Available online at www.sciencedirect.

com

ScienceDirect
IERI Procedia 7 (2014) 84 – 89

2013 International Conference on Applied Computing, Computer Science, and Computer

Engineering

Diesel Engine PT Pump Fault Diagnosis Based on the

Characteristics of Its Fuel Pressure
Xin-jun Wang *, Yan-ping Cai, Xu-ze Lin
Xi’an High Technology Institution, Xi’an 710025, China

Abstract

In order to solve the problem that the diesel engine PT fuel system is unable to field maintain, developed a portable signal
acquisition and analysis system for diesel engine PT fuel system. Firstly, the PT pump work Principle was analyzed, and
the PT pump failure mapping relation between reason and failure phenomenon was analyzed; Secondly, the diesel engine
PT pump failure fuel pressure characteristics were analyzed; Lastly, using the portable signal acquisition and analysis
system to diagnose the diesel engine PT fuel system, experiment results show that the system can correctly detect the
diesel engine PT fuel system state.

© 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
© 2013 WANG Xin-jun, CAI Yan-ping, LIN Xu-ze. Published by Elsevier B.V.
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and peer review under responsibility of Information Engineering Research Institute
Selection and peer review under responsibility of Information Engineering Research Institute

Keywords: Diesel engine; fuel system; fuel pressure; PT pump; Fault Diagnosis

1. Introduction

Because of the simple structure and high fuel injection pressure, PT fuel system is widely used in the
upward 10L Cummins diesel engine, such as all kinds of mobile power station, ship power station[1-2].
However, because of its specific structure design and working process, PT fuel system is easy to occur sudden
failure and its fault type and position is difficult to judge[3-4]. Besides, diesel engine unit failure rate resulting

* Corresponding author. Tel.: +86 02984741102.

E-mail address: caiyanping502@163.com.

2212-6678 © 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and peer review under responsibility of Information Engineering Research Institute
doi:10.1016/j.ieri.2014.08.014
 Xin-jun Wang et al. / IERI Procedia 7 (2014) 84 – 89 85

directly from PT fuel system performance variation is as high as 60% or more[5], once failure, the diesel
engine can’t on-site repair.
At present, the oil line pressure, cylinder head vibration and instantaneous speed signal are main three
commonly used signals to reflect the state of the diesel engine fuel system. While PT fuel system is
completely different from the traditional plunger fuel system in structure and working principle, plunger-type
fuel system diagnosis methods which have been better application based on above methods, but no longer
applicable to the PT fuel system. For diesel engine on-site maintenance problem, a set of portable diesel
engine PT fuel system signal acquisition and analysis system was developed, and diesel engine PT pump fuel
pressure characteristics was analyzed, which laid a technological foundation for the realization of the PT
pump failure on-site maintenance.

2. PT pump signal acquisition system

PT fuel system is made from the fuel tank, float tank, filter, governor, PT fuel pump, injector and low-
pressure pipeline and return pipe. When working, PT fuel pump is responsible for oil regulator, PT injector
completes the tasks that generate high pressure and timing injection. PT fuel pump provide low-pressure fuel,
high-pressure fuel exists only in the fuel injector. Due to the lack of high-pressure oil line, the fuel is not easy
to leak which ensure the tightness of the PT fuel system. Connection of the various components is shown in
Fig 1.

Fig.1. PT fuel system composition and connection

The PT pump signal acquisition system includes pressure sensor, amplifier, rotational speed sensor and
signal acquisition instrument. Pressure of low-pressure oil is about 0.8~2.1Mpa, measured by 4045A20 sensor
and 4618A0 amplifier produced in Switzerland KISTLER. The sensor has a measurement range of 0~3MPa,
output amplifier range is about 0~10V, the magnification times is 5. Signal acquisition instrument is the
DASP systems developed by the Beijing oriental vibration and noise research institute. The sensor specific
installation location is shown in Fig 2.

Fig.2. the sensor specific installation location

Fig 3 is a top dead point markers schematic diagram. When Calibrating, the 1-6TC mark on the belt pulley
match the gear wheel canopy formed mark, on this moment, the 1 and 6 cylinders are at the top dead point
position. Point A is the 1 and 6 cylinders in the power stroke 90° indicating mark, point B is the 2 and 5
cylinders in the power stroke 90° indicating mark, and point C is the 3 and 4 cylinders in the power stroke 90°
indicating mark. The reflective light paper is pasted on the point A.
86 Xin-jun Wang et al. / IERI Procedia 7 (2014) 84 – 89

Fig.3. top dead point markers schematic diagram

Hardware connection diagram of acquisition system is shown in Fig 4. The deliver direction of the signal as
follows: oil line pressure signal is converted into a voltage signal by the sensor 4045A20 which is cascaded in
the diesel engine low-pressure oil line, then amplified by the amplifier 4618A0 and input into the DASP signal
acquisition instrument, finally computer finish processing, signal de-noising, feature extraction and the
judgment of the failure mode.
Diesel engine
Sensor Amplifier
low-pressure
4045A20 4618A0
oil line

DASP Multi-channel
Computer data acquisition
system

Fig.4. hardware connection diagram of acquisition system

3. Experimental Designs

Experimental diesel engine model is Cummins NTA855-G1, whose ignition sequence is 1-5-3-6-2-4 and
the load for the rated power is 200kW. Settable rotational speeds are idle speed (900r/min) and rated speed
(1500r/min). Settable load are no-load, 60kW, 120kW and 200kW four kinds. In the experiment, due to the
lack of a priori knowledge of the frequency distribution of the pressure signal, the larger sampling frequency
is taken as 25.6 kHz. Each failure was collected at different speed in different load group of 20 low-pressure
common rail hydraulic signals and the sampling time is 3s. Because faults can affect the system's load
capacity, maximum load can only be added to 120kW of certain failure with speed and load fluctuations.
In order to easily analyze the effect of pressure waveform from oil line failure, taking the recurring failure
and the feasibility of setting, non-destructive and economical efficiency into account, experiments simulate the
six states. Each fault setting method and the abbreviation is shown in Table 1.

Table 1. Each fault setting method

System Status Setting method Abbreviation

Normal state
Injector clogging
The injector seals broken
EFC actuators inflexibility
Oil line clogged
Composite failure
System to work properly, without failure Normal
Replacement Clogged injector spray holes with 1 cylinder injector Missing-cylinder
The replacement of the disabled sealing ring on 2 cylinder seals Damaged
The external power supply to control the EFC actuators open Inflexibility
Anti-magnetic filter installed Clogged
1 cylinder injector clogging and second ring broken Composites
 Xin-jun Wang et al. / IERI Procedia 7 (2014) 84 – 89 87

4. The failure fuel pressure characteristics analysis of PT pump

4.1. Time domain analysis

The higher the speed, the greater the load, pressure changes caused by the fault will be more obvious. Some
failure load can not be added into 200kW, so signals are chosen to be analyzed at the rated rotation speed and
under the load of 120kW. Experimental diesel engine NTA855-G1’s supply advance angle is 19° and
reflective light paper is attached to the 1 and 6 cylinder power stroke 90° indicating mark, where point A mark
is located. The diesel engine rotates crankshaft two turns to complete a work cycle, during the time two pulses
will be received by the rotational speed sensor in 0.08s, in the case of the speed 1500r/min.
Intercepting a variety of fault signals collected in experiment of each working cycle to analyze pressure
changes, intercept two working cycle to analyze periodicity of pressure fluctuations. Speed varies under
different faults, so each fault lasting time is also various in a work cycle. Time domain waveform of the
normal signal is shown in Fig 5.
5

4
U/V

1
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
(a) t/s

4
U/V

1
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
(b) t/s

Fig.5. normal signal time domain waveform

The time domain waveform of the missing cylinder signals is shown in Fig 6. According to rotational speed
signal, a work cycle lasts 0.08s. As can be seen from Fig 6(b), the pressure waveform has obvious periodicity.
Compared with the normal signal, the amplitude of the signal of the missing cylinder is larger. Comparing the
time domain waveform in one of the working cycle of the normal signal with the missing cylinder signal, there
is a distinct minimum between 0.07s and 0.08s in the normal signal, but it does not appear in the missing
cylinder signal. For 1 cylinder clogging, the time domain waveform appears much differently with normal
signal.
6

4
U/V

0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
(a) t/s

4
U/V

0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
(b) t/s

Fig.6. time domain waveform of missing cylinder signal

88 Xin-jun Wang et al. / IERI Procedia 7 (2014) 84 – 89

Time domain waveform of EFC actuator inflexibility signal is shown in Fig 7. According to speed signal, a
work cycle lasts 0.0803s. For given 120kW load, the actually measured load is 117kW, the rotational speed is
1494r/min. As the actuator shaft is not completely open, it affects the load capacity of the system. When the
load is 200kW, the actually measured load is only 155kW, speed is 1362r/min, and this time EFC actuators
inflexibility has a serious impact on the load capacity of the system.
5

U/V
3

1
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
(a) t/s

4
U/V
3

1
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18
(b) t/s

Fig.7. time domain waveform of EFC inflexibility signal

Time domain waveform of oil line clogged signal is shown in Fig 8. According to speed signal, a work
cycle lasts 0.08s. From the point of view of waveforms, compared with the normal signalˈwaveform changes
of oil line clogged signal is more complex. What can be seen from Fig 8(a) is 6 minimum of pressure, the
apparent periodicity also can be found from Fig 8(b).
5

4
U/V

1
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
(a) t/s

4
U/V

1
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
(b) t/s

Fig.8. time domain waveform of oil line clogged signal

4.2. Fault signal frequency domain analysis

The frequency distribution of each fault signal can be obtained by Fourier transform. Since all signal
frequency don not exceed 2000Hz, we take frequency domain distribution within 0~2000Hz band of each fault
signal, which is shown in Fig 9. In Fig 9, (a) is a normal signal; (b) is a composite signal; (c) is missing-
cylinder signal; (d) is EFC actuator inflexibility signal; (e) is injector seal ring broken signal; (f) is oil line
clogged signal. It can be seen that the energy of the signal is mainly concentrated in the frequency band of
0~1000Hz. So the fault characteristic contained in the signal can be extracted by analyzing the changes in the
distribution of the frequency of the band.
 Xin-jun Wang et al. / IERI Procedia 7 (2014) 84 – 89 89

Fig.9. Fault signals frequency distribution

5. Conclusions

In this paper, the link between the cause of the fault and the phenomenon of each fault is analyzed in the
time and frequency domain. From the frequency distribution of the signal it can be seen that the change in the
amplitude of the fault frequency is mainly concentrated in the frequency band of 0~1000Hz when the fuel
pressure is a low-frequency signal. The frequency distribution of the different failures differs greatly, which is
feasible from the frequency domain to distinguish failure modes.

Acknowledgements

The research in this paper was supported by National Natural Science Foundation of China (61201449) and
Shaanxi Natural Science Foundation of China (2013JQ8023).

References

[1] Zhang Chunrun,Ma Hongwen, Zhao Jinsong, fault detection method based on the vibration intensity of
Diesel Fuel Injection System off the oil [J]. Armored Force Engineering University, 2008, 22(4) :62-63.
[2] Badaoui M EI, Daniere J, Guillet F, et al. Separation of combustion noise and piston-slap in diesel engine-
part I: separation of combustion noise and piston-slap in diesel engine by cyclic wiener filtering [J].
Mechanical Systems and Signal Processing, 2005, 19:1209-1217.
[3] Liu XH, Randall R B. Blind source separation of internal combustion engine piston slap from other
measured vibration signals [J]. Mechanical Systems and Signal Processing, 2005, 19: 1196-1208.
[4] Sun Yunling, Park A Zhe, Zhang Yongxiang.The internal combustion engine fault diagnosis research
based on instantaneous speed and bispectrum [J]. CSICE, 2004,22 (3) :241-244.
[5] Liu Yanbin, Liu Jianmin, Qiao Xinyong, et al. Diesel engine fuel supply advance angle detector research
based on the vibration signal of [J]. CSICE, 2007,25 (6) :565-568.