Emergence of Coated Piston Ring Scuffing Behavior on an Application-Oriented Tribological Model Test System

Abstract

A major problem in lubricated piston ring/cylinder liner contact sliding systems is the tribological failure mechanisms known as scuffing. In order to evaluate and better understand this damage phenomenon in these tribological systems, a tilted linear tribometer (TE77) for application-oriented reciprocating model tests was developed and validated with scuffed field engine parts. With precise oil lubrication, original engine parts, such as CKS-coated piston rings (chromium-based coating with included aluminum oxides), original liners and fully formulated lubrications, were tested under conditions similar to the most critical part of the internal combustion engines (ICEs), known as fired top dead center (FTDC). Various in situ measurements during the tests allowed for a detailed investigation of the damage processes (crack transformation) on the tribological components. For the coated piston ring, vertical cracks were attributed to residual stresses, while horizontal cracks resulted from shear stresses. The crack transformation and wear results from other studies were confirmed for the liner. The results from FIB (Focused Ion Beam) cuts, along with EDS and SEM analyses, revealed that Fe (deriving from material transfer) acts as a catalyst on the CKS layer for the tribopads and that zinc sulfides are not present everywhere.

1. Introduction

Reducing fuel consumption, increasing the power density and achieving CO₂ targets are still the main driving forces for new engine developments in the field of internal combustion engines (ICEs), especially in the field of large-engine systems [1,2]. To achieve such developments, a good fundamental understanding of the mechanism of the piston ring/cylinder liner system among others is required, considering the amount of lubricant and the lubricant’s composition. Within the piston group, the ICE loses most of its frictional energy, in percentage terms, and therefore the piston ring/cylinder liner system offers a promising opportunity for further improvements [3]. The cylindrical piston assembly consists of a piston that moves back and forth within a cylinder to create a sealed working volume with the cylinder head. This movement allows for the compression of the air-fuel mixture and the subsequent expansion of combustion gases, ultimately generating mechanical work. The key tribomechanical systems include the contact between the piston ring and cylinder liner, as well as the contact between the piston pin and connecting rod eye. Cost-saving pre-studies can be performed in a controlled environment using laboratory models tests for this complex system.

Some fundamental investigations have already been carried out for basic understanding of interactions between piston ring, cylinder liner and lubricant [4,5]. Some test setups have also been developed which enable testing as application-oriented as possible, such as those of Barber et al., Johansson et al., Lenauer et al. and Schiffer et al. [4,5,6,7]. As one of the model test benches, which are considered application-oriented, the ring-on-liner test setup has proven to be very suitable for testing various system combinations [8,9]. It is possible to use original engine parts, from which the test specimens are manufactured. Friction, wear and scuffing limits are important parameters to describe and evaluate the performance of the different parts and lubricant combinations. For this reason, these parameters are usually referred to when investigating effects and tribo-processes within a ring/liner system. The ability to adjust any parameter such as speed, normal force, system temperature and lubricant supply in a few seconds is the biggest advantage compared to fired engine tests [9].

Zhang et al. [10] conducted research on wear behavior to demonstrate the combined conditions of speed and contact pressure, considering lubrication, under which the cylinder liner experiences the most significant wear. By using various test rigs, studies in a wide range of specific research topics in the field of ring/liner systems have already been carried out. Thus, investigations were carried out with various lubricants [11] and particle-containing lubricants [12], and in order to create application-oriented conditions with regard to the amount of lubricant, tests were also carried out with low lubricant supplies [4,13]. Scuffing damage is still not fully understood, but Qu et al. observed that scuffing generally occurs at the reversal points [14]. Van et al. [15] also showed that the scuffing phenomenon usually starts at the TDC, where the lubrication conditions are worst, and Dyson et al. [16] considered that the failure of the lubricating film is the main reason for scuffing initiation. This is usually accompanied by a sudden increase in friction, temperature and vibration [17]. In addition, investigations on different cylinder liner honing structures and with different piston ring coatings [18,19] and surface textures [20,21] were carried out with a focus on scuffing phenomena, friction and wear behavior [22,23,24]. Damage analyses on fired engine components were also carried out [25].

In this study, various optimizations were made to the ring-on-liner test system and test strategies to create even more application-oriented test conditions. The test conditions were chosen as close as possible to the FTDC conditions because this area is characterized by high temperatures, rather low sliding speeds, a low lubricant supply and high normal forces from the combustion process. An original coated piston ring, original cylinder liners and a fully formulated oil were tested and validated via engine parts. The added value of this study lies in its contribution to the understanding of scuffing damage in selected piston ring/cylinder liner systems. Unlike most studies, which primarily focus on the scuffing behavior of the liner, this research also examines the scuffing behavior of the piston ring. This dual focus allows for a comprehensive evaluation of the damage processes occurring on the piston ring and how a tribological layer forms on its surface. By investigating both components of the friction pair, this study provides a more complete picture of the wear mechanisms at play, offering valuable insights into how scuffing damage initiates and propagates in real engine conditions. This knowledge is crucial for developing more durable materials and coatings, improving engine performance, and extending the service life of critical engine components.

2. Methodical Approach

For the research approach, multiple measures were carried out to ensure scientific quality and novel insights in this topic.

• The first step of this study focuses on analyzing the properties of all tribo-partners involved as piston ring, cylinder liner and lubricant in the initial conditions. The piston ring is examined regarding geometry, curvature, hardness, layer structure and surface texture. The cylinder liner is analyzed regarding hardness and surface roughness. As regards the lubricant, additive elements and physical properties are explored.
• The second step includes analyzing field engine parts. The aim of this analysis is to evaluate the damage mechanism of the application and to obtain reference data of an end-of-life condition for the validation of the application-oriented model test setup.
• The next step is to determine the relevant ICE parameters such as temperature, piston ring pressure, velocity and lubricant consumption. These parameters are converted to the size, geometry and capabilities of a model test rig. Also, parameters of other model test rigs are evaluated.

Based on this approach, performance screening under close-to-application conditions are carried out and validated to field engine parts to resolve the prevailing tribo-processes.

3. Analysis of New Parts

3.1. Piston Ring

A piston ring (Figure 1) with a chromium-based coating with included aluminum oxides (CKS36 [26]) and a spheroidal graphite cast iron (GOE 52 [26]) as substrate material was used. The base form of the piston ring cross section is rectangular. The piston ring has a diameter of 190 mm, an axial thickness of 5 mm, and a radial thickness of 7.35 mm. The coating thickness is in a range of 200 µm depending on the curvature and the cooling processes a hardness of about 900–1200 HV. The surface view (Figure 1—right side) in tribological contact area 1 shows a crosscut grinding profile from the manufacturing process. In areas 2 and 3, outside of the tribological contact area, there is a tangential grinding profile from the manufacturing process.

The surface roughness (

Table 1. Roughness parameters of piston ring surface.

Value	Sa [µm]	Sz [µm]	Spk [µm]	Sk [µm]	Svk [µm]	Smr1 [%]	Smr2 [%]
Average	0.176	10.04	0.266	0.511	0.352	7.71	86.43
Min	0.164	9.48	0.222	0.472	0.315	7.55	85.53
Max	0.182	10.8	0.297	0.546	0.379	7.83	87.36

) was measured on five piston rings at two points with an Alicona G4 Focus (Alicona Imaging GmbH, Raaba in Austria) according to the standards DIN EN ISO 4288 and DIN EN ISO 25178 [27,28]. The piston ring surface has an average value of mean surface roughness of 0.176 µm, associated with low values for the peak heights (S_pK) and low values for reef depths (S_vk). The surface view also shows some very small holes and a fine crack network, also shown in the cross section view by Wang et al. [29].

3.2. Cylinder Liner

For the cylinder liner specimens, a cylinder liner with a lamellar graphite base material (Figure 2) was used. Steadite (phosphide eutectic—Fe, P and C [30]), with a hardness of about 970 HV was also seen in the cross section. The surface has a honing structure with a honing angle of 60 degrees and the hardness of the material is in the range of 250–290 HBW30 [30]. The surface shows a few small breakouts from the honing process in the initial state.

The surface roughness of three cylinder liners (

Table 2. Roughness parameters of the cylinder liner surface.

Value	Sa [µm]	Sz [µm]	Spk [µm]	Sk [µm]	Svk [µm]	Smr1 [%]	Smr2 [%]
Average	0.500	26.11	0.321	0.985	1.653	6.97	79.31
Min	0.450	20.83	0.293	0.886	1.469	6.71	78.69
Max	0.570	31.44	0.342	1.113	1.937	7.10	79.98

) was measured at three points on each liner using an Alicona G4 Focus (Alicona Imaging GmbH, Rabba in Austria) in accordance with the standards DIN EN ISO 4288 and DIN EN ISO 25178 [27,28]. The surface of the cylinder liner has an average of S_a value of 0.5 µm, together with low values for the peak heights (S_pK). The oil retention volume is characterized by higher values for the reef depths (S_vk). Some very small breakouts from the honing process are also visible in the surface view.

3.3. Lubricant

The fully formulated lubricant is an SAE40 with a kinematic viscosity at 40 °C of 11.3 × 10⁻⁵ m²/s and at 100 °C of 1.31 × 10⁻⁵ m²/s [31]. The ICP (inductively coupled plasma atomic emission spectroscopy) data (

Table 3. ICP analysis of the tested lubricant.

Test	Method	Value	Unit
TBN	ASTM D2896 [33]	7.3	mgKOH g−1
TAN	ASTM D664 [34]	1.95	mgKOH g−1
NITROGEN	ASTM D5762 [35]	973	µg G−1
TBN	ASTM D4739 [36]	6.3	%m m−1
Chlor	ASTM D6443/C [37]	<0.0003	%m m−1
Al	ASTM D5185 [38]	<0.0006	%m m−1
Cr	ASTM D5185	<0.0001	%m m−1
Cu	ASTM D5185	<0.0002	%m m−1
Fe	ASTM D5185	<0.0002	%m m−1
Mo	ASTM D5185	<0.0005	%m m−1
Ni	ASTM D5185	<0.0005	%m m−1
Pb	ASTM D5185	<0.0010	%m m−1
Si	ASTM D5185	<0.0008	%m m−1
Sn	ASTM D5185	<0.0010	%m m−1
Ba	ASTM D5185	<0.0001	%m m−1
Ca	ASTM D5185	0.163	%m m−1
Mg	ASTM D5185	0.0008	%m m−1
Zn	ASTM D5185	0.0312	%m m−1
P	ASTM D5185	0.0277	%m m−1
S	ASTM D5185	0.221	%m m−1
Ba	ASTM D5185	<0.0004	%m m−1
K	ASTM D5185	<0.0040	%m m−1
Na	ASTM D5185	<0.0007	%m m−1
Mn	ASTM D5185	<0.0005	%m m−1
Ag	ASTM D5185	0.0004	%m m−1
Ti	ASTM D5185	<0.0005	%m m−1
V	ASTM D5185	<0.0001	%m m−1

) of the lubricant show 0.163%m m⁻¹ Ca, 0.031%m m⁻¹ Zn and 0.028%m m⁻¹ P, which are the main driving forces for the formation of additive layers [32].

4. Analysis of Field Parts

To evaluate the damage mechanisms, a few piston rings and a liner from a field engine were analyzed, which have been classified as scuffed by the OEM (original equipment manufacturer) assessment. The piston ring analysis with a digital microscope (Keyence VHX 5000 (Keyence International NV/SA, Mechelen in Belgium)) showed that mainly horizontal (means 90 degrees in the sliding direction) cracks starting from the ring gap end with some vertical (means parallel to the sliding direction) branches represent the beginning of the damage process (see Figure 3). The contact pattern is located in the lower part of the piston ring. The EDS and SEM analysis (EDS System—Oxford INCA (Oxford Instruments Nano Analysis & Asylum Research Ltd., Abdingdon in UK); SEM—Zeiss EVO MA15 (Carl Zeiss GmbH, Vienna in Austria)) of the piston ring shows traces of Fe and Si in these darker grey areas, which would suggest material transfer from the liner, because the coating thickness of the piston ring coating CKS36 is more than 200 µm and does not contain these elements initially. In addition, traces of lubricant additives such as P, Ca and S were found on the surface.

The analysis of the liner (Figure 4) in the TDC shows a completely destroyed surface with fine cracks and many breakouts. The initial honing structure is lost completely. The EDS and SEM analysis shows a mixture of smeared iron with additive elements (Ca, P, S and Zn) from the lubricant. Carbon deposits from the combustion process were also found on the surface. The scanning electron microscope images of the engine part were recorded at an acceleration voltage of 7.5 kV.

5. Experimental Details

5.1. Description of the Reciprocating and Application-Oriented Model Test Rig

Tests were performed on a linear tribometer TE77 (Phoenix Tribology Ltd., Kingsclere in UK) (details can be found in [6]), which provides precise control of speed, normal force and system temperature. For lubrication supply, a peristaltic pump (Watson Marlow 520 (Watson Marlow Ltd., Falmouth in UK)) and a syringe pump (Lambda VIT-FIT (Lambda Instruments GmbH, Baar in Switzerland)) were used.

The cylinder liner specimen, which is made from an original engine part (Details see Section 5.3), is clamped on a fixed specimen holder in an oil pan (Figure 5) that can be heated to a specific temperature with heating rods. Also, the piston ring specimens are made from original engine parts and are guided through a specimen holder that allows the piston ring specimen to move in the same way as in an original piston ring groove of the piston. This means same side and groove clearance and same clearance from the piston ring holder to the cylinder liner wall. The piston ring rotation is possible to ±2 degrees. The fully formulated original engine lubrication is filled into the tribological system about 1.5 mm away from the reversal point of the ring with a needle, which is named BDC (bottom dead center). This allows the lubrication to warm up before it comes into tribological contact.

As regards the base frame of the linear tribometer, a tilting mechanism was developed and through this equipment, it can be tilted to an angle of 28 degrees, which aims to mimic more upright alignment conditions of the engine situation between ring and liner. The result is a very low and continuous flow of heated lubricant through the tribological contact, which stands out from many tribological linear model test systems that are horizontally aligned.

Two temperature measurement points were integrated into the cylinder liner sample. The system temperature is the control variable for the heating system. The contact near temperature is measured about 1.7 mm under the reversal point of the piston ring, which is named TDC (top dead center). At the piston ring, about 1.5 mm under the surface, in the center of the contact area, a temperature measurement point is integrated. An electrical resistance measurement of the contact, called the contact potential (CP), is also implemented between the piston ring and the cylinder liner sample holders. The value represents the electrical isolation of the contact, where a value of 50 mV means completely isolated contact and a value of 0 mV indicates metallic contact (for a detail description, see [4,39]. The oil pan is connected to a force sensor that measures the frictional force during the test. The normal force is applied to the top of the piston ring holder. The normal force parameter and the friction force measurement can be used to calculate the coefficient of friction (COF).

5.2. Boundary Conditions and Materials

Two test strategies were developed to mimic different engine conditions. Both test strategies are load step tests where the normal force increases stepwise after ten hours of the run-in phase at a low normal force. These types of tests aim to characterize emergency running conditions and scuffing load limits. The run-in phase ensures the same starting conditions for the different tests. The motion frequency in the start-up phase reaches 20 Hz after 2 min and remains constant through the whole test. The mean piston speed at 20 Hz is 0.984 m/s. The normal pressure in the run-in phase is 2 MPa (applied on the projected contact area) and is increased by 1 MPa every 20 min in the load step phase until it reaches 28 MPa, which is the limit of the maximum normal force of the linear tribometer or until the system fails. The difference between the two test strategies is the system temperature and lubricant supply (see Figure 6). By varying these parameters, two test setups can be defined.

In test setup 1, the experimental parameters were carefully selected from published articles and the available literature, and thus a system temperature of 230 °C, obtained from Schiffer et al. [30], and a lubricant supply value of 42 µL/min, obtained from Pusterhofer et al. [11], were used.

For test setup 2, the experimental parameters were carefully matched to the field engine parameters for a different application as is evaluated in

Table 4. Overview of engine parameters.

ICE Parameters	Car Engine	Truck Engine	Large Bore Engine
Max. temperature @ first piston ring groove [°C]	~200–230 [40,43,44]	~200–230 [40]	~200 [44]
BMEP [bar]	Diesel + Loading ~ 20 [3] Otto + Loading ~ 22 [3]	Diesel + Loading ~ 24 [3]	Gas ~ 28 [3,45]
Mean piston speed [m/s]	Diesel ~ 15 [3] Otto ~ 20 [3]	Diesel ~ 14 [3]	Gas ~ 11 [46]
Lubricant consumption [g/h]	~1.3 [9]	-	~25–60 [42]
Piston speed @10° crank angle [m/s]	Diesel ~ 4.1 Otto ~ 5.6	Diesel ~ 3.9	Gas ~ 3.1

, so that the system temperature is reduced to 200 °C [40] and the lubricant supply value is reduced to 0.72 µL/min, which is a reduction of a factor of about 60. This lubricant supply rate of 0.72 µL/min corresponds to a lubricant consumption of about 40 g/h per cylinder. Vareka et al. [41,42], for example, reported a lubricant consumption between 25 and 60 g/h per cylinder for big bore engines. Mahle [40] showed that about 80 to 90 percent of the lubricant consumption is lost in the form of unburned hydrocarbon (HC) in the exhaust gas. This would indicate that the top piston ring is lubricated with at least 80% of the lubrication consumption value. So, this means the application-oriented lubricant supply for the top piston ring should be in a range of 20 to 48 g/h per cylinder.

The aim of both test setups (Figure 6) is to generate an application-validated scuffing phenomenon of the ring-liner system. The parameters used for both test strategies are listed and summarized in

Table 5. Test strategy conditions.

Test Setup Parameters	Test Setup 1	Test Setup 2
System temperature [°C]	230	200
Pressure range [MPa]	0–28	0–28
Mean piston ring speed [m/s]	0.984	0.984
Lubricant consumption [µL/min]	42	0.72

.

5.3. Test Procedure and Matrix for the Tribological Tests

For each test, a new piston ring specimen and a new cylinder liner specimen were manufactured out of the original engine parts (see Figure 7). The piston ring specimen exhibits the best geometrical conformity [47] with the cylinder liner geometry at the ring gap. The cylinder liner specimen was manufactured from the TDC area with a wall thickness of 3.3 mm, width of 7 mm and a length of approximately 35 mm. To maximize repeatability, the exact roughness parameters of each cylinder liner specimen were measured using a Mahr Pocket Surf PS1 (Mahr GmbH, Göttingen in Germany). The specimens with the best agreement of the roughness parameters (compared to the specification) were selected from the sample pool to reduce the scatter of the experiments. The selection criterion was ±10% of the S_a and S_k value of the measurement compared to the manufacturing specification.

By using a 3D printed ring-liner fixing tool and a pressure sensitive foil (Fujifilm Prescale), it is possible to ensure that there is geometric conformance for each test (see Figure 8). Through this process, edge loading phenomena or scratch tests, which would result in different tribo-processes and high scatter, are prevented. This check was performed at the TDC, BDC and the middle of the stroke from the linear tribometer (mid-stroke), and reduced the scatter of the experiments.

Figure 9 shows the complete test matrix including the type of test setup, the stop criteria used and the parameters measured during the test. For tests one to eight, scuffing phenomena occurred, which also were the test stopping cause. This means that the test was stopped when the measured COF value was above 0.25 or the measured temperature near the contact was above 300 °C. Tests 9 to 16 were user-controlled stopped tests at different normal loads to evaluate of the damage process up to the scuffing phenomena.

6. Results

6.1. Test Results—Scuffing Phenomena

Initially, tests were performed with test setup 1. The data analysis (Figure 10) shows the different parameters measured during the test. The CP shows the run-in of the system during the first two hours of the test, when the system is heated up. This means a stabilization of the CP value, which indicates the formation of tribopads and removal of roughness peaks from the manufacturing process. During the test period from two to ten hours, the CP remains relatively constant at 50 mV. In the load step phases, some break-in of the CP value occurs, which becomes more pronounced with an increased load. The contact near temperature is in the run-in phase about 10 °C higher than the system temperature. During the load step phases, the system temperature stays constant, but the contact near temperature increases continuously through friction heat. The COF is relatively constant in the run-in phase and decreases continuously during the load step phases. All these tests are stopped because the maximum normal load of the linear tribometer has been reached and no scuffing occurred yet. Also, the test runs stable until the end. Three tests were performed with this test setup (see Figure 9). All of these tests have high reproducibility, which means the test data and the corresponding surfaces of the tribological partners look the same. Therefore, a representative test diagram from only one of these tests is shown in Figure 10.

The corresponding contact surfaces of Figure 10 are shown like an open book around the dashed line in Figure 11.

The surfaces of this test setup show only marginal transfer from the liner and formation of additive layers from the lubricant (

Table 6. Chemical composition (at. %) of the piston ring surface—Test setup 1.

Spectrum	Area 1—S1	Area 1—S2	Area 1—S3
C	40.95	41.93	8.86
O	32.47	33.33	12.67
P	6.11	8.45	0.64
S	2.51	4.14	0.79
Ca	3.42	4.21	2.36
Cr	8.70	-	74.67
Zn	5.84	7.94	-

) on the piston ring (piston ring—area 1 in Figure 11). A vertical crack is also visible in the center of the surface of the ring. As regards the liner, a few breakouts are visible on the surface (liner—area 2). Furthermore, tribopads consisting of P, S, Ca and Zn are found on the liner surface. The tribopads in the liner of area 1 of the TDC and about 3 mm along the BDC look visually thinner compared to the rest of the surface (liner—area 2) and have lower amounts of the lubricant components (

Table 7. Chemical composition (at. %) of the liner surface—Test setup 1.

Spectrum	Area 1—S1	Area 1—S2	Area 1—S3	Area 2—S1	Area 2—S2	Area 2—S3
C	25.13	21.86	22.55	26.64	16.22	16.27
O	23.45	26.56	18.53	21.51	48.09	53.17
Si	1.72	1.63	1.85	0.92	0.48	-
P	0.47	2.12	1.52	3.43	8.54	8.76
S	0.98	2.37	2.21	1.65	1.10	1.05
Ca	1.88	13.70	8.54	9.35	17.63	20.59
Fe	46.38	31.76	44.80	35.96	7.72	-
Zn	-	-	-	0.54	0.22	0.16

), as found in area 2 along the BDC. The P level in area 1 is in a range of about 0.47 at. % to 2 at. %, in comparison to area 2, in which it is 3.43 at. % to 8.76 at. %.

In the tests with test setup 2 (Figure 12), in the first two hours of the test, when the system is heated up, the CP shows the run-in of the system similar to that of the tests with test setup 1. In the test period of about two to ten hours, the CP remains quite constant at 50 mV. At the beginning of the load step phases, the CP value, in comparison to the test setup 1, remains on the same maximum level. Immediately before the system is stopped, because of a contact near temperature higher than 300 °C, the CP value falls down to about 0 mV, which indicates metal contact. The contact near temperature is in the run-in phase about 10 °C higher than the system temperature. During the load step phases, the system temperature remains constant, but the contact near temperature increases similarly to that observed with the test setup 1. Similar to the CP, immediately before the end of the test, the contact near temperature decreases, which indicates less friction heat being produced during this period. The COF is relatively constant during the run-in phase and continuously decreases during the first load steps. Before the test stops, the COF value decreases slightly and due to the less frictional energy being introduced, the temperature also decreases, which has been mentioned above. At this point, the CP value is in the range of about 0 mV, which means metal–metal contact. Immediately after this decrease in COF, CP and temperature, the system starts to scuff, which implies a rapid re-increase in the COF and temperature. The system stops when the COF value is higher than 0.25 or the contact near temperature is higher than 300 °C. Due to the early stop, the surface is not completely destroyed and a surface analysis is possible.

The surfaces of this test (Figure 12) show material transfers from the liner, as well as horizontal and vertical cracks on the surface in the area of piston ring 1 (Figure 13 and

Table 8. Chemical composition (at. %) of the piston ring surface—Test setup 2.

Spectrum	Area 1—S1	Area 1—S2	Area 1—S3
C	14.85	8.73	37.73
O	19.32	9.04	12.26
Si	1.53	1.27	1.38
P	0.39	0.20	0.32
S	0.25	0.27	0.51
Ca	0.56	0.50	1.20
Cr	-	6.60	9.06
Fe	63.10	73.39	37.55

).

The liner shows a destroyed surface with cracks, breakouts and scuffing marks (Figure 13). In this case, the secondary and primary structure are worn away in the middle of the liner width. On the sides, only the secondary honing structure is worn away. EDS and SEM analysis (

Table 9. Chemical composition (at. %) of the liner surface—Test setup 2.

Spectrum	Area 1—S1	Area 1—S2	Area 1—S3
C	31.48	31.99	26.52
O	32.47	19.02	32.77
Si	1.71	1.20	1.84
P	0.39	0.19	0.32
S	0.39	0.38	0.39
Ca	0.90	-	0.61
Cr	-	-	0.40
Fe	32.66	47.23	37.15

) now shows a mixture of small amounts of additives and liner material in the scuffed areas.

An overview of the results of the two experimental setups is shown in Figure 14. The left side of the diagram shows the test end results of test setup 1. Each test of this test series of experiments was a run through to the maximum load limit of the linear tribometer without a scuffing phenomenon on the surface. The right side of the graph shows the results of test setup 2. In this test setup, the tests stopped between eleven and thirteen MPa because of a scuffing phenomenon on the surface.

6.2. Test Results—Controlled Stopped Tests

Stopped tests at various loads were also carried out to elucidate the damage process occurring in this coated ring liner until scuffing phenomena take places. Due to the fact that for test setup 1 (literature-based lubrication supply) no scuffing occurred and thus no damage progress can be observed, these stopped tests have only been carried out with test setup 2 (application-oriented lubrication supply). As already mentioned before, a new piston ring, liner and lubricant were used for each test. Each piston ring, each stopped at a different load, was examined with a digital microscope after the test and the surface development was compiled (Figure 15). After the run-in period, the piston ring shows no change compared to the initial surface. As seen, the ring surface starts to become slightly polished. The mirror surfaces indicate mild wear processes on the surface. Also, in the middle area, some dark areas are seen, which represent additive layers such as those presented in Figure 15. In the load range of about 6 MPa, the first fine cracks can be observed on the piston ring surface, which are mainly horizontal. After a load of about 9 MPa, the piston ring shows a fine vertical and horizontal crack network in the contact pattern, which is much deeper than the initial coating crack network. At 10 and 11 MPa, the crack network becomes deeper and more visible. When scuffing takes place at about 12 MPa, a very clear crack network with vertical and horizontal cracks is visible. The formation of horizontal cracks can be attributed to tribological shear stress. Vertical cracks are assumed to be driven by residual stresses. To prove this theory, selected rings are also examined in oven tests without tribological stressing (see Section 6.3). EDS and SEM analysis of the rings at twelve MPa shows traces of Fe and Si originating from the material transfer from the liner, as well as traces of additives from the lubricant (Ca, P and S). At higher loads, material transfer from the liner is visible in the spectra and Zn from the lubricant is also measured. The scanning electron microscope images are taken at an accelerating voltage of 7.5 kV.

The next step of the analysis focused on the structure and its change along the damage process. This was carried out by using a FIB (Focused Ion Beam) cross-section cut (Figure 16) to expose the structure in depth and a subsequent analysis by EDS and SEM. In order to be able to analyze the upmost tribological layers, the surface was sputtered with gold (Figure 16—1) so that the tribological layer would not be damaged when the surface was removed. The cross section also exposed the recrystallization (Figure 16—2) at the edge zone near to the tribological layer. The thickness of the tribological layer was lower in the test with two MPa as well as in the test with ten MPa compared to the test with six MPa.

The EDS analysis of the various piston rings from the different load levels showed the same structural composition of the element layers as in Figure 17. Above the CKS layer, on which there is also an additive anti-wear layer, a thin Fe layer was always found in between. Fe and Si elements from the liner were uniformly present in the remaining tribological layer but in low concentration. Furthermore, Ca, P, S and minimal Zn particles were found in the tribological layer. Above the Fe deposit, which overlies the CKS layer, an enrichment of the P elements took place. Above the enrichment of the P elements, an enrichment of the S elements took place. The Ca elements were uniformly distributed and traces of Zn were found only in local places in the tribological layer structure.

In Figure 18, a quantitative EDS area analysis of the tribological anti-wear layer is shown.

Table 10. Chemical composition (at. %) of the piston ring cross-section from the FIB-Cut.

Spectrum	Spectrum 2	Spectrum 3
C	18.33	14.68
O	53.72	51.59
Si	1.60	2.42
P	4.30	22.97
S	19.68	4.05
Ca	0.66	3.54
Fe	1.71	0.75
Zn	-	-

shows the results of the EDS areas from spectrum 2 and spectrum 3 on the FIB-cut surface. Fe and Si elements from the liner are fairly uniformly present in both EDS measurements. In spectrum 3, which is near the CKS layer, an enrichment of the P element takes place. In spectrum 2, the enrichment of the S element is shown.

The analysis of the liner was also performed with a digital microscope, EDS and SEM (Figure 19). After the run-in phase at 2 MPa, the surface looks almost similar to the initial condition. However, more small holes are visible compared to a new liner segment. Tribopads with high amounts of P (~3 at. %) and S (~8 at. %) and Ca (~16 at. %) were visible at the plateaus. High proportions of the liner base material (Fe (~62 at. %) and Si (~2 at. %)) as well as C (~20 at. %) and O (~8 at. %) were found in the primary honing structure. Traces of the additives (P (~0.11 at. %), S (~0.55 at. %), Ca (~0.76 at. %) and Zn (~0.15 at. %)) were also found there. The small breakouts on the surface have high amounts of C and O and in the sectional view; these breakouts or flakes are mostly on the top of graphite lamellae. At the test stop load of 6 MPa, more breakouts are visible at the surface and the breakouts become deeper than graphite lamellae, which lie deeper under the surface. Also, a fine crack network is visible on the surface, mostly starting at the breakouts. The tribopads at the plateaus now have lower amounts of P (~0.11 at. %), S (~0.5 at. %) and Ca (~1.2 at. %), and look thinner and smaller from a visual perspective. The element spectra of the primary and secondary honing structure remain in the atomic percentage range at the same level as at 2 MPa. At a load of about 9 MPa, more and larger holes are formed. The tribological elements in the spectrum of plateaus become a little bit higher, and the secondary and primary honing structures remain in the same range. Tempering colors are also seen on the surface. At about 10 MPa, the tribological elements in the secondary and primary honing structures remain fairly constant. At the plateaus, the tribopads become optically thicker and the tribological proportions of additive layers become higher (P (~3 at. %), S (~12 at. %), Ca (~25 at. %)); also, Zn (~0.5 at. %) is now visible in the spectra. From the surface view, the breakouts and crack network become more intensive and the tempering color also changes. In some areas, especially TDC reversal point, micro seizure marks become visible. In addition, the secondary honing structure is also removed in some areas. Before macro scuffing starts, the tempering colors changes to grey/light blue, which indicates a surface temperature of about 300 °C. A lot of huge breakouts and scuffing marks can be seen on the surface. The test is stopped at the beginning of the scuffing phenomena. In this case, in the middle of the liner width, the secondary and primary honing structures are worn away and the scuffing marks are clearly visible. On the outer sides, the secondary honing structure is partly worn away. The EDS and SEM analysis now shows a mixture of additives and liner material in the worn areas. The scanning electron microscope images of the liner segments were taken with an accelerating voltage of 7.5 kV.

6.3. Heating Tests

Heating tests were also carried out to evaluate surface changes and crack initiation due to residual stresses. The piston rings were installed in an original cylinder liner segment. The package was placed in the oven at room temperature. The heating ramp from room temperature to 300 °C took about two hours. The cooling ramp took about four hours. Before heating the new piston ring, the surface was examined with a digital microscope, focusing on the crack network of the surface. Figure 20 shows the temperature measurement from the single cycle test and the ten-cycle test. For the ten-cycle test, a new package of piston ring and cylinder liner was used.

The surface analysis (Figure 21) of the single cycle test showed that a crack network at the ring gap end grew and, in particular, vertical cracks became more intense and visually deeper. The analysis of the ten-cycle test showed that the crack network not only became more intense and visually deeper, with mainly vertical cracks at the ring gap end, but the vertical cracks on the outer surface of the entire ring also became more intensive and visually deeper.

7. Discussion

For the test development, the derivation of the parameters and the equivalency of damage phenomena to application were the main focus.

Test setup 1, using a lubrication supply from the literature of 42 µL/min, did not create large-area scuffing and was too mild for an evaluation of the scuffing phenomena. Hence, the tests were run through. The damage analysis shows some horizontal cracks on the piston ring and marginal material transfer from the liners and additives from the lubricant. The liners show marginal surface damage with additive layers from the lubricant. So, before scuffing occurred, the maximum load of the linear tribometer was reached, attributable to the higher lubrication supply, which creates a higher hydrodynamical effect in the tribological system.

Test setup 2, with the system temperature lowered and the oil supply reduced by a factor of 60, shows scuffing phenomena in the middle of the normal force range of the linear tribometer of around 11 MPa. The damage analysis compared scuffed field engine parts with the model specimen of test setup 2 and showed a good damage equivalency between model test and application. The contact areas from the model tests and from the field engines tended to be in the middle and lower area of the piston ring width. On the liner, the geometric conformity looks uniform across the width, and the results are similar to the liner and piston ring running patterns of Zima et al. [48]. Also, for the stopped tests, the contact area width of 7 mm formed from the beginning onwards. The data analysis of COF, temperature and contact potential showed an expected dependency on each other. After the run-in phase of the system, the load increase started and the contact near temperature increased. The COF value decreased to a value of about 0.075. The peaks from the CP, along the load step phase, became more intense and frequently increased. This is an indication that the system is in change, like in the description of the stopped tests. This change correlates with the investigations of Pusterhofer et al. [11] for uncoated systems. The damage analysis of the parts after the tests showed material transfer from the liner on the piston ring, as demonstrated by Kamps et al. [49] and Tas et al. [17], and that the piston ring formed a horizontal and vertical crack network as in the stopped tests’ description.

The heating tests of the rings showed that the vertical cracks were mainly caused by the temperature input into the system and the removal of residual stresses. The horizontal cracks were caused by the tribologically introduced shear stresses from the sliding motion along the stroke.

Due to the load step tests, the frictional forces increased step by step and thus initiated the crack propagation and this resulted in damage processes, which are very similar to those in the work of:

• Pusterhofer et al. [11], who used a piston ring with a removed coating.
• Saeidi et al. [50], who used lesser sliding speed and a 42CrMo4 steel counter body.
• Zhang et al. [10], who focused on wear and tear on the liner.

First, a fine crack network begins to emerge, starting mostly from small breakouts, which are mostly located above graphite flakes. With increasing of the contact area pressure, coming from the increase in the normal force, the crack network grows and deeper breakouts are formed. The tribopads become thinner and smaller. Before it comes to the end of the lifetime, it can be observed that the tribopads on the plateaus grow again. This would indicate that in this case, the growth of the tribological layers is promoted again. At the end of the lifetime, when the CP value has many peaks to 0 mV or drops constantly to 0 mV, micro scuffing marks start and become visible, mostly at the TDC reversal point, as observed by Qu et al. [14]. At the end of lifetime, at the scuffing phenomenon, the scuffing marks are clearly visible and the surface is destroyed. The surface shows a mixture of smeared iron with elements from the lubricant and sometimes material transfer from the piston ring.

The Fib-cut of the piston ring shows that in the edge zone, a recrystallization of the coating structure also takes place. This recrystallization hints at higher temperatures and temperature peaks (flash temperatures [51,52]) in the contact area. On the surface of the CKS coating, a thin layer of higher amounts of Fe is formed.

This indicates that Fe acts as a catalyst for the tribopads forming on the chrome-based coating. Fe and Si particles are present in small amounts throughout the tribological layer and originate from the material transfer of the liner. Above the thin layer of Fe, the element spectra show a gradient layer with different accumulation of different elements like Ca, S and P. In some analyzed areas, the element Zn was also found. On top of the thin Fe layer, the element P is present in higher amounts. Over this enrichment, an accumulation of the element S was detected. In comparison, Spikes [32] listed iron sulfides and zinc sulfides as the base layer on which phosphates form, and this structure is known as Zinc Dialkyl Dithio Phosphate (ZDDP).

Figure 22 shows a condensed tribological model which outlines the main steps from the damage process.

8. Conclusions

In this study, a model test method for scuffing investigations of piston ring/cylinder liner systems was developed, in which original engine parts, such as coated piston rings, can be used. Furthermore, the created damage mechanisms of the model test method were validated by comparison with damage phenomena of fired engines. With the used linear tribometer TE77, a precise control of the speed, normal force and system temperature and high-resolution measurements from contact near temperature, COF and CP can be guaranteed. The tilted system and precise pumps allow us to adjust the very low and continuous flow of heated oil through the tribological contact.

The following conclusions can be drawn:

• For the test method, a linear tribometer with a special designed tilting mechanism resulting in a very low and continuous flow of heated lubricant through the tribological contact from the higher-lying TDC to the lower-lying BDC was used.
• Sufficient validation between the damage mechanism from field engine parts in comparison to the damage mechanism from the model test was carried out to evaluate whether the same damage phenomena occur.
• The emergence of the cracks and breakouts on the liner was analyzed step by step until scuffing occurred. As a result, the damage on the liner surface is comparable with the previous work of Pusterhofer et al. [11], Zhang et al. [10] and Saeidi et al. [50], and fine cracks were increasingly found in the vicinity of the graphite. From there, the crack network and the breakouts spread.
• Also, tribopads and the material transfer behavior were observed. As a result, a thin layer of iron phosphate forms on the surface of the CKS coating with zinc sulfides only present in selected places. A phosphate layer forms on top of this and the sulfur content accumulates on the surface of the tribopad. The fact that zinc sulfides are only present in selected places differs from the work of Spikes et al. [32] and the ZDDP model.
• In addition, a targeted evaluation of the crack network on the coated piston ring surface was carried out. A furnace heating test concluded that vertical cracks mostly occur due to residual stresses. The coated piston ring forms at the model test horizontal cracks caused by shear stresses in the beginning. Due to the gradual increase in test load during the test, vertical cracks also form in the later stages of the test, which are caused by temperature input (flash temperatures) and the formation of cracks due to residual stresses.

Based on the findings of this study, which focus on both the liner and piston ring, further investigations can be conducted to develop a tribological system that delivers high overall performance.