CFA/DAGA'04, Strasbourg, 22-25/03/2004

Combustion Noise and Piston Slap Noise : Identification of Two Sources Responsible for Diesel
Engine’s Sound Signature

Cyril Renard1 , Laurent Polac2

1 Laboratoire d’Acoustique de l’Université du Maine (LAUM), Avenue Olivier Messiaen - 72085 Le Mans Cedex 9, France

Email : cyril.renard@renault.com
2 RENAULT S.A. Centre Technique de Rueil, 67 rue des Bons Raisins - 92508 Rueil Malmaison Cedex, France

Email : laurent.polac@renault.com

Introduction A significant difficulty concerning combustion noise and pis-

ton slap noise problems lies in the identification of these two
Nowadays, engines’ sound quality has become an impor- different phenomenons that occur in a very short time delay.
tant criterion in customer’s choice. Moreover, environmental Many works have been carried out about piston slap noise re-
agencies tend to limit the noise level emissions of motor ve- duction : De Luca et al. [1] have proposed a literature review
hicles. For these purposes, car manufacturers tend to work in about piston slap excitation both from an analytical and ex-
two ways : perimental point of view. Haddad et al. [2] have shown the
importance of piston design on the vibration level of Diesel
sound design which consists in making “good” noise engines. Okubo et al. [3] have measured cylinder wall trans-
components appear. For example, an engine will inspire fer function to compute a new piston design. Then they have
a “sporty” image to the audience by the rise of harmonic checked its effects on the piston slap noise reduction. Ryan
3 and 6 because it will sound like a 6 cylinders, et al. [4] have planned an experiment whose purpose was to
determine the effects of operational parameters such as spark
noise reduction which is more about lowering global timing on a petrol engine.
noise level by decreasing “bad” noise components, like
aerodynamic noise or combustion noise.
Test description
This work registers in the noise reduction part. It deals with The engine used for the experiments was a 4 cylinders, 1.9 l.
combustion noise and piston slap noise that are two unde- direct injection Diesel engine fixed to a test cell pallet using
sirable noise components responsible for the Diesel engine’s standard vehicle mounts. Measurements were made for the
sound signature. cylinder n 3, at 960 rpm, no load, in cold start conditions.
They provided the cylinder pressure, vibrations on both thrust
Piston slap noise phenomenon appears consequently to the and anti-thrust sides of cylinder block, vibrations on cylinder
rapid pressure rise in the combustion chamber due to the fuel head, and vibrations on cylinder block skirts.
self-ignition. Around the top dead center position, the piston
moves through its clearance to bore and hits the cylinder wall, By increasing or decreasing each source independently
causing the piston slap noise. This noise particularly occurs (namely combustion and piston slap), it has been possible to
under cold conditions (e.g. cold start at low temperatures) assess their respective contributions in terms of time localisa-
when clearances are wider, and then decreases as the engine tion, amplitude, and frequency.
gets hot conditions. Clearances at cold conditions are calcu- We set three types of combustion in cylinder n 3 by using a
lated with respect to dilatation coefficients which are different specific electronic control unit : one built with no combustion,
for pistons (made of aluminium) and cylinder block (most of the second built with a standard combustion and the third built
the time made of cast iron), in order to avoid scuffing while with a severe combustion.
working under hot conditions.
To control the piston slap noise source, two sets of pistons
were manufactured : a “low noise piston” characterized by a
% &' ( ( $

low piston to bore clearance, and a “high noise piston” with a

! "# $

wide piston to bore clearance.

! " #$ % & + , &(-! " # $ % &
' () * ' () *

! "# $ ! "# $

Results
A time-frequency analysis was performed on the vibrations
measured on the cylinder block for each configuration. Figure
2 shows the time-frequency map for a high noise piston and
. * / * , ) 01
! "# $
2 3 4 5 $ % &(3 ,
6 (% &3 , 0' 78 9
2 3 4 5 $ % &(3 , 0&# 8 , % 4 (% % (3 , 09 8 &"
no combustion in cylinder n 3. It shows that piston slap is
6 (% &3 , 0' 78 9 0&# 8 , % 4 (% % (3 , 09 8 &"
2 # 8 , : % " 8 ;&0# 3 &8 &(3 ,
responsible for vibrations around 2000 Hz.
Figure 3 shows the same representation for a low noise piston
Figure 1: Schematic representation of the test apparatus (view from and a severe combustion. Frequencies between 3500 Hz and
pulley side). 6500 Hz seem to be particularly impacted by the combustion.

607
CFA/DAGA'04, Strasbourg, 22-25/03/2004

Figure 4 summarizes the engine vibrations breakdown. Each

6
colored area represents the mean spectrum calculated around
Cylinder pressure #3
5
x 10
top dead center position in cylinder n 3. The grey area is the
4
spectrum calculated with a low noise piston and no combus-
tion in cylinder n 3. The blue area is the spectrum for a high
3
[Pa]

1
noise piston and no combustion in cylinder n 3. The red area
0 is the spectrum for a high noise piston and a standard combus-
50
Acceleration cylinder #3 Thrust side tion in cylinder n 3, and the green area represents the spec-
trum calculated for a high noise piston and a severe combus-
Amplitude [m.s−2]

0
tion.
Engine vibrations breakdown − cylinder #3 Thrust side
45

−50
40
Time−Frequency map of acceleration − amplitude [m.s−2]
7000
35

6000 30

5000 25
SEVERE
COMBUSTION
Frequency [Hz]

dB [m2.s−4]
4000 20 PISTON
SLAP

3000 15
COMBUSTION
10
2000

5
1000

0
0
0 0.02 0.04 0.06 0.08 0.1 0.12
Time [s] −5

−10
0 1000 2000 3000 4000 5000 6000 7000
Frequency [Hz]

0 5 10 15 20 25 30

Figure 4: Engine vibrations breakdown for the thrust side on the

Figure 2: Time-frequency map of vibrations measured on the thrust engine block.
side of the engine block. Configuration : high noise piston and no
combustion in cylinder n 3.
Conclusion
The method used to identify combustion noise and piston slap
noise contributions seems to be very efficient. Its real weak-
ness lies in the time necessary to plan such an experiment.
As a matter of fact, the combustion is easily controlled by
6
x 10 Cylinder pressure #3
6

4
the use of a specific electronic control unit. Nevertheless the
change of piston set implies a huge mechanical manipulation
[Pa]

1
of the engine. In further works we will try to realise the radi-
0 ated noise breakdown by the use of different signal processing
50
Acceleration cylinder #3 Thrust Side tools.
Amplitude [m.s−2]

0 References
[1] J.C. de Luca & S.N.Y. Gerges, Piston slap excitation :
−50 literature review, Society of Automotive Engineers paper
7000
−2
Time−Frequency map of acceleration − amplitude [m.s ]
n 962395, 1996.
6000
[2] S.D. Haddad & H.L. Pullen, Piston slap as a source of
5000 noise and vibration in Diesel engines, Journal of sound
and vibration (1974), 34(2), 249-260.
Frequency [Hz]

4000

3000
[3] M. Okubo, H. Kanda & T. Yonezawa, Analysis and re-
2000 duction of piston slap noise in Diesel engines, Society of
1000 Automotive Engineers paper n 890127, 1989.
0
0 0.02 0.04 0.06
Time [s]
0.08 0.1 0.12 [4] J.P. Ryan, V.W. Wong, R.H. Lyon, D.P. Hoult, Y. Sekiya,
Y. Kobayashi & S. Aoyama, Engine experiments on the
effects of design and operational parameters on piston
0 5 10 15 20 25 30
secondary motion and piston slap, Society of Automotive
Engineers paper n 940695, 1994.
Figure 3: Time-frequency map of vibrations measured on the thrust
side of the engine block. Configuration : Low noise piston and severe
combustion in cylinder n 3.

608