Vous êtes sur la page 1sur 6

14th Annual (International) Mechanical Engineering Conference May 2006

Isfahan University of Technology, Isfahan, Iran

Dynamic Response of an Overhead Cam Valve Train

H. Ahmadian1, M. Shahravi2
School of Mechanical Engineering, Iran University of Science and Technology
Narmak, Tehran, 16844, IRAN
E-mail ahmadian@iust.ac.ir

Abstract
This paper presents the dynamic analysis of a complete overhead cam (OHC) type valve train configuration
of an internal combustion engine to predict its dynamic response. The valve train model is developed using
flexible component approach. Within this approach, each component of the valve train is modeled as an
object with the 3D geometry having stiffness, mass, and damping properties. The stiffness properties of the
valve train components are calculated using the 3-D finite element model of each component. The dynamic
behavior of the valve train is analyzed for the rated speed and over speed of the internal combustion engine to
determine the dynamic forces acting in the interface of the valve train components. The results in the form of
nodal displacements and valve acceleration are studied to distinguish the occurrence of the jump
phenomenon at the rated speed and/or over speed. The model predictions are compared with the experimental
observations to insure the validity of the obtained model. Having established the validity of the model, the
dynamic forces acting in the interface of the valve train components are estimated which can be used as input
for the sustainability analysis of the valve train components at rated and over speeds. The effect of variation
in the cam profile on dynamic forces acting on the valve is also studied.

Keywords: Internal combustion engine - Overhead cam (OHC)- Dynamic response - Flexible
component model Valve lash Valve bouncing.

INTRODUCTION
Modern internal combustion engine are being
developed towards higher efficiency, more
concentrated power, and greater reliability. The
valve train is an important and rather complex
component in this development. Valve train design
and the valve timing directly affect the engine
performance. The main considerations in valve train
design are:
Engine performance: The main demands here are
the exact timings (Valve opening and closing), a
high volumetric flow (fat valve lift), fast opening,
and in some cases a short overlap period. The
usual considerations for design optimization are
WOT (wide open throttle), performance
characteristics for gasoline engines and full load
smoke for diesel engines.
Durability: The limits for the design are given by
Hertz contact pressure, the oil film thickness at the
cam/follower interface, impact velocity and force

1- Associated Professor
2- PhD student

www.me-en.com
mechanical engineering

at the valve seat, prevention of contact loss, and


valve bouncing.
Vibration and Noise: Optimum solutions are
achieved with high stiffness and correspondingly
high natural frequencies of the valve train. The
impacts and the inertia forces should be kept at
minimum level.
Conflicting demands in the design may occur due to
fast opening of valve, the valve train dynamics and
its effects on the gas exchange process. On one hand
big effective opening areas of the valve are required
to optimize the engine air flow and to minimize the
pumping losses for a given valve timing. This
requirement forces the designer to go close to the
limits of valve train durability. On the other hand the
resulting high forces in the valve train may cause
significant differences between the geometric and
dynamic valve lift curves depending on the stiffness
of the valve train. The gas exchange process of the
engine, however, is controlled by the dynamic valve

14th Annual (International) Mechanical Engineering Conference May 2006


Isfahan University of Technology, Isfahan, Iran

lift. In view of this conflict the best compromise can


only be found if accurate prediction of the system
performance can be made. This is particularly
important at high speeds.
High thermodynamic efficiency of the engine
requires quick opening and closing of the valve for
any particular speed and concentrated power calls
for high operating speeds. Both these requirements
are associated with high stresses on the valve train
components and are therefore contrary to the
demand on reliability and long life of the parts
involved. However, electronic computers may be
applied to satisfy these different requirements. There
are a number of reasons to use CAE methods in the
modern engine development process. Increase in
usage of virtual testing versus costly hardware
testing is primarily due to the need to cut cost.
Another important argument is the fact that certain
characteristics of the system cannot always be
measured without influencing the system. In a first
step the characteristics, which can be measured with
relative low effort, may be used to correlate and to
verify the validity of the calculation models. In
return these models deliver a large number of
characteristics whos measurement would be too
time consuming and costly in the frame of a regular
development project. Since hardware prototypes
are not available early in the design process
especially in the pre prototype phase, the usage of
virtual prototypes is required and represents the
most important scenario. During this period the
results from calculations and the engineering
experience of the designers are the only basis for
design decisions, which influence the entire design
process until production. The more innovative the
design is the less engineering expertise is available;
hence the importance of the calculation increases
dramatically in the concept phase. During the last
two decades the significance of new concepts with
respects to the engine mechanics has reached a new
level as shown by the introduction of engines with
variable compression ratio and valve trains with
variable timing [1, 2].

VALVE TRAIN SIMULATION


According to the application it is possible to
distinguish between predictive and validated
simulation models. The predictive models should
deliver results based on physical input parameters
early in the process, when no hardware prototype is
available. The validated models are first validated
against measured data using data deviating from the
physical input data. These models are usually used
in the detailed design phase, such that the
requirements on model accuracy are much higher. In

www.me-en.com
mechanical engineering

general, models with increasing levels of refinement


are necessary during the development process.
The dynamic simulation of a complete valve train is
performed using ADAMS/Engine software, version
12. The software consists of three modules: The first
is ADAMS/Engine template builder, which is for
building or selecting templates of valve train
mechanisms. The second module is ADAMS/Engine
standard, which consists of the assembly modeler
and solver. The third module is ADAMS/Post
processor, which is a postprocessor code for
showing the results.
ADAMS/Engine uses a flexible elemental approach
to model valve trains. Within this approach
individual parts of the valve train are represented by
modeling elements, which can be related together to
form the whole valve train model. The software
models each valve train component, having stiffness
and damping. The excitation to the model is
provided by a rotating element. Then the equations
of motion are solved simultaneously for subsequent
elements in the model.
The models of the valve train components are
developed using the representative 3D elements in
the ADAMS/Engine archive. The detailed model of
the single and complete valve train studied in this
paper are shown in figures (1) and (2). The
geometric and material data were taken from the
component engineering drawings. The stiffness data
was collected by carrying out the CAE analysis of
the valve train components. The 3-D models of the
components were made in CATIA.
The stiffness of each component is calculated as
follows. The camshaft mounting parameters are
calculated using beam-bending formulas to obtain
the stiffness model. Using the same assumptions, an
effective mass is also calculated. The stiffness is
combined with estimated bearing stiffness from
practical recommendations. The tappet stiffness is
calculated by finite element analysis for a range of
offset loads. The valve stiffness is calculated using
classical formulas for axial load on a bar and dishing
of the valve head. The valve is divided into two
parts at the position of the centre of gravity. The
stiffness of the valve head end is also combined with
the valve seat insert stiffness specified by industrial
guidelines. The spring data is calculated from the
measured characteristic after calculating the
relationship between spring rate and number of
active coils.
The damping values for the component are assumed
to be 3-5% of critical damping base on the
guidelines for relative damping factors given in

14th Annual (International) Mechanical Engineering Conference May 2006


Isfahan University of Technology, Isfahan, Iran

reference [3-4]. The damper coefficients are


preliminarily established based on previous models
of the valve train developed in references [5, 6].
The valve train parameter calculations are made
easier by coding them as spreadsheets. All the
stiffness parameters calculations are summarized in
table (1).
SIMULATION RESULTS
The performance of the valve train becomes critical
at higher engine speeds, particularly in relation to
the dynamic valve lift, velocity and acceleration.
The displacement, velocity and acceleration curves
for the valve are shown in figures (3-6) for two
different speeds of 2000 rpm and 2500 rpm.
It can be observed from Fig.4 that the dynamic valve
lift is comparatively lower than the kinematics lift.
This can be attributed to the flexural rigidity of the
component and individual component damping.
This phenomenon has an adverse effect on the
effective flow area and consequently leads to
reduction in the volumetric efficiency of the engine.
Breathing of the engine can be made effective by
increasing the dynamic valve lift. The dynamic
valve lift can be increased by reduction in the energy
losses in the valve train or by increasing the forces
acting on the valve to the limit of durability. For a
given valve train the energy losses can not be
reduced below a certain limit constrained by
material and friction. At this juncture it becomes
very important to check the valve train for the
maximum dynamic forces. A lag in the valve
displacement is observed with respect to the cam at
the start of lift. This is due to the inertia effect of
tappet, pushrod, rocker, valve and the spring preload. The valve velocity plot shows a usual trend of
reversal of velocity at the point of maximum valve
lift.
In valve train performance the valve acceleration at
over-speed gives information of the occurrence of
the shock phenomenon. A steep acceleration curve
signifies the probability of shock phenomenon,
which is one of the causes of valve breakage and
detrimental from the point of view of airflow into
the cylinder. The acceleration curve for the
simulated valve train does not show any probability
of shock. It is observed that there is abrupt change in
the acceleration during valve closure. This leads to a
minor shock at the over-speed, but that is not the
case at the rated speed.
Other results about this model are contact force
between cam and tappet, contact force between
tappet and valve and contact force between valve
and seat. Figures (7-9) show these forces and stress.

www.me-en.com
mechanical engineering

The dynamic forces on the valve follows the same


trend as that of valve acceleration and the maximum
value of the dynamic force acting encountered
during over-speed is well below the first mode
buckling load. The buckling load value is
established by carrying out finite element analysis of
the valve.

EXPERIMENTAL MEASUREMENTS
In order to be able to verify the model predictions of
the valve train dynamics an experimental
measurement is conducted using a test bench shown
in figure (10) designed and developed at the MD
Laboratory of the IPCO in co-operation with Iran
Khodro.
The test system purpose is to reproduce the actual
engine functional conditions. It can operate for
different cylinder head types, at high camshaft speed
and under high lubrication oil temperatures; it
adopts the same power belt transmission used on the
engine. In order to measure high valve
displacements and velocity, it was used a high speed
differential laser vibrometer AS SHOWN IN
FIGURE (11). The laser vibrometer made and
model are Keyence High Speed Vibrometer (LC2450). The laser vibrometer measuring distance
range is -8mm to +8mm and its operation
displacement is 50 mm.
In order to rotate the cam shaft an electromotor with
4.4 KW power and 3000 rpm velocity is used as
shown in figure (12). Next an inventor is used for
changing cam shaft velocity from 0 to 2500 rpm.
This velocity is half of crank velocity. To eliminate
raw vibrations of cylinder head support, as reference
plane was selected an area close to the valve. The
valve acceleration is not directly measured, but
obtained by means of numerical derivative. A
schematic view of the measurement set up is shown
in figure (13).
Figure (14) shows the predicted results and
corresponding observed behavior of the valve train.
A good agreement between ADAMS simulations
and experimental results are observed. The
difference between these two set results is 0.005
mm. This shows the developed procedure can be
used as a helpful tool in the analysis and design of
valve train.

CONCLUSION
A new approach to quantify the dynamic forces
acting on the valve train components focusing on the
valve train performance within the frame of
durability limits has been established. Based on the

14th Annual (International) Mechanical Engineering Conference May 2006


Isfahan University of Technology, Isfahan, Iran

obtained results it is observed that there is no loss of


contact between valve train components within the
IC engine operation speeds. The simulation results
are verified against measured response of the
system. Predictions of the model are in good
agreement with the durability testing done on engine
dynamometer.
REFERENCES
1- S. Uehara, L.R.Padovese, Dynamical Study of a
Cam-Follower System. Computational methods
in Engineering, Caixa Postal 61548-CEP,1999.
2- G. Dalpiaz and A. Rivola, A Model for the
Elastodynamic Analysis of a Desmodromic
Valve Train. Proceedings of the Tenth World
Congress on the Theory of Machines and
Mechanisms, Vol.4,(1999), pp. 1534-1542.
3- Turkish, Relationship of Valve Spring Design
to Valve Gear Dynamics and Hydraulic Lifter
Pump up. SAE conferece, Detroit, March 1953.
4- J.H. Nourse, R.C. Dennis and W.M.Wood,
Recent Development in Cam design. SAE
Summer Meeting, Chigago , June 1960.
5- A.P.Pisano, F.Freudenstein., An Experimental
and Analytical Investigation of the Dynamic
Response of a High-Speed Cam-Follower
System.
Part
2:
A
Combined,
Lumped/Distributed Dynamic Model, Journal
of Mechanisms, Transmissions, and Automation
in Design, Vol. 105,(1983),pp.699-704.
6- A.P.Pisano, F.Freudenstein., An Experimental
and Analytical Investigation of the Dynamic
Response of a High-Speed Cam-Follower
System. Part 1: Experimental Investigation,
Journal of Mechanisms, Transmissions, and
Automation in Design, Vol. 105,(1983),pp.692698.

Table.1. valve train component stiffness


Component
Stiffness [N/mm]
Tappet
7.2e5
Valve Spring
40
Intake valve
9.2e4
Exhaust valve
8.6e4
Cam-Tappet Contact Stiffness

www.me-en.com
mechanical engineering

Seat-Valve Contact Stiffness

1e5

Fig 1- single OHC valve train model

Fig 2- Complete OHC valve train model

Fig. 3- The valve lift in 2000 & 2500 rpm

14th Annual (International) Mechanical Engineering Conference May 2006


Isfahan University of Technology, Isfahan, Iran

Fig. 4- The valve bounce in 2000 & 2500 rpm

Fig. 5- The valve velocity in 2000 & 2500 rpm

Fig. 7- The contact force between cam & tappet

Fig. 8-The contact force between tappet & valve

Fig. 9-The contact force between valve & seat

Fig. 6- The valve acceleration in 2000 & 2500 rpm

www.me-en.com
mechanical engineering

14th Annual (International) Mechanical Engineering Conference May 2006


Isfahan University of Technology, Isfahan, Iran

Fig. 13- The measurement set-up

Fig. 10- The test bench.

Fig. 14- The valve lift obtained by ADAMS/ from TEST

Fig. 11- Installation of LDV

Fig. 12- The test bench

www.me-en.com
mechanical engineering