Squeeze film is a term denoting a hydrodynamic film that sustains a negative (1) (h is the film thickness, t is time) i.e. when the opposing surfaces squeeze together. An extremely useful characteristic of squeeze films is that they provide increased load capacity (although temporary) when a bearing is suddenly subjected to an abnormally high load, while another characteristic of squeeze films is that the squeeze film force is always opposite in direction to the motion of either bearing surface [1]. Evidence of squeeze film effect can be found in the minimum oil film thickness (MOFT) measurements where one can notice that the measurement profile is not symmetric. As Bolander et al [2] have pointed out, the point of absolute minimum of the oil film thickness measurement is shifted a few degrees from the TDC and BDC. Friction peaks correspond to asperity interaction, contact Aus Wissenschaft und Forschung 4 Tribologie + Schmierungstechnik · 69. Jahrgang · eOnly Sonderausgabe 2/ 2022 DOI 10.24053/ TuS-2022-0032 1 Introduction The study of overall load carrying capacity of the piston-ring and liner lubricated interface leads to the tribological explanation and evaluation of the different parameters affecting the lubricated conjunction and the surface interaction. The main focus is on optimising lubrication and promote effective lubrication of the surfaces in contact. As part of this, friction reduction, cavitation initiation and development, which in turn limits the load carrying capacity and measurement of the oil film thickness, flow rate and oil film pressure has become a priority that eventually leads to emissions reduction. The saviour in lubrication terms is the squeeze film effect at low velocities where a thick enough film to sustain the load capacity is non-existent. According to Stachowiak and Batchelor [1], an extremely useful characteristic of squeeze films is that they provide increased load capacity (although temporarily) when a bearing is suddenly subjected to an abnormally high load. As regards to the motion of either side of the bearing surface, the squeeze film force is always opposite in direction to their motion. 2 Squeeze Film: Minimum oil film thickness (MOFT) - Friction force measurements - Load capacity of piston-ring Bearings that are subject to transient dynamic loads (such as engine crankshaft bearings) are likely to withstand much larger instantaneous maximum loads than the ones derived from steady state analysis. The reason behind this, is that there is not sufficient time to develop pressure capable of spreading the lubricant layer before total load is reduced. This phenomenon is causing an obvious lubricant film “stiffness” as acting loads impose a continuously thinner film. Squeeze Film Investigations in a Simulating Piston-Ring Cylinder Liner Experimental Set-up Polychronis Dellis* The importance of these investigations lies to the combination of experimental results with cavitation initiation investigations and its development after the dead centers of the stroke as well as rheological behavior of different chemical additives with a view to establishing the likely performance gains in new lubricant formulations. Lubricant formulation plays an important role because at higher temperatures lubricant additives have a different interaction with contacting surfaces and in this manner the resulting effect of asperity contact is either increased or reduced. Keywords squeeze film, piston-ring lubrication, experimental test-rig, friction, oil film thickness, energy losses, lubrication modeling Abstract * Polychronis Dellis School of Mechanical Engineering Educators, ASPETE, Neo Iraklio, 15122, Athens, Greece between the piston-ring and liner surfaces where the boundary lubrication prevails. The lubricant begins to squeeze out of the contact area while the pressure generated through this squeezing motion shifts the profile towards the center line [2]. The interpretation of the minimum oil film thickness measurements is important as it identifies the lubrication rheology phenomena that in turn, affect the load capacity of a certain piston-ring configuration. It was also noted that the profile of the minimum film thickness is not symmetric. The point of absolute minimum is shifted a few degrees from TDC and BDC due to the squeeze film effect. This effect is also seen in the friction signals causing an asymmetry in friction spikes at the ends of the stroke [2]. Theoretical predictions for a Newtonian fluid have determined the effect of a number of independent system parameters on performance for an idealised system involving a Newtonian fluid and perfectly smooth surfaces [3]. The squeeze velocity is studied as a dependent system parameter, derived from the simulation of independent parameters such as contact geometry, angular speed, load and viscosity and surface parameters. The numerical parametric study of the system for the Newtonian fluid was derived with the Swift-Stieber boundary condition imposed on the Reynolds equation. Cavitation delay for specific oils was also considered for a range of boundary conditions and were quantified. With a magnitude of the sinusoidal velocity reaching 0.79, 1.31 and 1.83 m/ sec respectively for rotational speeds of 300, 500 and 700 rpm at the idealised single ring test rig system, the minimum cyclic film thickness is occurring a few degrees crank angle after the reversal due to the squeeze action. The impact of combined effects of speed and dynamic load results is movement of film thickness to be shifted towards BDC [3]. Dynamic effects due to the variation of speed depict the rate of change of film thickness at different speeds. This rate of change represents the squeeze velocity and the finding from the simulation was that increasing the speed from 300 to 500 rpm does not significantly alter the crank angles at which squeeze velocity changes sign, except in the vicinity of 90° CA where the maximum film thickness is reached later in the stroke at 500 rpm. Cross over points at 700 rpm near mid-stroke were found to be clearly shifted towards BDC due to the reduced dynamic load. Power losses due to the squeeze action explain why the power loss at the dead centers are very small but not quite equal to zero [3]. Different modelling conditions showed that in the vicinity and after the reversals, where the squeeze action is dominant, the prediction using the separation boundary condition is not significantly different from that with the Swift-Stieber condition. More knowledge is required on the onset of cavitation in dynamic systems before the mechanism of delayed cavitation that leads to thinner films than predicted when applying the Swift-Stieber condition, can be accurately modelled. Some further understanding can be obtained by comparing the measured friction with the values derived from simulation. The Swift Stieber and separation conditions lead to underestimation of the friction simulated values whereas Coyne and Elrod results is an overprediction. This is due to a corresponding increase of the shear rate in the fluid and the extent of the film over which the lubrication shearing occurs [3]. For the model applied (Greenwood and Tripp), asperity interaction reduces the rate of change of film thickness near the reversals when asperity interaction occurs. Squeeze action is also inhibited as increasing amount of the load is carried by asperities. This is evident in the (2) (h¯ T is the average film thickness (separation) in the mixed lubrication model) sign change to occur very close to the dead centers [3]. It was also stated that application of the quasi-steadystate assumption to an inherently dynamic system should be expected to lead to a degree of error, especially in those parts of the cycle where cavitation regions start to form, i.e. close to the dead centers of the stroke. Frictional losses are increasing due to extensive cavitating regions. The geometry of the piston rings contributes to different cavitation sizes and what needs to be clarified is whether the cavitation development at the beginning of the stroke plays a significant role in the friction peaks at boundary lubrication region [4]. The initiation of cavitation which is affected by speed, load, ring geometry, temperature and lubricant chemistry is affecting the squeeze film as experimental data have shown. Cavitation affects the squeeze film forces by the formation of compressible bubbles in the imcompressible lubricant. The appearance of bubbles is due to a much slower rate of bubble dissolution as compared to the rate of bubble formation. Eventually load capacity is reduced compared to the assumption of no cavitation effects [4]. The lubricant reformation is dependent on squeeze film effect-therefore when this is significant the lubricant reforms earlier. At these low velocities the hydrodynamic action is not sufficient to sustain the thick film and the lubricant begins to squeeze out of the contact as the surfaces move closer. The squeeze pressure is independent of the asperity contact pressure that begins to increase as the film continues to drop [1]. It has been verified that different forms of cavitation appear after the dead centers of the stroke that accompany the squeeze film (which is measured from the capacitance signal and also evidenced by the friction peaks at the same stroke area). Aus Wissenschaft und Forschung 5 Tribologie + Schmierungstechnik · 69. Jahrgang · eOnly Sonderausgabe 2/ 2022 DOI 10.24053/ TuS-2022-0032 a view to establishing the likely performance gains in new lubricant formulations. 2.2 Results Viscosity variation, physical and chemical properties are correlated to the measured friction force, at the boundary, mixed and hydrodynamic lubrication region of the stroke. Meanwhile, geometrical factors such as pistonring curvature variation, affect total friction, oil film thickness and oil film pressure measurements as well as the squeeze film effect. When the squeeze film effect is studied for lubricants with different HTHS viscosities, it was noticed that the oil with the higher HTHS viscosity produces a thicker film at all temperature testing with the squeeze film moving further down the stroke for the lubricant with the higher HTHS viscosity. For a specific oil, a similar trend was noticed for the squeeze film in oil film thickness curves at all testing temperatures, for the same speed and load test cases. In a set of experiments focused on high temperature testing, high friction peaks were noticed when oil viscosity changed to lower values as lubricant temperature increases and MOFT decreases significantly [7]. At higher temperatures the asperity interaction at the boundarymixed lubrication region is intense giving considerably higher friction results than the ones taken at lower temperatures. As MOFT decreases with high temperature, friction force peaks move closer to the dead centers of the stroke with absolute friction values that are significantly higher. This gives evidence that the squeeze film effect does not have such a strong impact at high lubricant temperatures. In previous publications it was shown that for the MOFT measurements, high load testing is combined with squeeze film movement towards the dead centers of the stroke. High temperature testing showed that the MOFT decreases significantly from ambient temperature (33 °C) to 50 °C and that the squeezing action is getting Aus Wissenschaft und Forschung 6 Tribologie + Schmierungstechnik · 69. Jahrgang · eOnly Sonderausgabe 2/ 2022 DOI 10.24053/ TuS-2022-0032 2.1 Experimental set-up In a simplified single-ring test rig, a steady piston-ring section is placed under a flat surface used as a reciprocating liner. The idealised simulation test rig benefits from simplified lubrication conditions compared to the real engine taking advantage of the simple design layout. As a result, solid and repeatable results are taken allowing the lubricant film characteristics to be examined in isolation. Sensors that measure oil film pressure, thickness (optical - LIF and electrical - capacitance), friction and imaging, provide the necessary parametric data to study the effect of speed, load, temperature, piston-ring curvature and variable lubricant properties. When the liner decelerates, the interface reaches a state of mixed lubrication and asperity interaction until the liner reaches boundary lubrication close to the dead centers, as the squeeze film prevails. As the liner accelerates away from the dead centers, the lubricated film begins to develop. While being close to the dead center, asperity interaction between the surfaces remains significant with the squeeze film effect also taking place, resulting in beneficial oil support as it is supported partly by the lubricant present in the contact [4]. Increase in temperature in a lubricant model investigation had the effect of decrease in oil film thickness and advanced the initiation of cavitation and enhanced its intensity [5]. Less viscosity results in less squeezing force from the oil around the dead centers and thus greater asperity contact force is generated to support the radial ring load. Less oil squeezing force is responsible for more asperity contact around the dead centers [6]. For the modelling performed the Swift-Stieber boundary condition imposed on the Reynolds equation resulted in good agreement but for a certain level of high temperature/ high shear (HTHS) viscosity coefficient the measured lubricant films after the reversals were thinner than predicted due to cavitation delay, i.e. the period within the cycle that the lubricant experienced absolute tension. The mixed lubrication models for both Newtonian and non-Newtonian shear thinning fluids were shown to be sensitive to the asperity interaction sub-model [3]. The purpose of this study is to show the effect of squeeze film variation and extract useful parametric results that show how different lubricants and setups impact on friction peaks / losses, correlate and verify them to other measurement techniques for the single ring set-up (such as MOFT measurements). Cavitation initiation and development is another factor that should be taken into account and assess whether cavitation development at the beginning of the stroke together with impeding or aiding factors, play a significant role in friction peaks and MOFT minima. Eventually, a clearer picture will be attained to the aspects of load carrying capacity of the ring and the rheological behaviour of chemical additives with Figure 1: Temperature effect on friction force peaks at 300 rpm, 3371 N/ m load, top dead center [7] marginal. The same action is shown in Figure 2 for the friction peaks [8] at 300 rpm, 971 N/ m and 70 °C. For a set of different lubricants, the properties of which can be found in Table 1, friction force peaks have different behaviour close to the dead centers. Figure 2 shows that for similar speed and load and temperature testing conditions friction force peaks move closer to the dead centers for the lubricant that has the lowest VI, V 40, V 100 and HTHS viscosity. For the same set of lubricants and same testing conditions, the measurements close to top dead center can be seen in Figure 3. Upstroke and downstroke measurements, in terms of cavitation area size and ring geometry (symmetric or non symmetric) can show some discrepancy. As it was shown in [9] the aforementioned factors can have a significant effect. In the case of the simulating single-ring test rig, motion dynamics and location of the liner in relation to the electric driving motor rotating shaft had an effect on the measurements and dynamic load on the liner due to the linear motion. This was quantified with an equation that contained the angular position [7]. For the set of lubricant tested, it was verified that oil 2A had the lowest friction peak at the boundary lubrication region and its location moves further in the stroke compared to the other three lubricants of the test matrix. Oil 2A had the highest V 100 , V 40 and HTHS viscosity according to Table 1. The friction signal for the lubricants tested as in Table 1 can be seen in Figure 4. This figure represents the friction signal for the whole stroke. It can be inferred that in terms of power losses, the most important factor is the behaviour of the lubricants at the dead centers, with the friction peaks’ variation being the major indicator of the friction losses. Hydrodynamic losses also play an important role as they spread throughout the majority of the stroke, but in magnitude are relatively even. Testing conditions are at high temperature 70 °C, 300 rpm and 971 N/ m load. It has been verified that different forms of cavitation appear after the dead centers of the stroke that accompany the squeeze film which is measured in the capacitance and friction signals [5, 7, 8, 10]. The geometry of the piston-ring affects the friction force as well. The flatter the piston-ring, the lower the friction force peak and they also appear earlier in the stroke [7]. In Figure 5, the shift in squeeze film close to BDC for oil 6E, at 400 rpm and 1159 N/ m load can be seen. There is a marginal shift closer to the dead center for 50, 60 and 70 °C, compared to the ambient temperature curve (35 °C). Further testing of oils in Table 1 showed the results in Figure 6. A slight shift towards BDC can be evi- Aus Wissenschaft und Forschung 7 Tribologie + Schmierungstechnik · 69. Jahrgang · eOnly Sonderausgabe 2/ 2022 DOI 10.24053/ TuS-2022-0032 Figure 2: Lubricant properties effect on friction force peaks at 300 rpm, 971 N/ m load, at bottom dead center Figure 3: Lubricant properties effect on friction force peaks at 300 rpm, 971 N/ m load, at top dead center Figure 4: Friction variation, for the whole stroke length all oils tested, at 300 rpm, 971 N/ m and 70 °C Blend Code 003B 006E/ 02 005A/ 02 002A/ 02 Grade 0W-30 0W-40 0W-20 10W-40 HTHS(mPas) 3.30 3.4 2.14 4.05 V 100 (cSt) 12.16 12.8 6.04 14.97 V 40 (cSt) 68.93 66.8 31 97.8 VI 182 196 146 160 Table 1: Oils tested for temperature-friction investigations boundary, mixed and hydrodynamic lubrication region of the stroke. - Large radius of curvature for the ring profile promotes effective squeeze action at the ends of the stroke, as the flatter ring enhances a stronger squeeze effect than the curved ring at the dead centers [7]. - Different oil blends produce different appearance for the friction peaks in terms of their distance from the dead centers and an obvious absolute friction peak measurement. - For the cavitating region of the lubricant, parameters such as speed, load and temperature affect its initiation and furthermore the number of the specific form of string cavities might accordingly apply to temperature effect results. - Low MOFT decreases the wedging action at the converging profile providing a lower load carrying capacity. - For the set of four lubricants tested, the one with the highest V 100 , V 40 and HTHS viscosity seems to shift the squeeze film further in the stroke. It remains to be seen how is this behaviour established, in a set of lubricants to be tested that have same SAE grade as lubricant 2A and relatively different HTHS, V 100 , V 40 and VI values. The importance of these investigations lies to the combination of previous experimental results with cavitation initiation investigations and its development after the dead centers of the stroke as well as rheological behavior of different chemical additives with a view to establishing the likely performance gains in new lubricant formulations. Lubricant formulation plays an important role because at higher temperatures lubricant additives have a different interaction with contacting surfaces and in this manner the resulting effect of asperity contact is either increased or reduced. References [1] Stachowiak G.W. and Batchelor A.D., “Engineering Tribology”, ELSEVIER, 1993. [2] Bolander, N. W., Steenwyk, B. D., Sadeghi, F. and Gerber, G. R., “Lubrication Regime Transitions at the Piston Ring - Cylinder Liner Interface”, Proceedings of the Institution of Mechanical Engineers Part J: Journal of Engineering Tribology, 219, No 1, 19-31, 2005. [3] Ostovar, P. “Fluid Aspects of Piston Ring Lubrication”, PhD thesis, Imperial College of Science, Technology and Medicine, 1996. [4] Dellis P.S., “Piston-ring performance: limitations from cavitation and friction, International Journal of Structural Integrityˮ, Vol. 10 No. 3, pp. 304-324, 2019. [5] Nouri J. M., Vasilakos I., Yan Y., “Cavitation between cylinder-liner and piston-ring in a new designed optical IC engine”, Int. J. of Engine Research, (on line first) 9 Apr 2021. Aus Wissenschaft und Forschung 8 Tribologie + Schmierungstechnik · 69. Jahrgang · eOnly Sonderausgabe 2/ 2022 DOI 10.24053/ TuS-2022-0032 denced from the magnification over the area of interest. The oils that the corresponding squeeze film moves towards the dead center are 2A and 3B, with oil 3B being the most prone to alter its squeeze film measurement towards the dead center. 3 Conclusions - The squeeze film effect between the liner surface and the piston ring, shift the friction force peaks, as it forces the lubricant flow to delay compared to the liner movement. - With an increase in load in every lubricant the flow reversal due to the squeeze film effect appears closer to the dead centers. That could be also attributed to the fact that the viscous film impedes the liner’s reciprocation. - Load capacity is affected by the cavitating region. - With an increase in reciprocation speed for constant load, friction force maxima have lower absolute measurements and appear at a greater distance from the dead centers. - Viscosity variation, physical and chemical properties are correlated to the measured friction force, at the Figure 5: Temperature effect on MOFT curves at BDC, for oil 6E, 400 rpm, 1159 N/ m Figure 6: All oils tested-squeeze film behaviour at 600 rpm, 1159 N/ m load 70 °C [6] Tian, T., Wong, V. W., and Heywood, J. B. “A Piston-Ring Pack Film Thickness and Friction Model for Multigrade Oil and rough surfaces”, SAE paper 962032, 1996. [7] Dellis P., “Effect of Friction Force between Piston Rings and Liner: a Parametric Study of Speed, Load, Temperature, Piston-Ring Curvature and High-Temperature, High- Shear Viscosity”, Proc IMechE, Part J: J Engineering Tribology, 224, No 5, 411-426, 2010. [8] Dellis P., “Oil Film Thickness Measurements Combined with High Temperature Friction Investigations in a Simplified Piston-Ring Lubrication Test Rig”, Tribology in Industry, 41 No. 4, 471-483, 2019. [9] Dellis P., “Cavitation development in the lubricant film of a reciprocating piston-ring assembly”, Proc. IMechE, Part J: J. Engineering Tribology, 218 No 3, 157-171, 2004, DOI: 10.1243/ 1350650041323340. [10] Dellis P., “Cavitation initiation and patterns in engine lubricants as a result of different operating conditions and lubricant properties”, STLE Virtual Annual Meeting and Exhibition, May 17-20, 2021, New Orleans, USA. Aus Wissenschaft und Forschung 9 Tribologie + Schmierungstechnik · 69. Jahrgang · eOnly Sonderausgabe 2/ 2022 DOI 10.24053/ TuS-2022-0032