Article citation info:

STOECK, T. Simplification of the procedure for testing common rail fuel injectors. Combustion Engines. 2020, 180(1), 52-56.
https://doi.org/10.19206/CE-2020-109

Tomasz STOECK CE-2020-109

Simplification of the procedure for testing common rail fuel injectors

The paper presents a simplified methodology for generating the characteristic curve of fuel doses for common rail injectors, which
consists in limiting the number of measurements on the test bench and calculating missing data using predefined (array) functions of the
Microsoft Excel spreadsheet. This allows checking the method of fuel delivery in a wide spectrum of predefined pressures and atomiser
opening times, while reducing the arduousness and time-consumption of the active experiment phase. The proposed solution is
particularly useful in problematic situations when standard manufacturer's tests, referred only to selected work points, make it
impossible to clearly assess the technical condition of the injector.
Key words: common rail injector, fuel doses characteristic curve, simplification of test procedures

1. Introduction
All tests of fuel injection apparatus requiring the dis-
mantling of its actuators from the engine belong to invasive
methods [13]. The most reliable and precise way to diag-
nose fuel injectors are test-bench tests, which are performed
on dedicated test benches [11]. They are universal bench
devices equipped with single or multiple measuring chain
systems. They can be compact in design or modular, ena-
bling the installation of equipment compatible with the
capabilities and the profile of operations of a given car
workshop or service company. The benches have all the
necessary adapters and connectors, both for fixing the fuel
injector and connecting the hydraulic and electric lines. This
allows not only for diagnostic tests, but also for thermochem-
ical flushing, i.e. the process of internal cleaning under high
pressure and at elevated detergent temperature [12].
Standard tests of the common rail fuel injector are per-
formed in automatic cycles or, much less often, with manu-
al settings. Usually they include several operating points for
different pressures of the supplied fuel and atomiser open-
ing (actuation) times, according to the manufacturer's re-
commendations [8]. Unfortunately, many years of work-
shop and laboratory practice indicates that such measure-
ments may prove to be insufficient, primarily due to im-
proper fuel dosing in other areas of engine operation. De-
tection of these irregularities is possible, but after preparing
the full characteristics of the fuel doses. Despite the un-
doubted benefits, this function is rarely used, only in justi-
fied cases, which results from the arduousness and time-
consumption of the test phase. As a result, the process of
checking a set from one engine becomes problematic and
economically unjustified. Therefore, an own methodology
was proposed, the idea of which is based on limiting the
number of measurements to the necessary minimum and
calculating the remaining data in the environment of a pop-
ular spreadsheet.
2. Methods
2.1. Test object and test beds
The test was carried out on the example of a Bosch
CRI1 electromagnetic fuel injector (Fig. 1), which was
dismantled from a Fiat 1.3 JTD 16V MultiJet diesel engine
with an operating mileage of 158,000 km. Fig. 1. CRI1 Bosch fuel injector design

52 Received 30 January 2020; Received in revised form 20 February 2020; Accepted 20 February 2020
Available online 30 March 2020
This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/)
 Simplification of the procedure for testing common rail fuel injectors

Fuel injectors of this type are commonly used in the entered into a Microsoft Excel spreadsheet, after which the
combustion engines of passenger cars, working at maxi- missing data were estimated in indirect points.
mum fuel injection pressures of up to 140 MPa [16]. Be-
2.2. Predefined (array) functions
cause the manufacturer has made available the technology
For the calculation of fuel injector fuel doses d that were
for their diagnostics, while offering a set of original spare
not included in the experiment phase, predefined functions:
parts, they can be regenerated to almost full extent [14].
LINEST and TREND, were used. The first of them allows
you to fit a straight line to a set of points, but it can be suc-
cessfully used to determine the coefficients in the multiple
regression equation in the form of a polynomial. For this
purpose, a column of t-values raised to the appropriate
power is prepared for each word, or array operations are
used, but the order of the exponents should be reversed by
creating the following formula:
= SUM(LINEST(known_d`s; known_t`s^{1\2} ∗
t^{2\1\0}) (1)
On the other hand, the TREND function is used directly
for data interpolation (e.g. polynomial, rational), hence the
following expression is used to calculate its value at any
selected point of the discrete interval:
= SUM(TREND(known_d`s; known_t`s^{1\2});
Fig. 2. Test bench STPiW-3 [17] t^{1\2}) (2)
In the next step, the addresses of the cells in which the
results of the measurements were entered are substituted,
i.e. the ranges of data known_t`s and known_d`s, as well as
the value of the sought indirect argument t.
Formulas (1) and (2) can be used interchangeably be-
cause they offer the same quality of approximation [5]. In
addition, when dividing a discrete area into two parts, low
degree polynomials are used to create the so-called spline
[6]. In this way, the risk of interference and oscillation
(local extremes) that could occur with a single interpolation
function [4] is eliminated.
3. Analysis results and discussion
3.1. Preliminary tests
Before starting the tests on the STPiW-3 test bench, the
fuel injector was dismantled into its components, which
were subjected to baths in ultrasonic baths and thoroughly
Fig. 3. Microscope Meiji FL150/70 [19] dried.
During the microscopic examination, traces of initial
The following systems and instrumentation were used in frictional wear of the surface of plunger and barrel assem-
the repair process: blies were found, which due to increased resistance limited
– STPiW-3 test bench (Fig. 2), the dynamics of the operation of the needle with the atom-
– 3-phase Bosch gear (CRR 120 pneumatic generator and iser (Fig. 4). This negatively affects the fuel delivery pro-
CRR 220 electric generator, CRR 420 digital sensor, cess and uneven engine operation, in particular at idle or
LAB/SM135 power supply, CRR 320 torque wrench), when setting low loads [7]. This type of defect is found
– Meiji FL150/70 microscope (Fig. 3), very often and is detected even at low operational mileage
– MIC-40700 multimeter, [1]. In addition, valve seat deformation (Fig. 5) was ob-
– ultrasonic baths (Carbon Tech Ultrasonic Bath S15/C2, served, which leads to an overestimation of the volume of
Elmasonic S10H), fuel going to the overflow. The occurrence of this malfunc-
– vices and gear required for the disassembly and assem- tion results from various reasons, including the wrong ball
bly phase. travel [22]. However, this parameter was checked after the
In order to prepare a full characteristics of the fuel doses plunger and barrel assembly, guide piston and control valve
in a simplified way, measurements on the test bench were assembly were replaced. Both in the mechanical AH (Ger.
carried out at the predefined actuation times (main nodes) t Ankerhub) and electric AHe (Ger. Ankerhub elektrisch)
[μs]: 200, 300, 600, 1200. The obtained test results were tests, the result was 0.051 mm, which is within the manu-
facturer's acceptable range (0.0030–0.060 mm) [18]. Then

COMBUSTION ENGINES, 2020, 180(1) 53

Simplification of the procedure for testing common rail fuel injectors

the fuel injector was mounted on the STPiW-3 bench, the predefined functions are much more flexible than the
where it successfully passed the approval procedures, in- classic interpolation methods that are being implemented in
cluding leakage and electrical tests. a spreadsheet environment [9]. Hence for t = 1000 μs we
obtained:
52.4 = SUM(LINEST(36.6: 60.3; 600: 1200^{1} ∗
1000^{1\0}) (5)
or
52.4 = SUM(TREND(36.6: 60.3; 600: 1200^{1});
1000^{1}) (6)
For obvious reasons, conducting two-way calculations
is not required, but it does not pose any major difficulties
and allows you to control the results obtained at every stage
of this process. The correctness of the formulas can also be
checked in another way, namely by substituting any select-
ed root node. In this case, the value of the interpolating
function must be equal to the measurement result at this
point [3]. For example, for t = 1200 μs, we obtained:
Fig. 4. Friction wear traces on the needle
𝟔𝟎. 𝟑 = SUM(LINEST(36.6: 𝟔𝟎. 𝟑; 600: 1200^{1} ∗
1200^{1\0}) (7)
or
𝟔𝟎. 𝟑 = SUM(TREND(36.6: 𝟔𝟎. 𝟑; 600: 1200^{1});
1200^{1}) (8)
Figure 6 shows the characteristic curve of fuel doses for
the tested injector d = f(t), which were generated based on
all results. By contrast, the measured points are marked in
colour, and the calculated points are left without any fill.
The shape and course of individual curves should be as-
sessed positively because they do not overlap or intersect
the entire working area in question [20]. This indicates that
there are no irregularities in the fuel delivery process [2]. In
addition, the implementation of pre-defined functions sim-
Fig. 5. Valve seat deformation plified the active experiment phase and enabled the estima-
tion of correct fuel doses values at any atomiser opening
3.2. Main tests time. For example, according to Bosch diagnostic coordi-
Table 1 presents the results of measurements carried out nates, the pilot dose VE (Ger. Voreinspritzung) and idling
on the test bench. The data obtained in the calculation pro- LL (Ger. Leerlauf) should be within 0.3–4.1 mm3/inj at
cess is also specified, the course of which is discussed on t = 260 μs, pinj = 80 MPa and 0.3–3.9 mm3/inj at t = 420 μs,
the example of selected fuel doses at fuel injection pressure pinj = 30 MPa [15]. In both cases these requirements were
pinj = 140 MPa. met because the values dVE = 3.9 mm3/inj and dLL = 2.3
The introduction of cell addresses in formulas (1) and mm3/inj were obtained.
(2) allowed to estimate the value of interpolation polynomi- It is worth emphasising that the detection of irregularities
al for the first indirect argument t = 250 μs: in this area would require the dismantling of the fuel injector
7.6 = SUM(LINEST(5.8: 36.6; 200: 600^{1\2} ∗ again and fuel delivery correction. In addition to adjusting
the ball travel, the air gap between the control unit disc and
250^{2\1\0}) (3) the coil body, the needle travel, and the valve spring force
or should be checked. The steps included in respective stages
regulation match the standard Bosch procedures.
7.6 = SUM(TREND(5.8: 36.6; 200: 600^{1\2});
4. Conclusions
250^{1\2}) (4) Despite the variety of diagnostic methods, more and
The fuel doses for the remaining atomiser opening times more often directed at conducting measurements under the
were calculated in a similar way. However, in the case of conditions of real operation on the engine, the basis of
the second discrete interval, the formula was simplified due modern research laboratories are test benches. They enable
to the smaller number of measuring points. In this aspect, a comprehensive assessment of common rail fuel injectors,

54 COMBUSTION ENGINES, 2020, 180(1)

 Simplification of the procedure for testing common rail fuel injectors

Table 1. Results of test-bench test and calculations for CRI1 Bosch injector
Injection pressure, Fuel dose
pinj [MPa] d [mm3/inj.]
Lp. Actuation
time, 140 120 100 80 60 40 30
t [μs]
1 200 5,8 5,2 4,5 3,1 1,7 0,2 0,1
2* 250 7,6 6,2 5,1 3,8 2,4 0,8 0,4
3 300 9,9 7,8 6,4 4,9 3,5 1,5 0,8
4* 400 16,4 13,1 11,0 8,7 6,6 3,5 2,0
5 600 36,6 32,0 28,3 22,2 16,8 9,7 5,7
6* 800 44,5 40,2 36,1 29,7 24,1 16,2 11,9
7* 1000 52,4 48,5 44,9 37,2 31,3 22,6 18,2
8 1200 60,3 56,7 53,2 44,7 38,6 29,1 24,4
* analytically estimated data

d [mm3/injection]
70

65 140 MPa 120 MPa 100 MPa 80 MPa 60 MPa 40 MPa 30 MPa

60

55

50

45

40

35

30

25

20

15

10

0
200
150

250

300

350

400

450

500

550

600

650

700

750

800

850

900

950

1000

1050

1100

1150

1200

1250
t [μs]

Fig. 6. Fuel doses characteristic curve for the injector tested

indicating at the same time areas of their incorrect operation view of their effect on the final results is obtained, which,
[10]. Therefore, improving test procedures seems to be the with a wide set of implemented predefined functions, gives
most reasonable, taking into account technical criteria (ac- a very useful tool for processing and presenting data. In the
curacy, repeatability, durability), as well as economic ones analysed case, two of them were used, simplifying the ac-
(cost, efficiency, labour consumption). The proposed meth- tive experiment phase by eliminating 28 measuring points.
odology can be useful to meet them, as it has been used in In addition, calculations in the digital environment did not
laboratory and workshop practice. require the presentation of the final form (formulas) of
One of the most important advantages of a spreadsheet interpolation polynomials, but only an estimate of their
is the ability to automatically recalculate all formulas after values for each indirect argument.
changing the input variables [21]. In this way, a quick pre-

Nomenclature
AH mechanical measurement of ball travel (Ger. An- d fuel dose
kerhub) CRI1 first generation common rail injector made by Bosch
AHe electrical measurement of ball travel (Ger. Ankerhub JTD multiJet Turbo Diesel
elektrisch) LINEST predefined linear regression function

COMBUSTION ENGINES, 2020, 180(1) 55

Simplification of the procedure for testing common rail fuel injectors

LL idling dose (Ger. Leerlauf) TREND predefined interpolating function

pinj injection pressure VE pilot dose (Ger. Voreinspritzung)
t actuation (atomiser opening) time

Bibliography
[1] ABRAMEK, K.F., STOECK, T., OSIPOWICZ T. Statistical tion methods. Combustion Engines. 2017, 170(3), 49-56.
Evaluation of the corrosive wear of fuel injector elements https://doi.org/10.19206/CE-2017-308
used in common rail systems. Strojniški vestnik – Journal of [13] OLSZOWSKI, S., CHOJNACKI, T., OLSZOWSKI, T.
Mechanical Engineering. 2015, 61(2), 91-98. Nieprawidłowa praca wtryskiwaczy common rail pomimo
https://doi.org/10.5545/sv-jme.2014.1687 pozytywnego wyniku badania. Autobusy – Technika, Ek-
[2] BUSZ, W., WALASZYK, A. Optimize the testing process sploatacja, Systemy Transportowe. 2017, 6, 1003-1007.
common rail fuel injectors. Combustion Engines. 2015, [14] OSIPOWICZ, T., ABRAMEK, K.F. Diagnosing methods
162(3), 978-981. common rail fuel injectors. Combustion Engines. 2017,
[3] CHENEY, W., KINCAID, D., Numerical Mathematics and 168(1), 56-61. https://doi.org/10.19206/CE-2017-109
Computing (7th Ed.). Books/Cole Publishing Company. [15] OSIPOWICZ, T., KOWALEK, S. Evaluation of Modern
2012, Pacific Grove. Diesel Engine Fuel Injectors. TEKA – Commission of Mo-
[4] GAUTSCHI, W. Interpolation before and after Lagrange. torization and Energetics in Agriculture. 2014, 14(3), 83-87.
Rendiconti del Seminario Matematico Università [16] REIF, K. (Ed.). Automotive Mechatronics. Bosch Profes-
e Politecnico di Torino. 2012, 70(4), 347-368. sional Automotive Information. Springer Fach-medien.
[5] GONET, M. Excel w obliczeniach naukowych 2015. https://doi.org/10.1007/978-3-658-03975-2
i technicznych. Wydawnictwo Helion. 2010, Gliwice. [17] STOECK, T. Application of the experimental design tech-
[6] HOFFMAN, J.D., Numerical methods for engineers and nique in fuel dose adjustment of common rail injector. Com-
scientists. Marcel Dekker, Inc. 2001, New York-Basel. bustion Engines. 2019, 179(4), 210-215.
[7] JUŚCIŃSKI, S., PIEKARSKI, W., CHOMIK, Z. Analysis https://doi.org/10.19206/CE-2019-435
of injection systems types used in agricultural machines. Ag- [18] STOECK, T., OSIPOWICZ, T. Analiza uszkodzeń wtryski-
ricultural Engineering. 2017, 21(4), 37-46. waczy Common Rail firmy Bosch stosowanych w silnikach
https://doi.org/10.1515/agriceng-2017-0034 wysokoprężnych pojazdów użytkowych. Logistyka. 2013, 6,
[8] KARPIUK, W., BOR, W., SMOLEC, R. Possibilities of 723-727.
analysis of condition and repair of common-rail system in- [19] STOECK, T., OSIPOWICZ, T. Przyczyny tworzenia się
jectors. Journal of KONES Powertrain and Transport. 2018, osadów we wtryskiwaczach Common Rail. Autobusy –
23(4), 209-216. Technika, Eksploatacja, Systemy Transportowe. 2016, 8,
https://doi.org/10.5604/12314005.1217208 139-141.
[9] KARRIS, S.T. Numerical analysis using MATLAB® and [20] VASILIU, D., VASILIU, G.C., RĂDULESCU, A. et al.
Excel® (3th Ed.). Orchard Publications. 2007, Berkeley. Predicting the dynamic performance of the fuel injectors by
[10] KLUCZYK, M., GRZĄDZIELA, A. Marine diesel engine numerical simulation. Scientific Bulletin. Series C: Electri-
common rail injectors monitoring with vibration parameters. cal engineering and Computer Science. 2018, 78(4), 173-
Diagnostyka. 2019, 19(3), 37-44. 188. University Politehnica of Bucharest.
https://doi.org/10.29354/diag/109793 [21] WALKENBACH, J. Excel 2013 Formulas. John Wiley &
[11] KLUCZYK, M., GRZĄDZIELA, A. Vibration diagnostics Sons, Inc. 2013, Hoboken, New Jersey.
of common rail injectors. Journal of Marine Engineering & [22] YORDANOV, N., HADJIEW, K., STANKOW, E. Experi-
Technology. 2017, 16(4), 177-184. mental simulation of common rail electromagnetic injectors
https://doi.org/10.1080/20464177.2017.1387088 wearing. Machines, Technologies, Materials. 2018, 12(4),
[12] KNEBA, Z., STRASZAK, P., JAKÓBCZYK, K. The effec- 208-211.
tiveness of fault detection in common rail injectors examina-

Tomasz Stoeck, DEng. – Faculty of Mechanical

Engineering and Mechatronics, West Pomeranian
University of Technology in Szczecin.
e-mail: [email protected]

56 COMBUSTION ENGINES, 2020, 180(1)