chiara, F., and Laforgia, D., “Analysis of the relation between
injection parameter variation and block vibration of an internal combustion
diesel engine“, Journal of Sound and Vibrations, Vol 295, pp. 141–164, 2006.

14. Wang, C., Zhong, Z., Zhang, Y., “Fault diagnosis for diesel valve trains based on
time-frequency images”, Mech. Syst. Signal Pr., Vol 22, pp. 1981–1983, 2008.

15. Sheng, G., Vehicle Noise Sound Vibration and Sound Quality, SAE interna-
tional, 2012.

16. Cohen, L., “Time-Frequency Distributions – A Review”, Proceeding of the
IEEE, Vol. 77, No. 7, 1989.

17. Ball, A., Gu, F., Weidong, L., “The Condition Monitoring of Diesel Engines
using Acoustic Measurements, part 2: Fault Detection and Diagnosis”, SAE
Special Publication SP 1501, 2000.

18. Daubechies, I., Ten Lectures on Wavelets. Philadelphia: Society for Industrial
and Applied Mathematics, 1992.

19. Chiollaz, M., and Faver, B., “Engine Noise Characterization with Wigner-
Ville Time-Frequency Analysis”, Journal of Mechanical Systems and Signal
Processing, Vol 75, pp. 375–400, 1993.

20. Chiatti, G., Chiavola, O., Fulvio, P. and Andrea, P., “Diagnostic methodol-
ogy for internal combustion diesel engines via noise Radiation’’, Energy
Conversion and Management, Vol 89, pp. 34–42, 2015.

21. Chiatti, G., and Chiavola, O., “Combustion Induced Noise in Single Cylinder
Diesel Engines”, Small Engine Technology Conference Graz, Austria
September 27–30, SAE Paper no. 2004-32-0071, 2004.

22. Chiatti, G., Chiavola, O., “Experimental analysis of combustion noise in spark
ignition engine”, NVC conference, Tranverse city, Michigan, USA, May 5-8,
SAE paper no. 2003-01-1442, 2003.

23. Wu, J., Chen, J., “Continuous wavelet transform technique for fault signal
diagnosis of internal combustion engines”, NDT International Vol 39,
pp. 304–31, 2006.

24. Wang, J., McFadden, P., “Application of Wavelet to Gearbox Vibration Signals
for Fault Detection”, Journal of Sound and Vibration 192 (5), pp. 927–939,
1996.

25. Lin, J., Zuo, M., “Gearbox Fault Diagnosis Using Adaptive Wavelet Filter.
Mechanical Systems and Signal Processing”, 17(6), pp. 1259–69, 2003.

26. Smith, T., “The Application of the Wavelet Transform to the Processing of
Aeromagnetic Data”, PhD Thesis, The University of Western Australia, 2000.

Liquid Piston Engines. Aman Gupta, Shubham Sharma, and Sunny Narayan.
© 2017 Scrivener Publishing LLC. Published 2017 by John Wiley & Sons, Inc.

7

Sources of Noise in
Diesel Engines

7.1 Introduction

NVH study of a diesel engine is important from point of view of over-
all design/planning of system. It includes testing variations under various
testing conditions as well as engine-to-engine variations. The theory of
NVH applications in diesel engines was introduced by Reinhart et al. [1].
In modern times there are five major simulation tools that have been used
in the NVH analysis which includes:

1. Boundary element method (BEM),
2. Gas dynamics,
3. Multi-body dynamics,
4. Engine cycle simulation and
5. Finite element analysis.
There are several sources of noise in diesel engines some of which are
discussed in following sections.

167

168 Liquid Piston Engines

7.2 Combustion Noise

Combustion noise is an important concern for design and performance
calibration of system design and forms a primary source of noise for direct
injection diesel engines. There are three modes of transfer of this noise to
surroundings which include from cylinder top, liner walls and connecting
rod assembly. It is related to rate of pressure rise in cylinders. Tung and
Crocker have studied combustion noise in turbocharged diesel engines
[2]. Structural attenuation also affects the combustion noise radiated
from engine which is difference between in cylinder pressure spectrum
and noise radiated from surface. High structural stiffness of cylinder bore
leads to higher values of resonant frequencies. Higher values of resonant
frequencies than combustion excitation frequencies can help to attenuate
high frequency combustion noise. Knocking or clatter is also an important
source of noise in low speed engines. Combustion noise can be reduced
either by increasing the structural attenuation of engine or by reducing
in cylinder pressure. Reduction in delay period in ignition leads to lower
values of in cylinder pressure and hence combustion noise. Other factors
which can help reduce the combustion noise include higher compression
ratio, increased intake boost pressure, higher exhaust gas recirculation
rates and increased structural attenuation of parts.

The transfer function of engine structural attenuation has been experi-
mentally determined by Shu [3]. Combustion noise optimization by use of
combustion noise meter was done by Wang [4]. Combustion noise assess-
ment by decomposition of cylinder pressure has been done by Torregrosa
[5] showing that this method is more accurate as compared to traditional
block attenuation curve. Effect of cetane number on engine noise was
measured by Machado and De Melo [6]. Engine noise during cold start
of engine was analyzed by Alt [7]. Cylinder CFD modeling of combus-
tion noise has been conducted by Blunsdon [8] and Luckhchoura [9].
Cyclic variations of combustion noise variations were studied by Gazon
and Blaisot [10]. Analysis of noise emissions from a medium sized diesel
engine has been performed in [11, 12].

7.3 Piston Assembly Noise

There are three types of noises that occur in piston assembly. These include
pin tickling noise, piston slap and piston rattle noise. Piston slap is a major
contributor towards mechanical noise sources in diesel engines which is
caused due to secondary motion of piston between skirt and liner bore.

Sources of Noise in Diesel Engines 169

Clearance between piston skirt top corners200 Piston skirt Scraping 720
and cylinder bore at thrust side (micron)corner 3 the bore
Piston skirt
540 630
150 corner 4

41
100

3 Piston 2
50 in bore

Slap
0

0 90 180 270 360 450
Crank angle (degree)

Figure 7.1 Simulation of piston secondary motion.

There may be several events of slap in an engine cycle with most prominent
occurring just after TDC firing position as shown in Figure 7.1.

Piston slap is affected by the following major factors which influence
piston secondary motion:

a. Piston side thrust force-lower speeds, lower piston assembly
mass, lower in cylinder pressures and higher crank radius to
connecting rod length ratio can help to reduce side thrust
force and hence piston slapping noise [13].

b. Moments about piston pin- lower inertia of assembly, proper
piston pin offset and crankshaft offset, proper supply of the
lubricant and the piston pin friction force moments can help
to reduce the piston slap noise Munro and Parker [14].

c. Allowable distance of travel before hitting liner wall – smaller
gap between piston skirt-to-bore can help to reduce the slap-
ping noise at the expense of increased shear frictional force.

d. Oil damping force –sufficient supply of oil on the skirt can
help to reduce piston slap significantly. Lower tension, lon-
ger skirt length and increasing the contact area in the piston
rings can help to increase the oil film thickness and hence
piston slap [15].

e. The stiffness and the damping of the parts –impact of softer
piston skirt causes lesser noise emissions due to due to larger
deformations. It is important to increase the gap between
top land and bore in order to avoid the contact which other-
wise would produce sharp rattling noise.

170 Liquid Piston Engines

The slap noise is most common at idle cold start conditions and high
load-low speeds. The proper skirt design is an important measure to mini-
mize the piston slap noise The two most commonly used methods include
reducing the gap between liner and skirt and offsetting the piston pin.

7.4 Valve Train Noise

This type of high frequency noise is a major issue in NVH analysis of
engine and includes the following three major excitation sources:

a. Cam acceleration-The opening and closure of cam excites
high frequency vibrations at high speeds due to inertial
forces.

b. Valve train impacts-These include impulsive impacts at
valve opening between cam and follower, between valve seat
and valve at closure of valves and valve train separation and
bouncing at high speeds.

c. Frictional Vibrations-This noise is dominant at low speeds
when asperity contact occurs cam and follower near nose of
cam when lubricant velocity becomes zero.

Valve train noise identification was done using valve acceleration–cam
angle diagram and a time–frequency diagram [16]. Anderton and Zheng
[17] found that the valve train vibrations had a major contributions towards
total noise at high engine speeds ranges of above 2000–3300 RPM. Savage
and Matterazzo [18] has done experiments on a 3.3 L gasoline engine to
show the effects of various factors like cam jerk level, valve spring load,
tappet-to-bore clearance, valve stem gap and surface finish, rocker arm
bearing clearance, valve overlap, and cylinder head mass and damping.
Use of high precision manufacturing of cam profile, greater oil film thick-
ness, higher valve train stiffness and smaller tappet-to-bore clearance is
very important for reduction in valve train noise [19, 20].

7.5 Gear Train Noise

The rattle noise in transmission drive trains is primary cause of concern in
NVH development. The engine transmits non uniform torque from crank
train to drive train that causes gear rattle noise. Clearances are provided
between meshing tooth of gears to account for thermal expansions and

Sources of Noise in Diesel Engines 171

manufacturing tolerances. When gears are lightly loaded at low speeds
with strongly oscillating torque, there is a high chance of meshing teeth
separation and resulting vibrational impacts. Other type of gear noise
include whine noise which is due to tooth deflection under load.

Gear train operational noise is dependent upon number of meshing
teeth, size of gear train, magnitude of torsional inputs and location of gear
train. Detailed investigation of gear train noise was done by Spessert and
Ponsa [21] and Zhao and Reinhart [22].

7.6 Crank Train and Engine Block Vibrations

The torsional vibrations in crank shaft, the thin sections in the engine block
and the covers and the cylinder block are important sources of noise and
vibrations. Commercial software like ENGDYN are available to analyze the
response of the crank train and engine block system taking into account
oil film lubrication models [23]. Detailed analysis of crankcase and engine
block vibrations was done by Russell [24], Ochiai and Yokota [25], and
Maetani et al.[26].

7.7 Aerodynamic Noise

Low frequency Intake noise is due to turbulent fluctuations in flow of air at
inlet ducts which depends on intake valve opening area and engine speed.
Gas dynamics based design of intake ducts has been done by Silvestri et al.
[27]. Exhaust noise is due to pressure variations in exhaust duct due to
periodic charging and discharging in engines. The noise due to mechanical
vibrations in exhaust pipe is known as shell noise. Pang et al has described
exhaust manifold design to control system vibrations [28]. Turbocharging
noise also forms an important part of aerodynamic noise.

7.8 Bearing Noise

Bearings in crankshaft as well as connecting rod have clearances which are
likely to generate noise under action of external excitation forces [29].

One of major noise sources due to bearing effect is rumbling noise
which is due to engine torsional and bending resonance induced by clear-
ances. Figure 7.2 shows the key bearing parameters effecting rumbling
noise. Key methods to control bearing noise include optimization of

172 Liquid Piston Engines

Concord bearing clearance

Thrust bearing
clearance

Main bearing Bearing beam
clearance stiffness

Figure 7.2 Various bearing parameters effecting engine noise.

clearances, application of optimal crank shaft damper and application of
flexible flywheel design.

7.9 Timing Belt and Chain Noise

Major sources of noise for timing chain include meshing impact and poly-
gon effects [30]. The meshing frequency depends upon engine speed and
number of meshing teeth.

The polygon effect causes elevation and drop of chain element and
hence leading to transverse and torsional vibrations of chain. The impact
speed of roller(WA) and chain sprocket can be estimated in terms of its
pitch (PH), number of teeth (D), pressure angle ( ) and number of sprocket
teeth (Z) as:

WA NPH Sin 360 (7.1)
3600 Z

The impact energy EA can estimated by linear density of chain (υ):

EA WA2 PH (7.2)
2000

Use of rubber ring sprocket in chain sprocket can help to reduce noise
as shown in Figure 7.3(a).

Combustion engines also have transmission belts systems (Figure 3.4)
which can exhibit number of modes of vibrations as seen from Figure 7.5.

Sources of Noise in Diesel Engines 173

100Sound pressure level (dB)
Standard tooth

0.5p 90

80

70 Asymmetric tooth
60
Compression Tooth
(a)
Rubber 50
0
1000 2000 3000 4000 5000 6000
(b) RPM

Figure 7.3 Schematic representation of timing chain and its noise spectra.

1

2 6

5
3

4

Figure 7.4 Timing belt transmission system(1: Sprocket, 2: Tensioner, 3: Fuel pump
sprocket, 4: Crankshaft sprocket, 5: Idler sprocket, 6: Water pump sprocket).

(a) Transverse vibration (b) Axial vibration

(c) Torsional vibration C
Figure 7.5 Timing belt vibration sources.
C
(d) Lateral vibration

174 Liquid Piston Engines

Due to differences in pitch of tooth belt and sprocket timing belt, the
meshing belt creates a meshing impact which is a periodic excitation hav-
ing frequency (fs) estimated as:

fs ZD (7.3)
60

Typical spectrum of belt noise. Meshing noise is dominant at low and
medium frequencies, whereas meshing frictional noise is dominant over
6 kHz frequency [31].

7.10 Summary

Internal combustion engines have several contributing noise sources, such
as noise due to combustion process, fuel injection process, piston slap noise
and valve operation noise. It is necessary to separate the contributions due
to various sources and then analyze each of them individually. Figure 3.7

Noise fe
amplitude

80 fn 2fn Tz Time
60 63 fz fe
40
SPL (dB)
30
200 630 2000 6300 20000
Frequency (Hz)

Figure 7.6 Timing belt noise spectra.

Noise sources Vibration generation Paths of transfer Radiation of noise
Head vibrations Noise radiation
In cylinder Gas force Block vibrations
pressure Piston slap
Vibrations in Crank shaft vibrations
Inertia of piston connecting rod

Figure 7.7 Mechanism of noise generation.

Sources of Noise in Diesel Engines 175

90 1 8
2

Noise level dB (A) 80 3 4 5
67

70

60

Figure 7.8 The total noise contribution (8) can be decomposed into contributions due
to combustion noise (1), contribution due to piston slap noise (2), contribution due
to fan noise (3), contribution to gear operation noise (4), contribution due to pump
operations (5), valve noise (6), other sources (7).

depicts noise generation mechanism for various sources with a typical con-
tributions of various sources of noise towards sound pressure level of engine
measured at a distance of 1m away from engine in an anechoic chamber
featured in Figure 7.8.

As evident from the above figures the combustion based noise and
piston slap contributes a major portion (about 80%) towards noise emis-
sions from engines [31], hence it is necessary to focus on these two
important aspects of engine acoustics as discussed in following units of
this work.

References and Bibliography

1. Reinhart, T E., “An evaluation of the Lucas combustion noise meter on
Cummins B series engines”, SAE paper 870952, 1987.

2. Tung,V., and Crocker, M J., “Diesel engine noise and relationship to cylinder
pressure”, SAE paper 820237, 1982.

3. Shu, G., Wei, H., and Han, R., “The transfer function of combustion noise in
DI-diesel engine”, SAE paper 2005-01-2486, 2005.

4. Wang, S.,Chalu, C., and Gautier, F., “Optimization of combustion noise of
modern diesel engines for passenger cars”, SAE paper 2007-01-2379, 2007.

5. Torregrosa, A J., Broatch, A., Martin, J., and Monelletta, L., “Combustion noise
level assessment in direct injection diesel engines by means of in-cylinder
pressure components”, Measurement Science and Technology, 18, 2131–2142,
2007.

6. Machado, G B., and De Melo, T C C, “Diesel cetane number versus noise
emission”, SAE paper 2005-01-2150, 2005.

176 Liquid Piston Engines

7. Alt, N., Sonntag, H., Heuer, S., and Thiele, R., “Diesel engine cold start noise
improvement”, SAE paper 2005-01-2490, 2005.

8. Blunsdon, C A., Dent, J., and Das, S., “Modelling the origins of combustion
noise in the indirect injection diesel engine”, SAE paper 952432, 1995.

9. Luckhchoura, V., Won, H., Sharma, A., Paczko, G., and Peters, N.,
“Investigation of combustion noise development with variation in start of
injection using 3-dimensional simulations by applying representative interac-
tive flamelet (RIF) model”, SAE paper 2008-01-0950, 2008.

10. Gazon, M., and Blaisot, J., “Cycle-to-cycle fluctuations of combustion noise in
a diesel engine at low speed”, SAE paper 2006-01-3410, 2006.

11. Chiatti, G., Recco, E., Chiavola, O., and Conforto, S., “Acoustic Assessment
in a Small Displacement Diesel Engine,” SAE Technical Paper 2014-32-0129,
doi:10.4271/2014-32-0129, 2014.

12. Chiavola, O., Chiatti, G., and Recco, E., “Analysis of the Relationship between
Noise Emission and in-Cylinder Pressure in a Small Displacement Diesel
Engine,” SAE Technical Paper 2014-01-1364, doi:10.4271/2014-01-1364,
2014.

13. Oetting, H., Pundt, D., and Ebbinghaus, W., “Friction in the piston group and
new ideas for piston design”, SAE paper 841299, 1984.

14. Munro, R.,and Parker, A., “Transverse movement analysis and its influence on
diesel piston design”, SAE paper 750800, 1975.

15. Ryan, J P., Wong, V W., Lyon, R H., Hoult, D P., Sekiya, Y., Kobayashi, Y.,
and Aoyama, S., “Engine experiments on the effects of design and operational
parameters on piston secondary motion and piston slap”, SAE paper 940695,
1994.

16. Suh, I-S., and Lyon, R H., “An investigation of valve train noise for the sound
quality of IC engines”, SAE paper 1999-01-1711, 1999.

17. Anderton, D., and Zheng, J H., “A new measurement method for separating
airborne and structure borne sound from an IC engine’s valve train mecha-
nism”, SAE paper 931335, 1993.

18. Savage, J., and Matterazzo, J., “Application of design of experiments to deter-
mine the leading contributors to engine valve train noise”, SAE paper 930884,
1993.

19. Hanaoka, M., and Fukumura, S., “A study of valve train noises and a method
of cam design to reduce the noises”, SAE paper 730247, 1973.

20. H. Kanda, M. Okubo, T. Yonezawa, “Analysis of noise sources and their trans-
fer paths in diesel engines,” SAE Technical Paper 900014, 1990.

21. Spessert, B., and Ponsa, R., “Investigation in the noise from main running
gear, timing gears and injection pump of DI diesel engines”, SAE paper
900012, 1990.

22. Zhao, H., and Reinhart, T E., “The influence of diesel engine architecture on
noise levels”, SAE paper 1999-01-1747, 1999.

23. Offner,G.,Priebsch,HH.,Ma,MT.,Karlsson,U.,Wikstrom,A.,andLoibnegger,
B., “Quality and validation of cranktrain vibration predictions  – effect of

Sources of Noise in Diesel Engines 177

hydrodynamic journal bearing models”, Multi-Body Dynamics: Monitoring
and Simulation Techniques-III, pp. 255–271, 2004.
24. Russell, M.F., “Reduction of noise emissions from diesel engine surfaces”, SAE
paper 720135, 1972.
25. Ochiai, K., and Yokota, K., “Light-weight, quiet automotive DI diesel engine
oriented design method”, SAE paper 820434, 1982.
26. Maetani, Y., Niikura, T., Suzuki, S., Arai, S., and Okamura, H., “Analysis and
reduction of engine front noise induced by the vibration of the crankshaft
system”, SAE paper 931336, 1993.
27. Silvestri, J., Morel, T., and Costello, M., “Study of intake system wave dynam-
ics and acoustics by simulation and experiment”, SAE paper 940206, 1994.
28. Pang, J., Kurrle, P., Qatu, M., Rebandt, R., and Malkowski, R., “Attribute anal-
ysis and criteria for automotive exhaust systems”, SAE paper 2003-01-0221,
2003.
29. Qatu, M. S., Abdelhamid, M. K., Pang, J., and Sheng, G., “Overview of
Automotive Noise and Vibration,” International Journal of Vehicle Noise and
Vibrations, Vol. 5, Nos. 1/2, 2009.
30. Young, J. D., Marshek, K. M., Poiret C., and Chevee P., “Camshaft roller chain
drive with reduced meshing impact noise levels,” SAE Paper No. 2003-01-
1666, SAE International, Warrendale, PA, 2003.
31. Sheng, G., et al., “A new mechanism of belt slip dynamic instability and noise
in automotive accessory belt drive systems,” International Journal of Vehicle
Noise and Vibration, Vol 2, pp. 305, 2006.

Liquid Piston Engines. Aman Gupta, Shubham Sharma, and Sunny Narayan.
© 2017 Scrivener Publishing LLC. Published 2017 by John Wiley & Sons, Inc.

8

Combustion Based Noise

8.1 Introduction

Combustion noise generated mainly depends on rapid rise of cylinder
pressure due to ignition delay period in combustion engines. Design of
combustion chamber as well as variations in various injection parameters
like injection pressure, amount of fuel injected and its timings also play a
crucial role in noise emissions [1]. Depending upon type of engine and
various operational parameters, overall noise emissions from a typical
engine are in range 80–110 dBA [2]. Anderton has investigated the effects
of turbo charging on noise emissions from engines [3]. Split injection using
electronic control reduces the premixed combustion and hence is an effec-
tive way to reduce overall noise emissions by about 5–8 dBA [4]. Head and
Wakes have shown that during transient operational conditions, overall
noise levels are 4–7 dBA higher as compared with steady state operations
[5]. Cold starting conditions lead to higher ignition delay period which
in turn causes more premixed combustion and hence an increase in noise
emissions [6]. Quality of fuel also affects combustion noise emissions from
engines. It has been seen that reduction of centane number of diesel from
50 to 40 causes a rise of 3 dBA in combustion based noise [7]. In gasoline

179

180 Liquid Piston Engines

engines, the ignition delay period is longer due to lesser compression ratio
which leads to lower temperature of charge and hence more noise [7]. For
a naturally aspirated engine the combustion noise depends upon amount
of fuel that mixes with air charge during injection delay period and hence
compression ratio of engine also plays a vital role [7].

8.2 Background of Combustion Process 
in Diesel Engines

Due to high efficiency, diesel engines have been a favorite choice for heavy
duty applications including trucks [8]. However they suffer from drawbacks
of high noise, weight and vibrations. These engines are of two types:

Direct Injection (D.I.) Engines
In direct injection engines

In the D.I. engines, the fuel is directly injected inside combustion cham-
ber and due to lesser time for mixing, a heterogamous mixture consisting
of both rich and lean parts is formed in the chamber.

Figure 8.1 shows the three phases of combustion in a conventional diesel
engine. The first phase starts with start of injection process and ends with
premixed combustion phase. The direct injection of fuel into combustion
chamber begins some crank angle degree before top dead center positions
depending upon engine operational conditions. As soon as cold jet of fuel
penetrates into chamber, it mixes up with hot compressed air. The droplets
begin to vaporize forming a sheath of vaporized fuel-air mixture around jet

Phase 1 Phase 2 Phase 3

InRjaetceteofdhmeaatssrrelaetaes(e·m(f ·uqe)l) Premixed combustion pcyl
Cylinder pressure ( ·pcyl) premixed peak
Ignition
delay q·
m· fuel

0 Injection 20 Crank angle ATDC (deg)

Figure 8.1 Phases of diesel engine combustion.

Combustion Based Noise 181

periphery. When temperatures reach around 750 K,the first break down of
Cetane fuel occurs. Further reactions produce C2H2, C3H3, C2H4, CO2 and
water vapors [9].

Resulting rise in temperature causes complete combustion of fuel-air
mixture. This sudden combustion causes rise in heat release rate and high
pressure gradient dP which further leads to high temperatures in pre-mixed

d
zone and NOx production. The premixed combustion consumes all mix-
ture around inner spray region where temperatures of ranges 1600–1700 K
are reached and all oxygen available for combustion is consumed [8].
Now various partial burnt particles diffuse to outer layers and are burnt
in a thin reaction region at periphery of spray which leads to formation of
diffusion flame.

This kind of combustion is known as diffusion controlled combustion
and is depicted by region 2 and 3 in the above figure. The high temperatures
along with lack of oxygen is ideal for soot formation.

The diffusion flame is fed with oxygen from surrounding environ-
ment and high temperatures of range 2700 K is reached consuming all the
soot formed. At outer region of flame there is enough oxygen content for
formation of NOx.

Figure 8.3 shows soot formation concentration as a function of crank
angle. Most of soot produced at early crank angles is consumed later
and final exhaust emissions have only a fraction of initial one. As seen
from Figure 8.1, the diffusion controlled combustion can be divided
into two sub phases. During phase 2 of combustion the burning rate
is controlled by mixing of fuel fragments and air and rate of reaction is
must faster. During the phase 3 the final oxidation of remaining unburnt

Soot NO~X formation
2700 K
Rich fuel-air mixture, oxidation

=4 Diffusion flame
825 K CO2, H2O

Air

950 K

Cold fuel
350 K

Ignition Products of
zone rich combustion,

CO, HC soot
~ 1600 K

Figure 8.2 Conventional diesel engine spray formation.

182 Liquid Piston Engines

Soot concentration

TDC Crank angle
Figure 8.3 Rate of soot formation.

Soot emission EGR Late SOI traSdoeo-to-NffOX
Diesel particulate filter (DPF) Low swirl

Reduced Early SOI
Intercooling

Water injection

Variable valve timing
Increase of pinj, decrease of hole size
Multiple injections (pre- and post-injections)
Rate shapping
aEfxthearutrsetagtamsDeenNt ox Increase boost pressure
...

NOX emission

Figure 8.4 Soot & NOx trade off.

particles and soot takes place, however due to decrease of gas temper-
ature during  expansion stroke and lack of oxygen the reaction rate is
much slower.

NOx and soot formation in combustion engines show opposite trends
as shown in Figure 8.4. In order to reduce NOx, it is necessary that local
temperatures must not rise above 2000 K. A possible way to do so is to
inject fuel late in cycle inside combustion chamber which shifts combus-
tion phase towards expansion phase and hence significant reduction in
chamber temperatures. However consumption of fuel and soot formation
increases due to late combustion.

Hence modern injection systems use multiple injection techniques in
order to reduce both soot as well as NOx emissions as seen from Figure 8.5
[8, 10, 11].

These generally use three phases of injection namely pre-injection
Period, Main – Injection Period & Post injection period as seen from

Combustion Based Noise 183

Injection pressure 2c 1 Pre-injection
2b 2 Main-injection
3a 2a rectangular type
2a
2b boot type
2c ramp type

3 Post-injection
3a early
3b late 3b

1

p0
TDC Crank angle

Figure 8.5 Multiple injection methods adopted for modern diesel engines.

Figure 8.5. There is delay period between the start of ignition process and
fuel injected inside diesel engine. The more this ignition delay, more is the
temperature during combustion and hence better condition for NOx for-
mation. To shorten the delay period, small amount of fuel is pre-Injection
before main injection during the pre-mixed combustion phase. The torque
and power produced in engine depends upon main injection period. It is
advantageous to vary injected fuel mass with time to reduce the specific
fuel consumption. This method is known as rate shaping. Rate shaping
may be rectangular, step or boot in shape. Post-injection of fuel is done
to reduce soot emissions and in some cases may be useful for Exhaust
gas recirculation treatment of gases [12]. It has been reported that post
injection reduces soot by about 70% without increasing the fuel consump-
tion [13].

8.3 Combustion Phase Analysis

This part of work presents experimental data in which signals from
accelerometers and cylinder pressure transducers are used for analy-
sis of combustion behavior of engines. Previous works have shown
that engine vibrations are sensitive towards change in engine injection
parameters. Accelerometer signals have been able to locate the impor-
tant features of combustion process in diesel engines [14, 15]. The aim
of present part of work is to explore the relationship between block
vibrations and in cylinder pressure development process. Experiments
were done at various load and speed conditions to explore the sensitivity
of vibration data.

184 Liquid Piston Engines

8.4 Combustion Based Engine Noise

The aim of this part is to provide an overview of combustion noise gen-
eration process. This was done by analyzing the methodology provided in
previous works done. Engine combustion noise originates from combus-
tion process taking place in cylinder. When fuel is injected inside the com-
bustion chamber where high pressure air is present, then part of ignitable
gas starts to burn causing a rapid rise in pressure as well as temperatures
inside combustion chamber. The pressure wave thus generated also strikes
the walls of combustion chamber causing resonance of structure. The
vibrations are radiated in air through engine structure and are perceived as
combustion noise.

In actual practice it is difficult to distinguish between combustion noise
and piston slap as both coincide near top dead center positions. For sake
of convenience it is assumed that combustion noise originates due to pres-
sure vibrations inside engine cylinder and is transmitted to cylinder cover,
piston, connecting rod, crank shaft and engine surroundings. Mechanical
noise includes noise from piston-liner impact, valve operations, pump
operations, injector operations as well as operation of various accessories
and valve trains. Generally for indirect injection diesel engines and gaso-
line engines, the combustion process is less severe as compared with diesel
engines, hence the combustion noise is lesser as compared with mechani-
cal noise.

The cause of combustion noise is rapid change in cylinder pressure dur-
ing combustion process. The effects of combustion process consists of high
frequency gas vibrations and dynamic load due to rapid pressure change.
The intensity of combustion noise (I) is dependent upon the values of max-
imum pressure value (Pmax) and maximum rate of pressure rise as [20]:

dP 2
dt
I Pmax (8.1)

max

In a direct injection diesel engine, combustion process occurs in follow-
ing four phases: retarded, rapid, early and late phase. Combustion noise
is mainly generated during rapid phase of combustion however retarded
phase has an indirect effect on combustion noise. During this phase of
combustion, the firing and fuel propagation results in impulsive pressure
wave that gets reflected multiple times after striking walls of chamber. This
process causes high frequency oscillations. The frequency (fg) can be esti-
mated from engine bore diameter (D) & Wave propagation speed (Cc) by:

Combustion Based Noise 185

fg Cc (8.2)
2D

Further the spectrum plot of cylinder pressure can be obtained from
acquired about in cylinder as seen in Figure 8.6.

The graph is marked by following three distinct regions [20]:

a. Region of low frequency-in this region the curve depends
upon peak cylinder pressure. Higher the maximum value of
cylinder pressure, higher is the peak in low frequency range.

b. Medium frequency range-in this part the pressure levels
decrease in logarithmic range with slope depending upon
rate of cylinder pressure Larger the value of pressure gradi-
ent, steeper is the slope of curve.

c. High frequency range-in this range rapid evolution of in
cylinder pressure occurs due to onset of combustion pro-
cess which results in high frequency vibrations of cylinder
structure having amplitude dependent on second cylinder
pressure derivative.

Cylinder pressure (bar) 80 d2p
d 2 max
60 dp
d max 100 200 300

40

20 pd

0

–20
–500 –400 –300 –200 –100 0
Crank angle

Cylinder pressure 102.44 pd d2p
spectrum (dB) 102.4 d 2 max
102.36
dp
d max

102.32

101 102
Frequency-Hz

Figure 8.6 Regions of combustion noise.

Attenuation (dB)186 Liquid Piston Engines

100
80

0.5 1 2 5 10
Frequency (kHz)

Figure 8.7 Attenuation curve of engine.

Combustion noise is not only dependent upon cylinder pressure but
also upon the structural response and damping effects of engine. The dif-
ference between in cylinder pressure and outside noise radiated in engines
is characterized by a decay which reflects the structural attenuation of
engine structure. Figure 8.7 shows a typical plot of structural attenuation
of an engine which is independent of exciting forces and cylinder pres-
sure spectrum. Various operational parameters of engine e.g. load, speed &
fuel injection parameters have significant effects on structural attenuation
property.

This curve can be divided into three distinct regions:

a. Below 2000 Hz-in this range the attenuation is very high as
most of engine parts have high stiffness with their natural
frequencies in low or mid ranges.

b. In mid frequency ranges of 2000–5000 Hz, the attenuation
is small as most of natural frequencies of engine parts fall in
this range.

c. Above 5000 Hz – natural frequencies of most of the parts
is above the natural frequencies of most of the parts, hence
attenuation is quite high.

8.5 Factors Effecting Combustion Noise

Controlling the rate of pressure is the key to control combustion noise
which mainly depends upon the ignition delay and quantities of combus-
tion gas formed during delay period. A shorter delay period means lesser
amount of combustible gas formed in the cylinder and hence lesser com-
bustion noise. Hence delay period must be reduced as much as possible
to reduce combustion noise. Structure and layout of engine also plays an
important role.

Combustion Based Noise 187

Increase in compression ratio and chamber temperature also shortens
the delay period. However increase in compression ratio may cause rise in
the piston slap noise. Various parameters of fuel injection system like angle
of fuel injection, injection pressure, number of nozzles and fuel supply rate
also effect the combustion noise in engines. An increase in the injector
pressure leads to an increase in the amount of fuel accumulated in the igni-
tion delay period and hence an increase in the combustion noise.

If other factors remain unchanged, an increase in engine speed increases
the fuel injection rate and hence greater quantity of fuel is injected in the
period of combustion delay which leads to an increase in combustion noise.

There are many approaches to control combustion noise. One of these
includes reducing cylinder pressure spectrum typically in middle and high
frequency ranges. Others include reducing ignition delay period or amount
of combustible gases formed during this period. Increasing the stiffness of
parts, use of turbo charging process and use of split injection methods in
engine cycle have also proved to be effective methods.

8.6 In Cylinder Pressure Analysis

The analysis of in cylinder pressure is one of the most common ways to
examine the combustion noise in diesel engines. Anderton and Priede were
first to observe that abrupt combustion led to high frequency contents of in
cylinder pressure spectrum [24]. Crocker suggested that frequency contents
up to 300 Hz were related to maximum cylinder pressure [25]. Between
300–2000 Hz they were related to first derivative of in cylinder pressure,
whereas above 2000 Hz were related to second derivative of cylinder pressure.

8.7 Effects of Heat Release Rate

Previous works have shown a relationship between peak of combustion
noise and overall heat release rate [26]. Rusell has developed a technique
based on block attenuation curve which is still most reliable one for study
of combustion noise [27]. It was observed that higher slopes of rate of heat
release curves led to higher combustion noise irrespective of fuel injection
timings [28]. There is a tradeoff between combustion efficiency and noise
generated due to combustion [29]. Efficient combustion leads to higher
heat release rate near top dead center which gives rise to high frequency
components in noise spectrum. Late heat release rates leads to lower in
cylinder pressures and hence low frequencies in combustion noise.

188 Liquid Piston Engines

Cylinder pressure (bar) Lean slope A weighted SPL
100 Sharp slope
50 100 200 500 1k 2k
Heat release rate 80 Crank angle
150
kJ/kg-deg.
60

50 100

40
50

00 20 5k 10 kHz
–60°–30° tdc 30° 60° 90° 20

Crank angle

Figure 8.8 Effects of heat release rate on combustion noise.

dB 180
170
160 Cycle 1
150 Cycle 2
140 Cycle 3
130
120 Frequency (Hz) 104
110

103

Figure 8.9 Cyclic variations in combustion noise.

It was observed that a 10 dB reduction in sound pressure levels was pos-
sible without change in fuel consumption. However, there is a fall in effi-
ciency of cycle and smoke emissions increase if ROHR is not terminated
50° BTDC [27].

8.8 Effects of Cyclic Variations

Combustion process in diesel engines vary with cycle which leads to varia-
tions in combustion noise [30]. These variations may be attributed due to
different fuel injection rates, compression ratios as well as difference in
fuel spray process, mixture formation and flame propagation. Torregrosa
has studied cyclic variations in combustion noise as given by Figure 8.9
[31]. Variations may be attributed to resonance phenomenon occurring in
combustion chamber.

Combustion Based Noise 189

First Second First
circumferential circumferential radial
mode (0.1)
mode (1.0) mode (2.0)

Figure 8.10 Various modes of combustion chamber cavity.

8.9 Resonance Phenomenon

Resonance process taking place in the combustion chamber effects the
noise radiated from engines. Grover has observed high peaks in noise
spectrum which may be attributed to this phenomenon [32].

Previous works have shown effects of gas temperature and affects of
resonant frequency on high amplitude combustion chamber oscillations
[33]. Hickling has observed several peaks in cylinder pressure spectrum by
filtering data with cutoff frequencies in range 20–1500 Hz which increase
in amplitude with increase of load [34]. The frequencies of these peaks
varied inversely with cylinder bore diameter. The resonant frequency (fr)
may be defined in terms of cylinder bore (B), axial length (L) and speed of
sound (C) as:

fr K2 q2 (8.3)
m,n

2L B

where m, n, k determine circumferential, axial and radial modes.

8.10 In Cylinder Pressure Decomposition Method

The signal processing methodology to decompose cylinder pressure sig-
nals was first proposed by payri [35]. In this methodology the cylinder
pressure signals were decomposed into three parts namely-Combustion
pressure, resonance pressure and compression pressure. The combustion
art which is strongly dependent on rate of heat release is defined by injec-
tion strategy in engines. Using suitable cutoff frequencies motored part of

190 Liquid Piston Engines

60

Total pressure

Motored pressure 102.8

Cylinder pressure -bar
Pressure level (dB)
40 Excessive pressure

20

102.7 Total pressure
0 Motored pressure
Excessive pressure
–20
–360 –240 –120 0 120 240 360 102 103
Crank angle Frequency (Hz)

Figure 8.11 Time and frequency decomposition of cylinder pressure signal.

pressure signal was isolated from excessive pressure signals. Figure 8.11
shows results of such a methodology adopted for a cylinder pressure signal.

As evident from these figures, motored pressure signal dominates the
pressure spectra at low frequency ranges. By subtracting the motored pres-
sure signal from total pressure signal, the excessive part of signal can be
obtained. The resulting signal has contributions due to both resonance
as well as combustion pressures. Resonance portion is clearly visible with
fluctuating peaks in pressure spectra. Hence it can be isolated by filter-
ing excessive pressure signal using high pass filter. The resulting resonance
signal can be subtracted from excessive portion to obtain contribution due
to combustion process. Figure 8.11 shows total decomposition of cylinder
pressure signal. It is evident from the lot that contribution due to com-
bustion portion dominates in mid frequency range whereas the resonance
phenomenon dominates at high frequency ranges [36].

The decomposed signals can be used to calculate noise indices defined
in terms of ideal engine speed as:

N
In n log (8.4)
N ideal (8.5)
(8.6)
dp dp

N dt pilot dt main
I1 Nideal
dp

dt motored

I2 10 log(106 P2 dt
residual

P2
motored

Combustion Based Noise 191

Cylinder pressure -bar 60 Total pressure
Motored pressure
50 Excessive pressure
40 Resonance pressure
30 Combustion pressure
20
10 –240 –120 0 120 240 360
Crank angle
0
–10

–360

Pressure level (dB) 1028 Total pressure
Motored pressure
Resonance pressure
Combustion pressure

1027

101 102 103
Frequency (Hz)

Figure 8.12 Total decomposition of cylinder pressure signal.

where
dP is maximum pressure gradient during pilot injection
dt

pilot

period,

dP is maximum pressure gradient during main injection,
dt main
Presidual is residual pressure and

dP is maximum pressure gradient in motored pressure
dt

motored

signal.

These indices can be further used to find overall noise (ON) emitted

from engine as:

ON – C0 + C1I1 + C2I2 + CnIn (8.7)
where constants C0, C1, C2 and Cn depend upon size of engine.

192 Liquid Piston Engines

8.11 Mathematical Model of Generation of
Combustion Noise

In wavelet transformation complex morlet wavelet can be used as mother
wavelet. A morlet wavelet may be defined in terms of central frequency
fc and bandwidth fb as:

(t) 1 e ei2 fct t2 (8.8)
fb

fb

Figure 8.13 shows real and imaginary parts of such a morlet wavelet
having fb = 1.5 and fc = 1

Figure 8.14 shows a transient model of combustion noise generation in
engines. This model can be analyzed by following three process: generation

Complex morlet wavelt cmor 1.5-1
0.5

0

–0.5
–8 –6 –4 –2 0 2 4 6 8
Real part

0.5

0

–0.5
–8 –6 –4 –2 0 2 4 6 8
Imaginary part

Figure 8.13 Complex morlet wavelet.

Rate of decay C(f ) Radiation area
Engine structure

Combustion impact Combustion chamber
power Wc
Vibration Combustion noise
Piston energy E(f) power Wn

Transmission Rate of
rate (f ) radiation

b(f )

Figure 8.14 Noise generation model.

Combustion Based Noise 193

of vibrational energy in combustion chamber due to combustion process,
decay of energy in the engine structure and finally radiation of energy
around engine surface.

The combustion process in engine generates combustion impact power
(Wc) which can be expressed in terms of in cylinder pressure (p), imped-
ance of medium ( c) and cylinder surface area (A) as:

Wc P2 A (8.9)

c

The available energy at engine surface can be expressed in terms of
transmission rate coefficient η(f) as:

E( f ) t (8.10)

( f ) 0 Wcdt

Differentiating both sides of this equation we have:

d (E( f )) (f )Wc (8.11)
dt

Taking into account the decay rate this equation gets modified as:

d (E(f )) (f )Wc C(f )E(f ) (8.12)
dt

where the decay constant C(f) may be defined as:

C( f ) d[log(Wn)] (8.13)
dt

d (Wn ) b( f )h( f )Wc C( f )Wn( f ) (8.14)
dt

Figure 4.85–4.89 shows the plots of combustion impact power plotted
for tests enlisted in Table 7.1 using reference value of W0 as 10–12 Watts.

8.12 Evaluation of Combustion Noise Methods

A combustion noise measurement technique has been proposed by
Lucas [7]. Austin & Priede have shown that combustion noise was most

194 Liquid Piston Engines

Structure response function Structure attenuation
in decibels re 20 mPa for 1 kPa excitation in decibels

80

80
70

90
60

100
50

110
40

120
30

130
20

140
10

0 150
100 200 500 1k 2k 5k 10k 20kHz
Frequency in herts

H.S.D.I.s I.D.L I.1l/cyl.DI.s I.2l/cyl.DI.s

Figure 8.15 AVL Structural response function and structural attenuation.

dominant in the frequency range 800–4 kHz [38]. Acoustic measurements
of noise outside the engine may be used for combustion noise analysis only
when the engine is operated in such a way that contribution of combus-
tion events towards the noise emissions is maximum. This can be achieved
without changing mechanical noise either by either operating the engine
with advanced injection timing or by changing Cetane number of fuel
used. Russell has used alkyl blended fuel to maximize in cylinder pressure
so that combustion noise becomes dominant [7].

Structural attenuation of engine structure also plays a vital role in deter-
mination of combustion noise. Austin and Priede were able to determine
this function by subtracting spectrum of in cylinder pressure level from
the spectrum of noise emissions recorded at a distance of 0.9 m from
engine structure [7]. Value of structure response functions fell by 10 dBA
in 500 Hz to 5 kHz range [7]. More recently AVL has developed a com-
bustion noise meter which based on analysis of engine indicator diagram
on frequency domain [39]. Good correlation was observed when the data
from this noise meter has been compared with that computed results from
computer programming. Figure 4.95 shows structural response functions
of a group of 9 engines as recorded by AVL noise meter. The response of
direct injection high speed diesel engines falls by 12 dB over 5 kHz fre-
quency range [7].

Shu has predicted this transfer function by setting an explosive charge
in cylinder at a locked crank angle position as seen from Figure 8.16 [40].