Downloaded from SAE International by Essam Allam, Monday, January 18, 2016

2013-01-9041
Published 09/20/2013
Copyright © 2013 SAE International
doi:10.4271/2013-01-9041
saeeng.saejournals.org

Consequence Analysis in Predictive Health Monitoring of

Automotive Diesel Engine Defects
Shawki Abouel-Seoud, Mohamed Khalil, Sameh Metwalley, Essam Allam, and Hany Assad
Helwan Univ

ABSTRACT
Reliability has always been an important aspect in the assessment of industrial products and/or equipments. Good
product design is of course essential for products with high reliability. However, no matter how good the product design is,
products deteriorate over time since they are operating under certain stress or load in the real environment, often involving
randomness. Maintenance has, thus, been introduced as an efficient way to assure a satisfactory level of reliability during
the useful life of a physical asset. The earliest maintenance technique is basically breakdown maintenance (also called
unplanned maintenance, or run-to-failure maintenance), which takes place only at breakdowns. A later maintenance
technique is time-based preventive maintenance (also called planned maintenance), which sets a periodic interval to
perform preventive maintenance regardless of the health status of a physical asset. The vehicle component is judged to be
safe depending on its reliability. Consequently, there is a general attention to establish approach by which the failure risk
has to be known. To achieve this objective, the present work introduces an approach to establish the vehicle engine risk
assessment using the excess air escaped inside the engine crankcase. The results indicate that the excess scalped air mass
flow rate corresponding to the failure rate values computed based on the Weibull distribution with assured reliability can
be considered to be guide for maintenance regime.

KEYWORDS: Diagnostic, reliability, hazard rate, prognostic, failure, Weibull distribution, deterioration

CITATION: Abouel-Seoud, S., Khalil, M., Metwalley, S., Allam, E. et al., "Consequence Analysis in Predictive Health
Monitoring of Automotive Diesel Engine Defects," SAE Int. J. Engines 6(3):2013, doi:10.4271/2013-01-9041.
____________________________________

INTRODUCTION resource planning systems, etc., have been developed for data
storage and handling. Collection of event data usually
Data acquisition is a process of collecting and storing requires manual data entry to the information systems. With
useful data (information) from targeted physical assets for the the rapid development of computer and advanced sensor
purpose of Conditions Based Maintenance (CBM) where technologies, data acquisition facilities and technologies have
CBM is a maintenance program that recommends become more powerful and less expensive, making data
maintenance decisions based on the information collected acquisition for CBM implementation more affordable and
through condition monitoring. This process is an essential feasible. One point the authors would like to make is that
step in implementing a CBM program for machinery fault (or event data and condition monitoring data are equally
failure, which is usually caused by one or more machinery important in CBM [2].
faults) diagnostics and prognostics. Data collected in a CBM For condition monitoring data, data errors may be caused
program can be categorized into two main types: the so- by sensor faults. In this case, sensor fault isolation is the right
called event data and condition monitoring data. Event data way to go. In general, however, there is no simple way to
include the information on what happened (e.g., installation, clean data. Sometimes it requires manual examination of
breakdown, overhaul, etc., and what the causes were) and/or data. Graphical tools would be very helpful to finding and
what was done (e.g., minor repair, preventive maintenance, removing data errors. A variety of models, algorithms and
oil change, etc.) to the targeted physical asset [1]. tools are available in the literature to analyze data for better
Maintenance information systems, such as Computerized understanding and interpretation of data. The models,
Maintenance Management Systems (CMMS), enterprise

1521
 Downloaded from SAE International by Essam Allam, Monday, January 18, 2016

1522 Abouel-Seoud et al / SAE Int. J. Engines / Volume 6, Issue 3(September 2013)

algorithms and tools used for data analysis depend mainly on optimizing the life of critical gearbox components. Statistical-
the types of data collected [3] based anomaly detection algorithms, noise measurement
The most common multidimensional data are image data related faults and degradation can now be developed for real-
such as infrared thermographs, X-ray images, visual images, time monitoring environments. Integration and
etc. Data processing for waveform and multidimensional data implementation of these technologies presents a great
is also called signal processing. Various signal processing opportunity to significantly enhance current gearbox health
techniques have been developed to analyze and interpret monitoring capabilities and risk management practices.
waveform and multidimensional data to extract useful However, methods for analyzing system status and health and
information for further diagnostic and prognostic purpose [4]. for predicting system life expectancy need to be made more
The consequence analysis techniques discussed in [5-6] is powerful, insightful, reliable, and robust for data collection
a general risk evaluation tool for CBM optimization based on onboard systems in real time. A developed robust an
various kinds of criteria. The most obvious and widely used analytical method for predicting remaining lifetime of
prognostics is to predict how much time is left before a transmission gears is presented in [12]. An investigation has
failure occurs (or, one or more faults) given the current been introduced with the aim of machine prognosis which is
machine condition and past operation profile. The time left to provide decision support for maintenance actions. The
before observing a failure is usually called Remaining Useful main idea of prognostics incorporating maintenance policies
Life (RUL). Most of the papers in the literature of machine is to optimize the maintenance policies according to certain
prognostics discuss only the type of prognostics, namely criteria such as risk, cost, reliability and availability. Risk is
RUL estimation [7]. defined as the combination of probability and consequence.
A logistic regression model has been used to calculate the Usually, consequence can be measured by cost. Since cost
probability of failure for given condition variables and an criterion applies to most situations, it is not surprising that the
Auto Regressive Moving-Average (ARMA) time series literature in CBM optimization is dominated by cost-based
model to trend the condition variables for failure prediction. CBM optimization [13].
Then a predetermined level of failure probability was use to The aforementioned review has shown that almost all the
estimate the RUL. Moreover, to track sensor-level test-failure efforts which have been done in predictive health monitoring
probability vectors instead of the physical system or sensor was directed towards the studying prognostics and diagnostic
parameters for prognostics. A Kalman filter with an theoretical or experimental using vibration, acoustic and oil
associated interacting multiple models was used to perform analysis. However, the objective of the present work is to use
the tracking. Two statistical models in survival analysis, the vehicle engine cylinder's excess air escaped (i.e. blow-by
Prognostics and Health Management (PHM) and the gas) for predictive health monitoring which hitherto has been
Proportional Intensity Model (PIM), which adopts a given little consideration.
stochastic process setting and assumes a similar form to the
intensity function of the stochastic process, are useful tools PREDICTIVE HEALTH
for RUL estimation in combination with a trending model for MONITORING
the fault propagation process. In addition, PIM with covariate
extrapolation has to be applied to estimate bearing residual Diagnostic and Prognostic Methods
life. The Hidden Markov Model (HMM), is a statistical Many industrial systems exhibit increasing wear and tear
model that can be used to analyze data gathered from, among of equipment during operation. For example, an automobile
other things, mechanical components, is also a powerful tool has many pieces of equipment, such as the engine, gear box
for RUL estimation [8, 9, 10]. and valves that exhibit various types of performance
All automotive fleets' ownerships want to increase their degradation due to erosion, friction, internal damage and
fleets' availability; therefore, they resort to consider the cracks. Prognostics are viewed as an add on capabilities to
conventional maintenance policy. Despite of, this policy diagnosis; they assess the current health of a system and
increases fleets availability but the cost will be high. This is predict its remaining life based on features that capture the
due to replace the parts before the end of their lifetime. On gradual degradation in the operational capabilities of a
the other hand, this policy is based on time where it is not system. Prognostics are critical to improve safety, plan
cost-effective. Consequently, this conventional maintenance successful missions, schedule maintenance, reduce
policy is not adequate to fulfill the needs of high availability maintenance cost and down time. Unlike fault diagnosis,
for automotive fleets. The concept is based on predicting the prognosis is a relatively new area and became an important
system degradation and calculating the reliability function for part of Condition-Based Maintenance (CBM) of systems.
the component. An application carried out on internal
combustion engine crankshaft main bearing to illustrate the Hazard Rate Model Based Data-Driven
effectiveness of this technique [11].
Real-time, integrated health monitoring of gearboxes that
Technique
can detect, classify, and predict developing gearbox faults is Before modeling the hazard rate process of an engine
critical to reducing operating and maintenance costs while cylinder kit where the engine cylinder kit is the piston, piston
 Downloaded from SAE International by Essam Allam, Monday, January 18, 2016

Abouel-Seoud et al / SAE Int. J. Engines / Volume 6, Issue 3(September 2013) 1523

rings and the liner, it is crucial to design a suitable signal

indicator that captures the evolution of the cylinder kit's
degradation. The degradation signal that we considered is (5)
focused on the pressure inside the crankcase of the engine.
The reason for choosing this signal is the relation between the Based on Weibull distribution and the rotational vibration
pressure inside the crankcase during the operating condition acceleration data measured for a faulty gear at different
and the clearance between the cylinder kit parts. There is a operation conditions, the failure Weibull probability density
proportional relationship between the pressure inside the function is written as following:
crankcase and cylinder kit parts clearance. The degradation
measured of the engine cylinder kit can be calculated by
using the hazard function. The probability of failure for a (6)
system in a given time interval [x1, x2] can be expressed in
From equation.(1), then
terms of either the unreliability function as [14, 15, 16]

(7)
(1)
From equations (3.3) and (3.4), the hazard given by
or in terms of the reliability function

(8)
(2)

Hazard (also called hazard rate or failure rate) is the

probability of an item failing at any given instance. Hazard
may change in time as result of many factors. The
maintenance cost depends on the probability distribution Individual hazard rate as a mean of fault
function (pdf), if the probability distribution function is diagnosis
described mathematically by f(x), then the cumulative To obtain the remaining traveling distance (RTD) of the
distribution function F(x) can drive by continuous integration operated faulty engine, the decision maker has to use
as the following: equation (8) with respect to a specific traveling distance
though calculated hazard rate.
For air mass flow rate, equation (8) become
x=q
(3)
Whereas the sum of the reliability and the cumulative
distribution function equal one then the equation can be
written as (9)

(4)

Failure rate as a mean of prognostic based on air

mass flow rate
To obtain the failure rate of the operated faulty engine, the
decision maker has to use equation (8) with respect to the air
mass flow rate though calculated hazard rate. The failure air
mass flow rate will be the air mass flow rate corresponding to
the hazard rate of 1.0 (100% the reliability function).
 Downloaded from SAE International by Essam Allam, Monday, January 18, 2016

1524 Abouel-Seoud et al / SAE Int. J. Engines / Volume 6, Issue 3(September 2013)

For air mass flow rate, equation (8) become predicted with certainty, then the engine cylinder kit
Put x = Q components will be replaced before this wear-out phase
begins. The hazard function is taken to be a one parameter
from the life time distribution that is obtained it through
statistical life that is usually conducted under ideal laboratory
(10)
condition. The life time's data of any mechanical component
may also be collected during the operating to estimate the
lifetime distribution. Therefore statistical procedures in
estimating the lifetime distribution parameters and decisions
related with replacement and distributions.
Failure rate as a mean of prognostic based on
vehicle driving distance Survey of the Vehicle Fleet Faulty Engines
To obtain the vehicle driving distance (VDD) of the A total of 30 vehicles were considered in the survey,
operated faulty engine, the decision maker has to use where 2 operate up to 100000 km, 6 operate up to 200000
equation (8) with respect to the driving distance though km, 16 operate up to 300000 km, 4 operate up to 400000 km
calculated hazard rate. The failure driving distance will be the and 2 operate up to 500000 km as shown in Fig. 2. All
driving distance corresponding to the hazard rate of 1.0 installed engines in these vehicles have the same technical
(100% the reliability function). specifications. The engine technical specifications is
For vehicle driving distance, equation (8) become presented in the Table 1
Put x = D
EXPERIMENTAL WORK AND
INSTRUMENTATION SYSTEM
(11)
Description of Experimental Method
Figure 3 illustrates the layout of diagnostic and prognostic
processes which indicates the path of the blow-by gases for
an engine equipped with Positive Crankcase Ventilation
BACKGROUND KNOEWLEDGES (PCV) system. The PCV system operates by the engine
vacuum at the intake manifold to create air circulation.
General Consequently, this circulation will help to remove the
Reliability is the probability that an item will perform unwanted vapor mixtures, which contains burned and
satisfactorily for a specified period of time or kilometer under partially burned combustion gases. It is well known that the
specified operating conditions. In the field of reliability, there PCV system creates vacuum inside the crankcase and this
are many types for life time distributions. Some of these vacuum is directly proportional to the engine speed. If any
types are normal and Weibull distribution. The reliability component from the PCV system has defect such as a hidden
function of any mechanical component is obtained through crack in a hose connection between the crankcase ventilation
statistical life testing procedures that are usually conducted and intake manifold, or the PCV valve has slightly defect
under ideal laboratory condition. On the other hand, the life then the taken measurement will not be accurate. Therefore,
time's data can also be collected during the components the effect of any defective parts of PCV system should be
operation to estimate it's distribution parameters and avoided in the measurements by removing the whole PCV
decisions. There has been growing interest in the recent years system during the tests and reinstalled it later. On the other
in reliability and maintenance models where the main hand, during the combustion process inside the vehicle
emphasis is placed on the so called intrinsic age rather than engine cylinders, there is some gas flow inside the engine
its real age. The hazard rate is changed over the life time for crankcase. The quantities of these gases are proportional to
any mechanical components as shown in Fig. 1. The figure the vehicle engine cylinder condition or the clearance
explains the engine main bearing life time, as an example between the cylinder kit components.
[11], the first interval of time from (t0) to (t1), represents
Description of Instrumentations Tools
reducing the hazard rate due to surface roughness between
the main bearing and shaft. The second interval of time, from Sensors and sensing strategies constitute the foundational
(t1) to (t2) represents the operating normal conditions where basis for fault diagnosis and prognosis. Strategic issues
arising with sensor suites employed to collect data that
the hazard rate is constant approximately. On the other hand,
eventually will lead to online realization of diagnostic and
the portion of the curve beyond t2 represents the wear-out
prognostic algorithms which are associated with the type,
failure or increasing the hazard rate due to increase the number, and location of sensors; their size, weight, cost,
clearance between the engine cylinder kit components dynamic range, and other characteristic properties; whether
(piston, cylinder liner, and piston rings), if time t2 could be
 Downloaded from SAE International by Essam Allam, Monday, January 18, 2016

Abouel-Seoud et al / SAE Int. J. Engines / Volume 6, Issue 3(September 2013) 1525

they are of the wired or wireless variety; etc. Data collected out the predictive maintenance regime and consequently to
by transuding devices rarely are useful in their raw form. save money and look promising.
Such data must be processed appropriately so that useful
information may be extracted that is a reduced version of the Prognostic Results
original data but preserves as much as possible those
characteristic features or fault indicators that are indicative of
Vehicle engine failure rate
the fault events to detect, isolate, and predict the time Table 4 presents the chosen 6 vehicles engines to be used
evolution. Furthermore, the sensor providing the data must be in the prognostic process, the values of the shape and scale
validated; that is, the sensors themselves are not subjected to parameters of Weibull distribution for used engines in this
fault conditions. Fig. 4 shows a photograph of hot-wire air paper were obtained from analysis of the collected data about
mass sensor, while Fig. 5 shows the interior details of the this engine. These data that were collected from the Agent is
sensor. considered a random sample from the failure engines from
this type and the age of these engines by Kilometers at
Test Procedure and Calculations failure. After collection these data and by using the Weibull-
In this study, the experimental work was carried out on all Ease14.0 software, we can obtain the shape and scale
the vehicles chosen. Kilometers reading are collected from parameters. The value of the shape parameter β is equal to 3.5
speedometer. The sensor was used to calculate the quantity of and the scale parameter η is equal to 393750 km (Driving
loss air escaped from the clearance between the engine Distance) or 7.795 g/s (Air mass flow rate) by using the Fig.
cylinder and piston rings. The air mass flow rate sensor is 7. After measuring the blow-by gas for any engine we can
connected to the engine oil pan output of engine breather and determine the engine failure rate. The next step we go to the
when the engine starts and the air breathing exists from the graph that illustrates the relation between the engine failure
engine and pass through the sensor hot wire. The results are rates with kilometers to predict the remaining driving
displayed in terms of Voltage output, where the air mass flow distance for this engine.
rate is obtained from the calibration curve of the sensor. The Figures 8, 9, 10, 11, 12, 13 show the measured values for
measurements were taken when the vehicle engine is in idle blow-by gas for these engines by air mass flow sensor to use
condition at engine speed of 750 rpm. Table 2 tabulates the them in the following step that the values of the remaining
values of measured vehicle actual driving distance and engine driving distance of these engines according to (equation 10).
air mass flow rate based on the arrangement shown in Fig. 3. Bearing in mind that the vehicle engines were tested are at
These data will be used to calculate the individual hazard rate idle condition with speed of 750 rpm. More observations can
in order to carrying out the fault diagnosis process based on be seen in Fig. 14
equation (9). Fig. 6 shows an example of this calculation. On Remaining driving distance
the other hand, these data will be also used for calculating the
Figures 15, 16, 17, 18, 19, 20 show the remaining driving
failure rate (hazard rate) and remaining life distance based on
distance, while in Table 5, the remaining driving distance for
equations (10) and (11) respectively.
the vehicle engines considered. These values are computed
RESULTS AND DISCUSSION based on the Weibull distribution with assured reliability. On
the other hand, the remaining driving distance for engine No.
Diagnostic Results 5 behaves the highest value followed by engine Nos. 7, 9, 11
and 12 with the lowest value for engine No. 13. These values
Table 2 tabulates the details of air mass flow rate, driving
display that engine No. 5 has the least risk assessment with
distance and calculated individual hazard rate for 16 vehicles
the highest for engine No. 13 (see Fig. 20. Bearing in mind
engines, where Fig. 7 shows the relationship between the
that the vehicle engines were tested are at idle condition with
measured air mass flow rate and the driving distance. Curve
speed of 750 rpm. More observations can be seen in Fig. 21.
fitting was carried out where the relation equation is y =
It can be seen that the remaining driving distance worked
118.11x - 526.95. Knowing the Weibull distribution the
effectively. It captured the system behavior exactly and gives
cumulative distribution function and the individual hazard
an alarm signal about possible irregularity before the vehicle
rate of the vehicle engine cylinder/piston ring components
engine actually broke. That is a very valuable indication for
can be calculated. The individual hazard rate and the vehicle
the vehicle engines health condition monitoring (diagnostic
engine number. In terms of individual hazard rate (equation
and prognostic).
9), it is clearly seen that the individual hazard rate varies from
vehicle engine to another according to both the driving CONCLUSIONS
distance and air mass flow rate. However, the significant
change around the value of new engine hazard rate indicates 1. Reasons that advanced maintenance technologies have not
the deterioration in condition, but it would only have been well implemented in industry might be: (1) lack of data
negligible influence on the average levels which is governed due to incorrect data collecting approach, or even no data
by the measured signal. This can be an effective way to carry collection and/or data storage at all; (2) lack of efficient
communication between theory developers and practitioners
 Downloaded from SAE International by Essam Allam, Monday, January 18, 2016

1526 Abouel-Seoud et al / SAE Int. J. Engines / Volume 6, Issue 3(September 2013)

in the area of reliability and maintenance; (3) lack of efficient 16. Rausand, M. and Hoyland, A. A “System reliability theory: Models and
statistic methods” John Wiley and Sons Inc., 2 edition, ISBN
validation approaches; (4) difficulty of implementation due to 0-471-47133-X, 2004.
frequent change of design, technologies, business policies
and management executives. Table 1. The Technical Specification of Engine
2. It is clearly observed that air mass flow rate corresponding
to the failure rate values computed based on the Weibull
distribution with assured reliability can be considered to be
guide for maintenance regime. On the other hand, the air
mass flow rate difference for engine No. 5 behaves the
highest value followed by engine Nos. 7, 9, 11 and 12 with
the lowest value for engine No. 13. These values display that Table 2. The chosen vehicles actual driving distance and
engine No. 5 has least risk assessment with highest for engine engines air mass flow rate measured at engine idle
No. 13. conditions
3. It can be seen that the remaining driving distance worked
effectively. It captured the system behavior exactly and gives
an alarm signal about possible irregularity before the vehicle
engine actually broke.

REFERENCES
1. Jardine, K.S., Lin, D. B. and Banjevic, D. “A review on machinery
diagnostics and prognostics implementing condition-based
maintenance” Mechanical Systems and Signal Processing 20, pp.
1483-1510, 2006.
2. Davies, C. and Greenough, R.M. “The use of information systems in
fault diagnosis” Proceedings of the 16th National Conference on
Manufacturing Research, University of East London, UK, 2000.
3. Xu, R. and Kwan, C. “Robust isolation of sensor failures” Asian Journal
of Control 5, pp. 12-23, 2003.
4. Wang, W.J. and McFadden, P.D. “Early detection of gear failure by
vibration analysis II. Interpretation of the time frequency distribution
using image processing techniques” Mechanical Systems and Signal
Processing 7, pp. 205-215, 1993.
5. Campbell, J.E., Thompson, B.M. and Swiler, L.P. “Consequence
analysis in predictive health monitoring systems” Proceedings of
Probabilistic Safety Assessment and Management, Vols. I and II,
Amsterdam, pp. 1353-1358, 2002
6. Castanier, B., Berenguer, C. and Grall, A. “A sequential condition-based
repair/replacement policy with non-periodic inspections for a system
subject to continuous wear” Applied Stochastic Models in Business and
Industry 19, pp. 327-347, 2003.
7. Farrar, C.R., Hemez, F., Park, G., Robertson, A.N., Sohn, H. and
Williams, T.O. “A coupled approach to developing damage prognosis
Solutions” Damage Assessment of Structures-The Fifth International
Conference on Damage Assessment of Structures (DAMAS 2003),
Southampton, UK, 2003.
8. Yan, J., Koc, M. and Lee, J. “A prognostic algorithm for machine
performance assessment and it's application” Production, Planning and
Control 15, pp. 796-801, 2004.
9. Phelps, E., Willett, P. and Kirubarajan, T. “A statistical approach to
prognostics, in: Component and Systems Diagnostics” Prognosis and
Health Management, Vol. 4389, Bellingham, pp. 23-34, 2001.
10. Vlok, P.-J., Wnek, M. and Zygmunt, M. “Utilizing statistical residual
life estimates of bearings to quantify the influence of preventive
maintenance actions, Mechanical Systems and Signal Processing 18, pp.
833-847, 2004.
11. Ahmed, I., Khalil, M., Allam, E., and Abouel-Seoud, S., “Remaining
lifetime Evaluation for Crankshaft Main Bearings,” SAE Technical
Paper 2008-01-1654, 2008, doi:10.4271/2008-01-1654.
12. Saad, A., “Predicting Remaining Lifetime of Transmission Gears,” SAE
Technical Paper 2010-01-0903, 2010, doi:10.4271/2010-01-0903.
13. Lugtigheid, D., Banjevic, D. A. K and Jardine, S. “Modeling repairable
system reliability with explanatory variables and repair and maintenance
actions” IMA Journal Management Mathematics 15, pp. 89-110, 2004.
14. Lin, D., Wiseman, M., Banjevic, D. and Jardine, A. K. S. “An approach
to signal processing and condition-based maintenance for gearboxes
subject to tooth failure” Mechanical Systems and Signal Processing 18,
pp. 993-1007, 2004.
15. Heng, A., Zhang, S., Tan, A.C.C. and Mathew, J. “Rotating machinery
prognostics: State of the art, challenges and opportunities” Mechanical
Systems and Signal Processing 23, pp. 724-739, 2009.
 Downloaded from SAE International by Essam Allam, Monday, January 18, 2016

Abouel-Seoud et al / SAE Int. J. Engines / Volume 6, Issue 3(September 2013) 1527

Table 3. Details of air mass flow rate, driving distance Table 5. The remaining driving distance of the vehicles
and calculated individual hazard rate engines

Table 4. The failure rate of the chosen vehicles engines Fig. 1. Air flow inside a vehicle engine crankcase
based on air mass flow rate

Fig. 2. Collected Survey of the vehicle fleet faulty

vehicles engines
 Downloaded from SAE International by Essam Allam, Monday, January 18, 2016

1528 Abouel-Seoud et al / SAE Int. J. Engines / Volume 6, Issue 3(September 2013)

Fig. 5. Interior details of air mass flow rate of the sensor

Fig. 3. The layout of diagnostic and prognostic processes

Fig. 6. Examples of hazard rate in digital form

(magnification value x 100)

Fig. 4. Photograph of five-wire air mass sensor

Fig. 7. The relationship between air mass flow rate and

the driving distance
 Downloaded from SAE International by Essam Allam, Monday, January 18, 2016

Abouel-Seoud et al / SAE Int. J. Engines / Volume 6, Issue 3(September 2013) 1529

Fig. 8. Failure rate Vs air mass flow rate for vehicle Fig. 11. Failure rate Vs air mass flow rate for vehicle
engine No. 5 engine No. 11

Fig. 9. Failure rate Vs air mass flow rate for vehicle Fig. 12. Failure rate Vs air mass flow rate for vehicle
engine No. 7 engine No. 12

Fig. 10. Failure rate Vs air mass flow rate for vehicle Fig. 13. Failure rate Vs air mass flow rate for vehicle
engine No. 9 engine No. 13
 Downloaded from SAE International by Essam Allam, Monday, January 18, 2016

1530 Abouel-Seoud et al / SAE Int. J. Engines / Volume 6, Issue 3(September 2013)

Fig. 14. Difference of air mass flow rate for different Fig. 17. Failure rate Vs vehicle driving distance for
engine number vehicle engine No. 9

Fig. 15. Failure rate Vs vehicle driving distance for Fig. 18. Failure rate Vs vehicle driving distance for
vehicle engine No. 5 vehicle engine No. 11

Fig. 16. Failure rate Vs vehicle driving distance for Fig. 19. Failure rate Vs vehicle driving distance for
vehicle engine No. 7 vehicle engine No. 12
 Downloaded from SAE International by Essam Allam, Monday, January 18, 2016

Abouel-Seoud et al / SAE Int. J. Engines / Volume 6, Issue 3(September 2013) 1531

Fig. 20. Failure rate Vs vehicle driving distance for

vehicle engine No. 13

Fig. 21. Remaining driving distance for different engine

number