ANALYTICAL METHODOLOGY FOR TESTING COMMON RAIL FUEL INJECTORS IN PROBLEMATIC CASES

The paper presents a new methodology for testing common rail fuel injectors, which extends the standard diagnostic procedure to include the analysis of the resultant fields of dosed fuel. The calculations were made on the basis of Gaussian formulas, also known as the shoelace formula. Their implementation in the digital environment was convenient from the practical side, as it eliminated the need to modify the test bench software. Thanks to the easy modification, the presented algorithms can be used in testing fuel injectors of other types or generations, for which clear assessment and verification of the technical condition is sometimes difficult. It is not required for this to increase the number of measuring points of the active experiment, which is one of the greatest advantages of the proposed solution.


INTRODUCTION
Common rail fuel injectors are the most susceptible to damage elements of the fuel systems in modern compression-ignition engines [5]. For this reason, intensive research works on the development of various diagnostic methods have been carried out for several years. This applies to both non-invasive techniques, allowing to locate a malfunctioning fuel injector without the need to remove it from the engine, and invasive techniques, thanks to which it is possible to identify the causes of failure and eliminate them in tests on test benches. In the latter group of techniques, standard manufacturer's tests are usually used, which are commonly regarded as the most precise and reliable [7]. They are carried out in automatic cycles, and mainly concern the method of fuel delivery at several critical operating points [6]. However, there are problematic cases in which such measurements may turn out to be insufficient, because the fuel injector doses fuel incorrectly in other areas of the engine operation, despite the positive test result. As a result, the regeneration process requires extended diagnostics. It is particularly troublesome with the so-called injection characteristics, generated with the full spectrum of operating parameters, as the time consumption of the experimental phase increases several times [12]. In this respect, an interesting alternative are interpolation methods or methods based on linear regression, which effectively reduce the number of additional measurements [3,17]. However, they are less practical than the manufacturer's procedures, as they are performed with manual settings and, moreover, require some laboratory and workshop experience.
The above considerations justify the search for further solutions, which, based solely on the initial data, will enable the verification and assessment of a malfunctioning fuel injector. To achieve this goal, the analytical methodology based on Gaussian formulas was selected. It was assumed that the base points from the standard diagnostic test would be located and connected in a Cartesian coordinate DIAGNOSTYKA, Vol. 22, No. 2 (2021) Stoeck T.: Analytical methodology for testing Common Rail fuel injectors in problematic cases 48 system. In this way, the resultant fuel dosing areas of the tested fuel injector were calculated and compared with the reference fuel injector. The mathematical operations were carried out in the environment of Microsoft Excel spreadsheet, which enabled quick analysis of the experimental data. This approach was convenient from the practical point of view, as it allows taking into account the results from any test bench without having to interfere with its software.

RESEARCH METODS
The tests were carried out on the example of a Bosch CRI 2.1 fuel injector, which was removed from a 2.4 D5 compression-ignition engine of a Volvo C30 (533) vehicle with an operational mileage of 156,000 km. Electromagnetic fuel injectors of this type operate at maximum fuel delivery pressures of 160 MPa [8]. Their characteristic feature is the lack of a spring under the valve disc (Fig. 1). The simplified structure allows the regeneration to be carried out almost to the full extent. The manufacturer made available the diagnostic technology as well as a complete set of original spare parts [11].

Test benches
At individual stages of the research, the following measuring equipment and instrumentation were used, which included, among others: -EPS 200 test bench (Fig. 2), the so-called Bosch 3-phase gear, e.g. CRR 120, CRR 220, CRR 320, CRR 420 ( Fig. 3), LAB / SM135, -SZM-168 laboratory microscope with a camera for digital image recording on a PC, ultrasonic baths (Carbon Tech Ultrasonic Bath S15/C2, Elma Elmasonic S10H), presses, vices and fuel injector disassembly and assembly tool kits, tools and torque wrenches. Before starting the tests, the fuel injector was mounted on a test bench and rinsed thermochemically. This decision was made in order to remove possible impurities and Internal Diesel Injector Deposits (IDIDs), the presence of which adversely affects the method of fuel delivery [4,9,15].
The diagnostic process was carried out in accordance with the manufacturer's requirements. Therefore, the research included a dedicated Bosch 3-phase repair kit, which ensures the highest measurement accuracy and precision when correcting fuel dosage, being a standard equipment of the injection system regeneration laboratory.

Gauss formulas
In the proposed methodology, the area of the polygon, which is presented in the Cartesian coordinate system, is calculated on the basis of the coordinates of contour turn points [1]. Assuming that the vertices x1, y1), (x2, y2), …, (xn, yn) are marked clockwise, the area of figure A can be determined using Gaussian formulas in the general form [10,14]: where: Apolygon surface area, nnumber of vertices, xi, yicoordinated of the i-th vertex.
Formulas (1) and (2) are also known as the socalled shoelace formulas because the vertex coordinates of the polygon are multiplied crosswise [2]. They should be used together in order to control the calculations made. On the other hand, it is most convenient to put the formulas created in the spreadsheet in the form of table 1, which organises and simplifies the analytical process.

Preliminary tests
Based on the data presented in Table 2, it can be concluded that the fuel injector passed the test procedure. The results obtained were within the limits specified by the manufacturer. However, after assembly, the engine was characterised by hard, uneven operation, particularly at idle and light loads.
For this reason, it was decided to implement the proposed methodology. The results of the volume measurements were located in the Cartesian coordinate system. The connection of the base points 1-2-3-4 made it possible to create an irregular quadrilateral whose surface area was estimated using the formulas (1) and (2). For this purpose, calculation formulas were entered into a Microsoft Excel spreadsheet. After substituting the numerical valuesconstituting the vertices of the analysed figure, the resultant fuel dosage area was obtained in the preliminary test APT (Table 3).  The calculations for the reference fuel injector were carried out in a similar way, using the data provided by the manufacturer (Table 4). In turn, Figure 4 shows a graphic interpretation of the results of the preliminary tests.
The disturbance of the fuel injection process causes a clear shift of the quadrilateral 1-2-3-4, and the position of individual vertices may indicate the cause of the malfunction. First of all, the very low idle dose value (point 4) is noteworthy. In such a situation, initial needle wear in the fuel injector atomiser often occurs, and thus problems with overcoming the nozzle spring tension after applying the lowest pressure on the test bench (pinj = 25 MPa). It is worth emphasising that with the increase in this operating parameter, the symptoms almost DIAGNOSTYKA, Vol. 22, No. 2 (2021) Stoeck T.: Analytical methodology for testing Common Rail fuel injectors in problematic cases 50 completely disappeared. This is evidenced by the similar location of points 1-1`, which correspond to the so-called pilot doses.

. Graphical interpretation of the preliminary test results
On the basis of the performed calculations, it was found that the difference between the resulting fuel delivery fields for the tested fuel injector APT and the reference fuel injector ASI was 10.06%. Undoubtedly, this was due to the limited operating range of the control valve. At this stage of the research, however, it is not possible to clearly answer why the values of full load dose 2 and emission dose 3 were underestimated.

Main tests
In the first step, the valve ball travel AH (German: Ankerhub) was checked. The obtained result was 0.047 µm, which was within the limit set by the manufacturer (0.046-0.056 µm [13]). Therefore, there was no need to perform a setting correction that would require changing the thickness of the valve shim at the disc (Fig. 2).
Further basic tests were preceded by the disassembly of the fuel injector into its component parts, which were bathed in ultrasonic bathes. After drying them, microscopic examination was performed under high magnification. No damage was found to the valve assembly, as well as to plunger and barrel assembly, i.e. the needle with the atomiser (Figs. 5, 6 and 7). Considering the low operational mileage of the engine, it was decided that no controls and actuators would be replaced with new ones. However, an adjustment was made, which consisted in increasing the idle doses and full load doses. For this purpose, the thickness of the needle washer was changed from DIAGNOSTYKA, Vol. 22, No. 2 (2021) Stoeck T.: Analytical methodology for testing Common Rail fuel injectors in problematic cases 51 1.14 µm to 1.10 µm, and then the nozzle spring disc were changed from 1.46 µm to 1.26 µm.  The regeneration of the tested fuel injector should be assessed positively, as the factory settings have been restored (Tables 5 and 6). The resultant fuel dosage field AMT and the standard fuel dosage field ASI are comparable, as the difference between them was only 1.33%. Therefore, the quadrilaterals 1`-2`-3`-4` and 1``-2``-3``-4`` practically overlap (Fig. 8). The vertices of the two figures have a very similar position on the graph, and the shifts so characteristic for the preliminary test are almost absent. The performed correction also had a positive effect on the pilot and emission doses (half load).

CONCLUSION
The proposed methodology of calculations makes it possible to consider specific cases of failure of common rail injectors that work incorrectly despite meeting the required manufacturer criteria. It is worth emphasising, however, that the resultant fuel dosage fields should be treated hypothetically (arbitrarily), as they do not reflect the actual fuel delivery method at intermediate points, i.e. beyond the vertices of the generated figures. Nevertheless, they can constitute a completely new diagnostic parameter, the analysis of which allows for the verification and assessment of the technical condition of the tested design, as shown in this particular example. Moreover, in the workshop and laboratory conditions, there is no need for a graphic interpretation of the obtained results, it is enough to compare them only, hence the drawings presented in the text are only illustrative. The added value is also that there is no need for additional measurements and calculations outside the test bench, without affecting its software in any way. For the above reasons, the presented solution meets the needs of the fuel injection system maintenance service market, which have been signalled in recent years. In addition, implementation in a digital environment allows the presented algorithms to be reused in research with a similar profile [16]. This approach is convenient from the practical point of view, because it enables a very quick analysis of experimental data that may come from diagnostic tests of fuel injectors of various types or generations.
In the analysed case, there were no malfunctions related to the wear of individual fuel injector component parts. The underestimated fuel dosage resulted from the presence of internal impurities, which were removed only when cleaning the disassembled elements in ultrasonic baths. On the other hand, the control and adjustment process made it possible to restore the original (factory) settings, as their values were comparable to the reference ones.