In various areas of drive technology, such as in internal combustion engines, it has always been a major aim to minimize friction and wear, in order to optimize the efficiency and durability of the components used [1
]. Chain drives are frequently used components in traction drive solutions and power transmission. Among the various types of chains, a distinction is made between bush chains and other representatives such as roller chains or toothed chains [3
]. In particular, bush chains are successfully used in the timing systems of internal combustion engines, to drive the camshaft with defined opening and closing times for the valves, as well as various auxiliary devices, such as water or oil pumps, etc. [3
A bush chain consists of several individual chain link joints, whose core elements are the bush and the chain pin. The bush is pressed into the inner link plates, and the pin is pressed into the outer link plates. In Figure 1
, a picture of several bush chain links, as well as a schematic depiction of the subcomponents in assembled condition, are shown. In the chain joint itself, the contact between the pin and bush is tribologically loaded through load transmission and the change of direction using the normal load and the relative motion between the pin and bush. This contact is generally considered one of the major friction losses in a bush timing chain [5
Wear in this contact is also a critical factor, since, on the one hand, the sum of excessive wear across all chain joints of a chain drive can lead to unacceptable chain elongation, affecting the valve timing. In this regard, chain elongations of up to 0.3% to 0.5%, depending on the application, are defined as limits [6
]. On the other hand, wear in individual joints can, in the worst case, also lead to the breakage of the joint and thus to the breaking of the chain.
A literature review revealed that several investigations have already been carried out in this area and are mentioned below in an overview of the already existing findings. Fink et al. [5
] conducted extensive research on the subject of friction losses in timing chains, which included the influence of various pin coatings on friction losses, as well as the influence of design aspects in the chain drive on the resulting friction losses. In particular, they also showed that an optimized chromium-plated pin coating provides friction advantages in the chain drive of up to 10–20% compared to the standard chrome coating over a wide speed range. In this regard, a further fundamental tribological characterization of chrome-based coatings is given in [7
], showing lower friction and wear with a physical vapor deposition (PVD) CrN coating over an electrolytic hard chrome coating. Tandler et al. [8
] also dealt with the friction losses of bush chains and investigated the influence of the lubricant on the friction losses in a special test rig, without the use of guide rails. The results of this work showed higher friction losses with used oils than with fresh oil. Meffert et al. [9
] investigated the friction effects in bush chains, whereby the investigations were carried out on a single joint-test rig. The results showed the influence of the pin/bush surface topographies of different chains and the choice of lubricant on the friction losses in the chain joint. Sappok et al. [10
] focused more on the wear testing of chain joints and the creation of an accurate method for measuring the wear of chain joint components. In another study by Sappok et al. [11
], they studied the friction and wear phenomena of the pin/bush contact (both contact partners were case hardened components) on a model tribometer, owing to the benefits of the more isolated investigation condition with this contact. In this regard, the results showed an improvement of friction and wear with a zinc dialkyldithiophosphate (ZDDP) additive compared to a base oil reference. A further demonstration of the positive hardening effects of chain components concerning wear resistance was presented by Thongjitr et al. [12
]. A numerical approach for wear prediction and estimation was presented in further work by Tandler et al. [13
The present work aimed to develop a model testing methodology that enables specific investigations of tribological phenomena (e.g., wear processes, friction assessment of various pins and plates, lubrication effects such as the influence of soot, fuel dilution, etc.) and materials occurring in the pin/bush contact area, with a high degree of visualization and exclusion of environmental effects. A special focus was put on the equivalent damage wear characterization of the pin components compared to engine applications.
2. Analysis of Engine Parts
In order to develop a reliable laboratory testing methodology for tribological engine contacts, it is absolutely necessary to investigate and understand the friction processes and tribological damage phenomena of the application to be tested. This represents a decisive factor in the validity of the methodology, especially when using a system reduction of the test technology to laboratory model scale. For this reason, commercially available components from the corresponding application, viz. a diesel test engine from an original equipment manufacturer OEM, were analyzed to understand the damage process in the present study. In addition, new parts were also investigated as a reference condition.
shows an Olympus BX51M (Olympus Austria GmbH, Wien, Austria) light microscope (LIMI) image of a pin surface in its initial condition. The curved surface (pin diameter is 4.4 mm) shows a fine but not completely smooth surface. Some pores or holes can be seen on the surface, as well as some surface scratches. In this context, scratches in the axial direction and scratches in the circumferential direction can be observed, where the scratches in the axial direction probably originate from disassembly.
In the microsection of a pin, see the scanning electron microscope (SEM) picture in Figure 3
—taken with a ZEISS EVO 15 from Carl Zeiss GmbH(Wien, Austria)—it can be seen that the pin consists of a substrate material, on which a thin coating of 2.5 µm thickness has been applied. An elemental measurement in the microsection using energy dispersive X-ray spectroscopy (EDS or EDX) from Oxford Instruments GmbH(Wiesbaden, Germany)—see line measurement A-B in the microsection—identified the substrate material as an iron-based material (Fe amount) and the coating as a CrN monolayer, which in the literature is generally attributed with good tribological properties [14
]. It can be clearly seen that the elemental composition of the iron-based spectrum changed to a 1:1 ratio of Cr and N.
In contrast to the pin, the bush is made of an uncoated steel, as the analyses in Figure 4
show. EDX measurements confirmed the presence of several alloying elements in addition to iron, such as Cr, Ni, Si, and Mn (see both element spectra in Figure 4
). The surfaces of the bushes were quite smooth, which was deduced from the optical appearance, but even more so from the roughness measurements with an Alicona InfiniteFocus IFM G4 optical roughness meter from Alicona Imaging GmbH (Raaba, Austria). The roughness measurements on this device were carried out in the form of line measurements according to the ISO 4287 standard and resulted in Ra values around 0.056 µm and Rq
values around 0.072 µm.
In order to better to understand the tribological contact and the prevailing conditions, nanoindentation measurements using a Keysight Nanoindenter G200 from Keysight Technologies Inc. (Santa Rosa, CA, USA) of the bushes and pins were also carried out. The results are summarized in Table 1
, where the hardness and the determined modulus are listed. The bush values were measured with an indentation depth of 1000 nm, with an effective evaluation range between 800 and 900 nm. Due to the low layer thickness, the CrN was measured at 10 times lower indentation depths and evaluation ranges, in order to exclude environmental influences on their measured values. For the verification of the measured values, the substrate material on the pin was measured with both measurement parameters. The results showed that the CrN layer had significantly higher modulus and hardness values, whereby the hardness value was higher by a factor of 3 and the modulus value was about 20% higher. In this regard, the nanoindentation measurement results of CrN were confirmed by the literature [18
Based on the studies of the initial condition, a damage analysis of engine components that had been in operation provided further information, in which the tribological mechanisms had to be mimicked by a proper model testing technique. In this context, parts with higher wear were examined from the representative engine application, namely a timing chain of a diesel three-cylinder engine. The verification of the wear situation was performed via the measured values from the chain elongation measurements of individual joints before and after engine operation. In this regard, higher wear was defined as individual joint wear above 30 µm. The chains were dissembled carefully, in order to not damage the running surfaces of interest.
shows a photograph of a cut pin, with the pin aligned so that the tribologically loaded area can be seen. On the right side a link plate was still press-fitted, which, in contrast, had been removed on the left side. It is already clear from this overview image that the pin surface had smoothen out during operation. This can be seen from the shiny surface, in comparison to the tribologically unstressed area on the left side, where the permanently press-fitted link plate was mounted. In the axial direction, different areas can be seen on the pin surface. Distinctive phenomena include pronounced scratches on the right side of the running surface. In this context, the pin surface in Figure 5
is more or less a representative surface condition of all the links in the investigated chain, with natural variations in individual joint wear.
Detailed pictures of the pin surface phenomena are shown in Figure 6
. These are light microscopic and scanning electron microscopic images of certain areas from the overview picture of Figure 5
, which it should be emphasized again, can be considered a representative pin condition after an engine test run, where significant chain link wear was generated. In the central areas (Figure 6
a), extremely mirror-smooth areas without noticeable sliding grooves can be seen. The holes, which also existed in the initial condition, appeared dark due to the extreme contrast difference with the smooth surface, which had a high backscattering of incident light. In these regions, hardly any traces of deposited products or tribofilm layers were found. Figure 6
b shows an off-center area with many grooves and a large area of wear. The exposure of the Fe substrate can be clearly seen in the material contrast of the scanning electron microscope image. This is evident due to the different shades of grey, with lighter areas representing more heavy elements. By means of EDX element measurements, it could be confirmed that in these areas the Fe substrate had been exposed and the CrN layer was obviously worn (Figure 6
b, Table 2
, Spectrum 2, 3), which showed an advanced state of wear. In some other areas on the pin (see Figure 6
c), very deep sliding grooves were seen and, in some cases, break-outs from the coating. Particularly deep grooves or break-outs were also repeatedly filled with lubricant residues. Other pin surfaces of parts with increased single chain link wear showed similar surface conditions; thus, the previously discussed phenomena can be classified as representative.
In the case of the counter-bodies, namely the bushes, a different damage pattern could be seen. Figure 7
a shows an overview of a tribologically stressed area of a bush. It can be seen that the bush surfaces appeared almost mirror-smooth and only some discolorations can be seen locally in the center. Very sporadically, fine areas of grooves can be seen. With higher magnification of the light microscope, the smooth surfaces were better visible (Figure 7
b). Furthermore, the discolorations are also clearly visible, which show brown and blue areas. In the scanning electron microscope and corresponding element analyses (Figure 7
c and Table 3
), it became clear that the discolorations (Spectrum 1) should be assigned to wear protection layers, namely P- and S-based layers. Outside the discolorations (Spectrum 2 and 3), essentially only the basic element (Fe) of the bush could be measured (alloying elements with low concentrations are difficult to quantify due to the low acceleration voltage, but it is not necessary to resolve this at this point). In addition, oxygen was measurable in lower concentrations, which indicated oxidation of the surface layer. In the non-tribologically stressed area, the bush surface was littered and covered with deposits (lubrication components, soot, and particles), see Spectrum 4. On chain links with less individual link wear, the bush surfaces were somewhat more roughened and not completely smooth. Tribo-chemically, however, similar phenomena were documented (slightly oxidized bush and local zones with wear protection layers). The consideration of these different stages of wear suggests a multi-stage sequence of damage in the case of the timing chains considered for diesel passenger car engines. First of all, local break-outs form on the pin, due to the tribological contact with insufficient lubrication, which leads to stronger groove formation and further wear on both surfaces. The wear particles are increasingly removed from the contact and become deposited at the non-tribologically loaded side of the bush. Further wear on the pin is generated in an adhesive progression, rather than the previous abrasive mechanism, and due to the contact of the exposed Fe-based substrate with the bush material. On the bush, the surfaces are subsequently polished, starting from the roughening that occurs at the beginning, due to layer break-outs of the pin and thus to the further fine adhesive wear. In this context, the result is an inadmissible elongation of the chain, which represents a risk for the reliability of engine operation. Overall, soot is likely to be a major factor in the process, as possible tribological wear protection layers are largely prevented from forming. This could also result in increased friction in the interface and thus facilitate layer defects and chipping. In the subsequent course of the wear process, possible further prevention of layer formation by additive adsorption or corrosive-abrasive mechanisms, as suggested in the literature [20
], is possible but could not be substantiated for the investigated component parts.
The experimental tests carried out on the model testing system could also be described and discussed with the aid of proven quantitative friction and pin wear parameters, and their surface phenomena could be compared with those from the application, in order to refine the presented testing technique, as well as to assess the new findings according to known classifications.
As defined in the methodology chapter, a quantitative assessment of the wear could be made via the measurable width of the wear cap on the pin. From this parameter, the wear volume and the linear wear height could then be derived via the geometric relationship through the cylinder shape. Based on other recorded measurement parameters such as frictional values, the results were further evaluated and discussed in accordance with well-known and accepted friction and wear evaluation concepts, such as those of Archard [30
] and Fleischer [31
]. Thus, proven characteristic values for classifying the preceding tribological phenomena were derived and can made available to the reader.
One the one hand, quantitative assessments regarding friction were carried out by calculating the dissipated friction energy. Figure 17
a shows the mean dissipated friction energy per test hour for the respective test conditions in a bar chart. In this regard, a linear relationship with the duration of the test could be determined, with all other conditions being equal. This can be read from the bar chart, as the related values for TC 1 and TC 3 are approximately at the same level, and thus a linearity of the friction energy on the total test time is seen. With an increased load, a significant increase (~ doubled) was also seen here in comparison with the reference level, which indicates that the frictional force increased relative to the applied normal load (i.e., similar coefficient of friction), thus showing that the system was sensitive to an extraordinary increase in normal force.
b shows the pin wear coefficients of the various test conditions according to Archard’s calculation principle. It can be seen that the coefficient decreased slightly with a longer test duration, although the substrate material was already exposed for longer test periods and more unstable friction processes were observed. This could have been due to the reduction of the pressure load with the increasing contact area and thus the reduced critical wear formation energy. This would mean that the wear and damage processes were higher at the beginning, which was also observed in a similar way in the analyses of the engine components. For higher normal forces, the wear coefficient according to Archard was also significantly increased, which indicated more intensive wear processes, as the other test system conditions were considered constant compared to the reference test condition. The effect of the increased load was also seen in the wear rate of the counter body. While in the case of a moderate load, no significant wear was measured for the plate, namely only 0.6 and 0.4 mg via the gravimetric evaluation method before and after the test, at high loads significant wear, 15 mg was measured. This clearly shows the more severe wear processes due to the increased load.
Jointly considering the friction–energetic processes, as well as the wear processes, the tests could also be evaluated according to Fleischer’s classification. For this purpose, the apparent friction energy density and the linear wear intensity were calculated according to calculation methods and plotted against each other in Figure 18
. It can be seen that all the tests were in similar ranges, which can be considered reasonable based on the measured parameters, the phenomena found, and the evaluation approaches. The trends described and discussed above could also be seen with this evaluation approach. Higher loads led to lower friction energy densities and higher linear wear intensities (i.e., more intensive solids contacts). Longer test times showed the opposite to the reference test condition. According to the wear classification, all tests were in the range of category 1 level 2 according to Fleischer, which means a moderate wear without thermal energy accumulation effects. This agreed with the observed surface phenomena and in situ measurement parameters, showing no signs of severe damage processes such as scuffing but rather moderate wear of deformation, slight adhesion, and sparse abrasive marks.
With regards to the informative value of the model testing technique, the comparison of the prevailing tribological mechanisms on the surfaces involved in the friction process was particularly decisive. This shows that damage equivalent conditions could be documented between the engine application and laboratory testing, which validated the applicability of the model testing technique with regard to the tribological characterization of pin wear. For pin surfaces with respect to moderate wear conditions, in both the engine and the model test technique, a significant smoothening of the CrN layer, without strong groove marks, break-outs, or deformation, could be observed (see Figure 6
a compared to Figure 15
a). Interactions with the lubricant were largely absent in such conditions or were not detectable with the applied analysis techniques. The pores present in the initial state could be detected in the engine, as well as in the test for these wear conditions. In the case of an increased wear state (in the test due to an increase in the test time or load; in the engine presumably due to lubrication conditions, geometric scattering of components and associated higher loads) the CrN layer was worn away and the Fe substrate was exposed (for comparison see Figure 6
b,c and Figure 15
d–i). Furthermore, in this case, break-outs of the CrN layer could also be seen, which most likely preceded the increased wear rate. When the steel substrate was exposed, an increased number of grooves was seen on the contact surface. This phenomenon was prevalent in both engine parts and test specimens. In addition, an increased proportion of deformations was seen in these areas. These equivalent conditions validated the developed model testing technique with regards to being an efficient laboratory screener for pin wear. A condition equivalence between the application and model test was also be observed on the mating body surfaces. In both cases, the steel surfaces were smoothly polished and slightly oxidized. Grooves were only observed locally. In addition, local wear protection layers on the P and S bases were equally measurable.