New insight to the tribology-structure interrelationship of lubricating grease by a rheological method

Abstract

By controlling the heat-treatment process, three types of lithium greases with different thickener fiber morphologies were synthesized via the saponification reaction. With increasing the cooling rate, the dimension of the thickener fiber decreased. The relationship between microstructure and tribological performance of the lubricating greases was investigated via a rheological method. The results indicated that the fiber dimension determined the level of physical entanglement and the evolution of fiber network in the friction process, further influencing the final lubricity. The grease with the large fiber dimension displayed good tribological performance under low frequency and high load conditions due to its large-scale strongly crosslinked structure. In addition, the grease with the small fiber dimension yielded a low level of fiber physical entanglements and displayed low structure strength and fast response in the rheological test. It could obviously improve the tribological performance under high frequency conditions. Through tuning the microstructure of the thickener fiber according to the lubricating conditions, the tribological performance could be obviously improved.

---

1. Introduction

As the core parts of great equipment, rolling bearings play a crucial role in the equipment manufacturing industry due to its broad applications in the field of aviation, aerospace, national defense, and transportation. In fact, about 90% of rolling bearings are lubricated with grease.¹ Over the past decade, the investigation of grease has attracted intense research attention, which was mainly focused in two areas: the innovation of research methods^2–10 and the improvement of test equipment.^11–16 Among these works, the search for a novel research method to reveal the relationship between the microstructure and tribological performance of lubricating grease plays a crucial role in developing high performance and long service life lubricant. It is well known that greases are highly structured colloid systems.^2,3,10 Both the rheological properties and tribological performances are intrinsically determined by the base oil (e.g. viscosity and functional group) and the thickener (e.g. concentration, morphology and the fiber network structure).

In the previous studies,^2–4,7,8,10 the influence of grease microstructure on the rheological performance was widely investigated. It has been demonstrated that the rheological properties can be controlled via tailoring the composition^2,3 and manufacturing process.^10,17 However, the influence of microstructure on the tribological performance has not been well studied. As we know, lubrication can be interpreted as a dynamic evolution process of the grease microstructure, which is mainly a deformation and flow phenomenon. The rheological behavior will directly influence the distribution and evolution of grease in the lubrication point and determine the final tribological performance. Therefore, the investigation on the rheological behavior of lubricating grease may contribute to elucidate the way in which the lubricating grease works under different operating conditions.

In this work, we study the interrelationship between microstructure and the final tribological performance of grease via a rheological method. Based on the same base oil, three greases with different thickener fiber morphologies were synthesized by controlling the cooling rate. The corresponding rheological behaviors were characterized to reveal the feature of flow such as the structure strength and response rate. Based on the investigation on the relationship between rheological properties and tribological performances, the influence of microstructure on the tribological performance was clarified.

2. Experimental section

2.1. Synthesis of lithium grease

The lithium grease was prepared as follows: proper amounts of base oil (PAO10, kinematic viscosity at 40 °C: 61 mm² s⁻¹) and 12-hydroxy-stearic acid were added to a container with stirring at 90 °C. Then, proper amounts of lithium hydroxide aqueous solution were added into the abovementioned solution. After the saponification at 100–120 °C for 3 h, the mixture was heated to evaporate water at 160 °C, followed by the addition of proper amounts of PAO10 at 200 °C. Finally, the weight percentage of lithium soap was 12 wt%. Then, the mixture was ground using a triple-roller mill three times. The cooling rate was controlled by controlling the amount of cold oil. The synthesized greases were donated as LG, MG and HG according to the cooling rate and the corresponding properties are reported in Table 1.

Table 1 Typical properties of synthesized grease under different cooling rates

Properties	LG	MG	HG	Test method
Cold oil (g)	30	40	60
Cone penetration (0.1 mm)	202	191	223	ASTM D217
Roll stability	+32	+18	+30	ASTM D1831
Dropping point (°C)	217	220	218	ASTM D556

---

2.2. Characterization and property measurements

The morphologies of the grease thickener fiber were characterized via a field emission scanning electron microscope (FE-SEM, JEOL JSM-6701F). A small amount of synthesized product was immersed in the petroleum ether to extract the oil. The solution was then dropped onto a copper grid. Afterward, the copper grid was coated with gold and characterized via FE-SEM.

Rheological measurements were carried out using an Anton Paar MCR302 rheometer (Austria) using plate to plate geometry (24.985 mm diameter and 1 mm gap). In the oscillatory measurements, an amplitude sweep test at a frequency of 1 Hz was performed to determine the linear viscoelastic region, prior to the following frequency sweep test (from 0.1 to 600 rad s⁻¹). In particular, stress relaxation test was conducted to characterize the time to reach the equilibrium state under constant strain (10%). In the steady shear experiments, the thixotropy shear test was performed to investigate the recoverability after exposure to a high instantaneous shear rate (from 500 to 3000 s⁻¹).

The tribological performances of the synthesized greases were evaluated via an Optimol-SRV IV oscillating friction and wear tester. The test was conducted in a conventional reciprocating “ball-on-block” mode with an oscillating upper ball (AISI E52100 steel, 10 mm in diameter, HV 710-730) and a fixed lower disc (AISI E52100, ø 24 mm × 7.9 mm, HV 710-730). The test conditions were mainly designed to investigate the difference in tribological performances. To evaluate the anti-wear property, a MicroXAM 3D non-contact surface mapping profiler was employed to measure the wear volumes of the wear scars on the lower disc.

3. Results and discussion

3.1. Materials characterization

As shown in Fig. 1, the morphologies of the grease thickener fiber under different cooling rates were characterized via FE-SEM. The synthesized lubricating lithium greases are a highly structured colloid system with thickener fiber randomly distributed in the base oil. These structural units are entangled to form a complex network. However, important differences in the structural skeleton (the entanglement level and dimension of thickener fibers) can be observed. With increasing the cooling rate, the fiber dimensions decrease evidently. It is well known that dimension has a direct influence on the flexibility and interaction of thickener fibers.⁵ The thickener fiber of grease under a low cooling rate (LG) displayed a loose entanglement structure due to its greater rigidity from larger dimension (Fig. 1a). With increasing the cooling rate, the fiber dimension decreased and resulted in the improvement of flexibility. Then a compact network was formed, as shown in Fig. 1b. However, after further increasing the cooling rate, the formation of a crosslink network in the large range was obviously restricted due to the weak interaction of the short fiber (Fig. 1c). Thus, MG exhibits a relatively high level of entanglement, compared with HG and LG.

Fig. 1 FE-SEM images of grease thickener fibers under different cooling rates: (a) LG, (b) MG, and (c) HG.

3.2. Rheological properties

In this study, a rheological method was used to provide more detailed information about the responsiveness of the microstructure under different conditions. Based on the same base oil, it mainly depends on the morphologies of thickener fibers. As shown in Fig. 2a, amplitude sweep experiments were carried out first to locate the borders of the linear viscoelasticity (LVE) region, where the stress is proportional to the strain. It is well known that the modulus G′ is a measure of the energy stored and is recovered per cycle, while the modulus G′′ is a measure of the energy dissipated or lost as heat per cycle of sinusoidal deformation.¹⁸ The relative strength of the thickener fiber entanglement could be reflected by the measurement of G′ and G′′. It was observed that the varying trends of the modulus in the test range were basically similar. In the LVE region, the G′ and G′′ are stable due to that the grease microstructure remains close to its equilibrium state or initial state. By further increasing the shear strain, the modulus to decrease sharply and there exists a structure transformation point from the solid-like (G′ > G′′) to liquid-like (G′ < G′′) outside the LVE region, indicating the destruction of the crosslink network. For the grease with the high entanglement level, the shear stress (τ) of the transformation point was high. This indicates that the MG shows the optimal shear resistance, which is in accordance with the results of roll stability (Table 1). In addition, the G′ and G′′ increased initially and decreased later with the decrease of fiber dimension (diametrical size and axial size). The corresponding sequence of structure strength was MG > LG > HG, which is in accordance with the FE-SEM observation and the results of cone penetration. The greases with large fiber dimension always possessed a stronger microstructure due to the high entanglement level. However, the structure strength will become low when the fiber diametrical size is too large due to poor flexibility, as shown in Fig. 1a. In addition, the HG exhibited the lowest structure strength due to the restricted interaction of the small axial size.

Fig. 2 (a) Storage modulus (G′), loss modulus (G′′) and shear stress (τ) of the synthesized greases as a function of amplitude. The linear viscoelasticity region is also indicated (▲). (b) Frequency dependence of storage modulus (G′) and loss modulus (G′′) of the synthesized greases.

For more information about the microstructure of the synthesized grease, frequency sweep tests were carried out under constant strain (5%), as shown in Fig. 2b and 3. Fig. 2b shows the curves of G′ and G′′ as a function of frequency. It can be observed that the G′ and G′′ of synthesized grease increased with the increase of entanglement level, which is in accordance with the results of amplitude sweep experiments. With increasing the frequency, the modulus G′ increased stably and the modulus G′′ showed a similar trend in the low frequency range. For the high frequency condition, the G′′ of HG and LG decreased sharply. It is well known that the modulus G′′ reflects the energy dissipated. Based on the stable G′ value, the decrease of G′′ might be due to the microstructure adjustment rather than destruction. For further investigating the influence of frequency on the microstructure, the complex viscosity curves under different frequency ranges are shown in Fig. 3. The data fitting procedure was also performed through the power-law model (Ostwald-de Waele model) (r² > 0.99):

where k and n are the consistency and flow indexes, respectively. The consistency index increases with the structure strength and the flow index could reveal structure changes. Fig. 3a shows the curves in the low frequency range. It can be observed that the consistency indexes increased with the structure strength and the flow indexes showed a contrary tendency. For the high frequency range (Fig. 3b), the flow indexes of HG and LG decreased significantly. The results indicate that the structure adjustment of HG and LG are prone to occur due to their low structure strength.

Fig. 3 The complex viscosity of the synthesized greases as a function of angular frequency.

In the actual friction process, the nonlinear viscoelastic (non-LVE) model is more suitable compared with the LVE model. For revealing flow features and the evolution of microstructure outside the LVE region, the internal structure relaxation time becomes an essential parameter,^3,19,20 which could reflect the structural rearrangement speed or stress–strain response. As shown in Fig. 4, the stress relaxation tests under constant strain (10%) were conducted, which are quantified by introducing the curvature. The time corresponding to the maximum curvature values could reflect the time to reach equilibrium state to some extent. It can be observed that the greases with low structural strength display a quick response. The varying tendency of response was HG > LG > MG, which is contrary to that of structural strength.

Fig. 4 The relaxation modulus and curvature of the synthesized greases as a function of time.

For further verifying the results mentioned above, thixotropy shear tests were performed, as shown in Fig. 5. By comparing the viscosities before and after high-speed instantaneous shear, the influences of fiber morphology on the thixotropy were studied. Under a low instantaneous shear rate (500 s⁻¹), shear stress is not destructive to network structure. The viscosities could recover to the initial state after structural rearrangement as shown in Fig. 5a. For the high instantaneous shear rate condition, the corresponding thixotropy largely depends on the response speed of the grease microstructure. It can be observed that the grease with quick response displays good recoverability (Fig. 5b and c). For the grease with high structure strength, the adjustment speed of the thickener structure was too low for the instantaneous high shear rate, which resulted in the structure destruction. The corresponding recoverability decreases sharply with the increase of shear rate (Fig. 5d).

Fig. 5 The thixotropy shear measurements of the synthesized greases under different instantaneous shear rates.

3.3. Tribological properties

Based on the characterization of the grease microstructure via a rheological method, the study herein focuses on the influence of microstructure on the tribological performance. Fig. 6 presents the friction coefficient curves under different frequencies from 25 to 70 Hz at a constant applied load (50 N). Under a low frequency, the greases with different thickener fiber structure gave the same friction coefficient and displayed a good friction-reducing property except for some instantaneous seizure (Fig. 6a). When the frequency was increased up to 40 Hz, obvious differences in the tribological performance could be observed. The curve corresponding to HG displays a sharp fluctuation in the initial stage, while the greases (LG and MG) with a high level of fiber entanglement show stable and low friction curves for the entire duration of the test. Further increasing the frequency, it is interesting to observe that the dynamic friction coefficient curves recover to the stable state and coincide with each other. For verifying these results, similar tests under a high applied load (100 N) were performed, as shown in Fig. 8. It can be observed that the tendencies are basically similar.

Fig. 6 Dynamic friction coefficient curves of the synthesized greases under a fixed applied load (50 N) and different frequencies (a, 25 Hz; b, 40 Hz; c, 55 Hz; d, 70 Hz).

Fig. 7 Wear volume of the lower disc lubricated by the synthesized greases under a fixed applied load (50 N) and different frequencies.

Fig. 8 Dynamic friction coefficient curves of the synthesized greases under a fixed applied load (100 N) and different frequencies (a, 25 Hz; b, 40 Hz; c, 55 Hz; d, 70 Hz).

For better understanding the phenomenon, the anti-wear properties were investigated, as shown in Fig. 7. At low frequency, the greases with stronger structure (LG and MG) display the optimal anti-wear properties, particularly at the frequency of 40 Hz. With increasing the frequency, the anti-wear property of HG improved significantly, while the wear volumes of LG and MG continued to increase. When the frequency was increased up to 55 Hz, there exists a performance change point and the grease with low structure strength (HG) showed the optimal anti-wear property, compared with LG and MG. The change trends, as shown in Fig. 9, are almost the same, except that the performance change point of LG transfers to higher frequency (55 Hz).

Fig. 9 Wear volume of the lower disc lubricated by the synthesized greases under a fixed applied load (100 N) and different frequencies.

Tables 2 and 3 show the three-dimensional morphologies of the wear scars and the wear scenario under different test conditions can be clearly observed. There exists an optimal application range for all the synthesized greases. Outside the range, the corresponding wear scar is very wide and deep, indicating serious wear occurred. Moreover, the change point can also be observed.

Table 2 3D images of the wear scars under different frequencies and a constant applied load (50 N)

	LG	MG	HG
50 N, 25 Hz
50 N, 40 Hz
50 N, 55 Hz
50 N, 70 Hz

---

Table 3 3D images of the wear scars under different frequencies and a constant applied load (100 N)

	LG	MG	HG
100 N, 25 Hz
100 N, 40 Hz
100 N, 55 Hz
100 N, 70 Hz

---

For better understanding the mechanisms, frequency ramp tests were conducted, as shown in Fig. 10. Under the applied load of 10 N (1.22 GPa), the friction coefficient curves of all synthesized greases decrease initially and increased later with the increase of frequency (Fig. 10a). This tendency is in accordance with the stribeck curve, as shown in Fig. 10d. It can be observed that the lubricating state undergoes a shift from boundary lubrication to mixed lubrication and then to hydrodynamic lubrication. These results indicate that the lubricating condition tends to transform to the hydrodynamic state with increasing frequency. Through the comparison of the morphologies of the wear scar in the Tables 2 and 3, the transformation of lubricating conditions can be directly observed. In addition, the transformation trend is gradually weakened with the increase of applied load. The overall lubricating state gradually transfers to boundary lubrication, as shown in Fig. 10b and c. When the applied load was increased up to 100 N, an obvious transformation trend could not be observed (Fig. 10c). Based on the tests of this part, the tribological behavior in the former part can be explained. It has been demonstrated that the tribological performances largely depend upon the lubricating conditions. With the frequency below 55 Hz, it is mainly under mixed or boundary lubrication and the corresponding lubricating film thickness is relatively thin. Therefore, the strength of the film is crucial to lubricity, which depends on the structure strength of grease. Moreover, under a low shear rate, the shear action had no obvious influence on the grease structure, as shown in the rheological tests. Thus, LG and MG with the stronger structure display optimal tribological performance as shown in Fig. 6a, b, and 7. With increasing the applied load, the effect of structure strength on the tribological performances is more remarkable (Fig. 9). In addition, when the frequency was up to 55 Hz, the lubricating condition began to transform to the hydrodynamic state and there exists an obvious performance change point. The grease with the low structure strength and quick response displayed the optimal tribological performances, as shown in Fig. 6c, d, and 7. In addition, the structure of LG and MG is more prone to be destroyed under a high shear rate due to the low response rate of their high structure strength. Therefore, the performance change might be due to the fact that the effect of the response rate is superior to that of the structure strength with the increase of the thickness of lubricating film. Under a higher applied load (Fig. 9), the performance change point of LG transfers to high frequency (55 Hz). This is due to the fact that the overall lubricating state will transfer to boundary lubrication with the increased applied load.

Fig. 10 (a–c) Evolution of friction coefficient of the synthesized greases during a frequency ramp test from 10 to 70 Hz under different applied loads. (d) Typical stribeck curve showing the types of lubrication.

4. Conclusion

In summary, three types of lithium grease with different thickener fiber structure were synthesized by controlling the cooling rate. It can be concluded that the dimensions of the thickener fiber directly determine the level of entanglement. The thickener fiber of LG displayed loose entanglement structure due to its greater rigidity of larger dimension. With increasing the cooling rate, the diametrical size of the fiber decreased obviously and resulted in the improvement of flexibility. A compact network was formed. Upon further increasing the cooling rate, the formation of a crosslink network in the large range was obviously restricted due to the weak interaction of the short fiber. More importantly, the relationship between the thickener fiber structure and the final tribological performance was established via a rheological method. It can be observed that the tribological performance largely depended on the lubricating conditions. The grease with the high entanglement level displayed high structure strength but a lower response rate. While for the grease with the low entanglement level, a contrary trend was observed. The structure strength plays a crucial role in mixed or boundary conditions. In addition, the effect of response rate is superior to that of structure strength under hydrodynamic conditions. In addition, it has been demonstrated that the thickener fiber could directly participate in the friction process and the microstructure should be designed according to the lubrication conditions.

Acknowledgements

This work was funded by the National 973 Program of China (grant no. 2011CB706602) and the National Natural Science Foundation of China (grant no. 51205384).

References

1. P. M. Lugt, Grease Lubrication in Rolling Bearings, John Wiley & Sons, West Sussex, 1st edn, 2012.
2. M. A. Delgado, C. Valencia, M. C. Sanchez, J. M. Franco and C. Gallegos, Ind. Eng. Chem. Res., 2006, 45, 1902.
3. L. Salomonsson, G. Stang and B. Zhmud, Tribol. Trans., 2007, 50, 302.
4. G. Moreno, C. Valencia, M. V. de Paz, J. M. Franco and C. Gallegos, Ind. Eng. Chem. Res., 2006, 45, 4001.
5. H. Kimura, Y. Imai and Y. Yamamoto, Tribol. Trans., 2001, 44, 405.
6. E. Kuhn and C. Balan, Wear, 1997, 209, 237.
7. M. A. Delgado, J. M. Franco and E. Kuhn, Ind. Lubr. Tribol., 2008, 60, 37.
8. M. A. Delgado, C. Valencia, M. C. Sanchez, J. M. Franco and C. Gallegos, Tribol. Lett., 2006, 23, 47.
9. S. K. Yeong, P. F. Luckham and T. F. Tadros, J. Colloid Interface Sci., 2004, 274, 285.
10. M. A. Delgado, M. C. Sanchez, C. Valencia, J. M. Franco and C. Gallegos, Chem. Eng. Res. Des., 2005, 83, 1085.
11. P. M. Cannt, B. P. Williamson, R. C. Coy and H. A. Spikes, J. Phys. D: Appl. Phys., 1992, 25, 124.
12. P. O. Larsson, R. Larsson, A. Jolkin and O. Marklund, Tribol. Int., 2000, 33, 211.
13. V. Mota and L. A. Ferreira, Tribol. Int., 2009, 42, 569.
14. T. Cousseau, M. Björling, B. Graça, A. Campos, J. Seabra and R. Larsson, Tribol. Int., 2012, 53, 53.
15. G. E. Morales-Espejel, P. M. Lugt, H. R. Pasaribu and H. Cen, Tribol. Int., 2014, 74, 7.
16. L. Huang, D. Guo and S. Z. Wen, Tribol. Int. DOI:10.1016/j.triboint.2014.11.005.
17. J. M. Franco, M. A. Delgado, C. Valencia, M. C. Sánchez and C. Gallegos, Chem. Eng. Sci., 2005, 60, 2409.
18. J. D. Ferry, Viscoelastic properties of polymers, John Wiley & Sons, New York, 3rd edn, 1980.
19. H. Watanabe, S. Ishida, Y. Matsumiya and T. Inoue, Macromolecules, 2004, 37, 1937.
20. H. Watanabe, Q. Chen, Y. Kawasaki and Y. Matsumiya, Macromolecules, 2011, 44, 1570.

---