Influence of grease bleed oil on ball-on-disc lubrication

World Tribology Congress 2013 Torino, Italy, September 8 – 13, 2013

Influence of grease bleed oil on ball-on-disc lubrication

Tiago Cousseau

¹⁾

, Markus Björling

²⁾

, Beatriz Graça

³⁾

, Armando Campos

⁴⁾

, Jorge Seabra

^1)*

and Roland Larsson

²⁾

1)

FEUP, Universidade do Porto, R. Dr. Roberto Frias s/n, 4200-465 Porto, Portugal

2)

LTU, Lulea University of Technology, 971 87 Lulea, Sweden

3)

INEGI, Universidade do Porto, R. Dr. Roberto Frias 400, 4200-465 Porto, Portugal

4)

ISEP, Instituto Superior de Engenharia do Porto, R. Dr. António Bernardino de Almeida, 431, 4200-465 Porto, Portugal

*

Corresponding author: jseabra@fe.up.pt

1. Introduction

Lubricating greases are already considered to be a

“fundamental” component of the bearing arrangement and thus, as important as the bearing, housing and sealing. However, grease lubrication mechanisms are not yet fully understood. The general consensus is that the active lubricant during the early stage of grease lubrication is governed by the releasing lubricant by bleeding, shear, or shear-induced bleeding [1]. Although, after this initial stage, the active lubricant depends of the running time, the combination bearing – grease, bearing housing, operating conditions and contaminants [2, 3].

In both situations, the rheological properties of the active lubricant are unknown. This uncertainty is one of the main barriers to model grease performance with regard to film thickness and traction coefficient as well as with regard to rolling bearing internal friction torque.

Recently, it was shown that the oil extracted from greases (bleed-oil), using the bleed test IP 121, generates the same film thickness of the corresponding lubricating grease under fully flooded lubrication [4]. It suggests that the active lubricant in the first stage of grease lubrication can be characterized using the bleed-oil properties.

Here, the initial stage of grease lubrication was simulated for 3 different lubricating greases and their base and bleed-oils, in terms of film thickness, traction coefficient and rolling bearing friction torque. A complete characterization of the greases, its base and bleed-oils were performed, and a relationship between bleed-oil properties, film thickness, traction coefficient and friction torque during the initial stage of grease lubrication was found.

2. Grease characterization

Three lubricating greases were tested: LiM1 formulated with mineral base oil and lithium thickener;

LiCaE formulated with ester base oil and thickened with lithium and calcium; and PPAO grease, thickened with polypropylene, with an elastomer as co-thickener and formulated with polyalphaolefin base oil.

The lubricating greases, their base oils and bleed-oils were rheologically characterized with an AR 1000-N rheometer from TA Instruments with rough plate-plate geometry (2R = 25 mm; Ra ≈ 20 µm) at two different temperatures (40 ºC and 80 ºC). While all base and bleed-oils behaved as Newtonian, the bleed-oil of PPAO presented a shear-thinning behavior, as presented

in Figure 1, where the viscosities measured at 80 ºC are shown as markers and the fitting curves, obtained with Carreau model (Eq. 1), as continuous lines.

The low-shear viscosity, Carreau adjustable factors (Flow index -n and Critical stress -G

_cr

) and other grease characteristics are presented in Table 1.

 

 

1 2 2 1 2

1

2

1



 





 





 

 



 









n

G

cr



 





 ^

(1)

Figure 1 Base and bleed-oil viscosity at 80ºC

Table 1 Physical characteristics of lubricating greases, their base and bleed-oils

name LiM1 LiCaE PPAO

Base oil Min Est PAO

Thickener Li Li+Ca PP

Base Oil Properties

Viscosity

40ºC

-mm

²

/s 211.05 89.84 42.78 Viscosity

80ºC

-mm

²

/s 34.49 25.11 12.36 Piezoviscos.

_80ºC

-GPa

^-1

23.9 12.1 11.4 Spec.Gravity-g/cm

³

0.909 0.919 0.843

Bleed-Oil Properties

Viscosity

40ºC

-mm

²

/s 186.20 107.78 587.84 Viscosity

_80ºC

-mm

²

/s 30.26 26.21 159.42 Piezoviscos.

80ºC

-GPa

^-1

36.8 20.3 12.8 Spec.Gravity-g/cm

³

0.903 0.952 0.828

Flow index, n [-] 1 1 0.76

Critical stress Gcr [Pa] - - 517 Grease Properties

NLGI Nº (DIN 51818) 2 2 2

Operating temperature -20+130 -30+120 -35+120

2 3. Film thickness

The film thickness of the greases and their base and bleed-oils were measured in a ball-on-disc test rig, through optical interferometry in a WAM machine, in fully flooded lubrication. The tests were performed at 40, 60 and 80 ºC, with a maximum Hertz pressure of P

₀

=0.5 GPa and pure rolling conditions.

Figure 2 shows the comparison between greases, base oils and bleed-oils at 80 ºC for LiM1, LiCaE and PPAO, respectively. Grease and bleed-oils generated very similar film thicknesses, which are significant higher than the corresponding base oils. These trends were also observed at 40 and 60 ºC. Therefore, grease film thickness was predicted using the bleed-oil properties according to the methodology presented by Van Leeuwen [5], i.e., the film predictions are based on determining the pressure-viscosity coefficient (α

_film

) that gives the best R

²

fit with the experimental film thickness values. Katyal and Kumar [6] film thickness equation (Eq. 2) was used to obtain α

_film

because it takes into account the shear-thinning effect under pure rolling conditions. The values used (viscosity, flown index and critical stress) to calculate the base and bleed-oil film thickness and the obtained α

_film

are given in Table 1. The predictions are shown as continuous lines in Figure 2.

R W

G U

R

H

_0c

 1 . 098538 

_x



^0.652



^0.557



^0.0415



 

^{1 2}

2 . 1

92 . 1 77 . 0

264 . 0 69 . 0

325258 .

1 1

n

G

cr

G W R U





 

 







 



 (2)

4. Traction coefficient

The traction coefficients of the lubricating greases were also measured in the WAM machine under fully flooded and starved lubrication. The tests were performed at 40, 60 and 80 ºC, with a maximum Hertz pressure of P0=1.86 GPa, SRR from 0.05 to 5.45% and entrainment speed from 0.37 to 2.8m/s.

Figure 3 shows the traction curves of the lubricating greases under fully flooded and starved lubrication at 80 ºC and SRR ≈ 1.3 %. Under fully flooded lubrication, the greases are positioned in a well-defined order of magnitude of the traction coefficient, where LiM1 consistently generated the highest traction coefficient, followed by the LiCaE grease, that is only slight higher than PPAO (LiM1 > LiCaE ≈ PPAO). This trend was verified for all operating conditions (speed, SRR and temperature).

The starved traction curves, also presented in Fig. 3, were always higher than the ones measured under fully flooded condition, but here no trends were observed.

Due to the non-controlled replenishment, it was observed sudden and significant changes on the traction values over time (speed), which is attributed to local replenishment. In general, the traction values under starved lubrication are very similar for all greases up to the moment that replenishment takes place and a quick reduction of the traction values occurs. In the cases replenishment didn’t occur, traction values over a pre-set traction limit (0.11) and the test was shut aborted.

Figure 2 Grease, base and bleed-oil film thickness.

Figure 3 Grease traction curves. Fully flooded (lines)

and Starved lubrication (dots).

3 5. Rolling Bearing Friction Torque

The internal friction torque occurring in rolling bearings is of major concern when energy saving and bearing performance optimization are global requirements. Therefore, full bearing tests were performed in a modified 4-ball machine with a thrust ball bearing 51107. A description of the machine and test methodology is reported in a previous work [7].

All the tests were performed with contact pressure of 2.3 GPa, rotational speed from 100 to 5500 rpm and self-induced temperature.

Figure 4 shows the evolution of the measured friction torque (M

_exp

) and operating temperature (Temp) with rotational speed. All lubricating greases showed a friction torque decreasing and a temperature increasing with rotational speed. The LiM1 grease reached the highest values of torque and temperature for all rotational speeds, and PPAO grease reached the lowest ones, while LiCaE are in between LiM1 and PPAO. The friction torque decreasing with rotational speed is due to temperature increasing, which leads to a reduction of several lubricant parameters related with friction, such as viscosity, pressure-viscosity and limiting shear-stress.

Figure 4 also shows the friction torque predictions using the latest SKF friction torque model [8].

The friction torque model does not take into account grease formulation (interaction between thickener-base oil-additives) and suggests constant full film friction coefficients (µ

_EHD

) whatever the operating temperature.

Therefore, few optimizations are suggested in this work, such as the use of the bleed-oil properties, in order to account for grease formulation and the measured traction values as µ

_EHD

, since it is well known that the friction coefficient depends on the operating conditions.

The traction coefficients from Fig. 3 were extrapolated for the bearing operating conditions considering a power law approximation for speed and temperature.

The torque loss for a thrust ball bearing 51107 is given by Eq. 3 and detailed in Appendix A:

 _   _{  }  ^ _{   } ^





 



sl

rr M

sl sl M

rr rs ish

t

G n v G

M     

^0.6

  

, (3)

where M

_rr

and M

_sl

are the rolling and sliding losses, respectively, shown in Figure 4. M

_rr

is mostly related to the bleed-oil viscosity and replenishment factor (φ

_rs

), therefore, the higher the bleed-oil viscosity, the higher is M

_rr

up to the point that the product φ

_rs

.φ

_sl

becomes dominant, reducing the rolling torque. M

_sl

is mostly related to the coefficient of friction (µ

_sl

), which in turn depends mostly on base-oil type and additive package.

Both of the friction losses are strongly connected to the film thickness, which was calculated with Eq. 2. The specific film thicknesses are shown in Figure 5 and indicate full film lubrication for all operating conditions (Λ>2), except at 100 rpm.

At full film condition µ

_sl

= µ

_EHD

, thus M

_sl

is governed by the full film friction coefficient and therefore, follows the same behavior of the traction curves.

Figure 4 Friction torque. Comparison between measured and predicted values.

Figure 5 Specific film thickness prediction.

4 6. Discussion

Combining single contact tests (film thickness and traction coefficient) with full bearing tests (friction torque) has shown to be a valuable approach to estimate the efficiency of grease lubricated rolling bearings. The use of the bleed-oil properties and the traction coefficient values were satisfactory on friction torque prediction.

It was verified that under full film lubrication, the bleed-oil is the active lubricant in rolling bearings running over a short period of time (bleeding phase) in case of full film lubrication. Therefore, at the earlier stages of grease life, the tribological behavior of lubricating greases can be estimated using the bleed-oil properties and their traction values.

Under starved lubrication and thin-films a different methodology is required. For such conditions the bleed-oil no longer is the active lubricant and the thickener-additive-surface interaction governs the separating film and thus, the traction values and rolling bearing friction torque.

To confirm this work, bleed-oil traction curves has to be performed at the same operating conditions of the rolling bearing tests and deeper analysis of bleed-oil properties is required to improve the understanding on grease lubrication mechanisms. It will lead to simple and reliable predictions of bearing operating conditions in the earlier stages of grease lubrication. Further studies on degraded greases, following the same methodology, may contribute on predicting grease tribological performance in later stages of grease lubrication.

7. References

[1]. Lugt, P.M., "Grease Lubrication in Rolling Bearings". Wiley, 2013

[2] Cann, P.M., et al., "Grease degradation in rolling element bearings". Tribology Transactions, 2001.

44 -6(3): p. 399-404.

[3] Cann, P.M., et al., "Grease degradation in R0F bearing tests". Tribology Transactions, 2007. 50 -6(2): p. 187-197.

[4] Cousseau, T., et al., "Film thickness in a ball-on-disc contact lubricated with greases, bleed oils and base oils". Tribology International, 2012.

53 -6(0): p. 53-60.

[5] van Leeuwen, H., "The determination of the pressure-viscosity coefficient of a lubricant through an accurate film thickness formula and accurate film thickness measurements".

Proceedings of the Institution of Mechanical Engineers, Part J: Journal of Engineering Tribology, 2009. 223 -6: p. 1143-1163.

[6] Katyal, P. and P. Kumar, "Central film thickness formula for shear thinning lubricants in EHL point contacts under pure rolling". Tribology International, 2012. 48 -6(0): p. 113 - 121.

[7] Cousseau, et al., "Experimental measuring procedure for the friction torque in rolling bearings". Lubrication Science, 2010. 22 -6(4): p.

133-147.

[8] Morales-Espejel, G.E. and A.W. Wemekamp, "An engineering approach on sliding friction in full-film, heavily loaded lubricated contacts".

Proceedings of the Institution of Mechanical Engineers, Part J: Journal of Engineering Tribology, 2004. 218 -6(6): p. 513-527.

APPENDIX

A – Friction Torque Model

 ^G   ^{n v}

^0.6



M

_rr

 

_ish

 

_{rs rr}

 ^(A.1)

54 , 0 83 , 1

1 m a

rr R d F

G   

 



^{1,28 0,64}



10

9

84 , 1 1

1

 







 

_

m

ish

n d

(A.2)

   

D d D K

d n Krs rs

z

e

^













2

1



 ^(A.3)

sl sl

sl G

M    (A.4)

3 / 4 05 , 0

1 m a

sl S d F

G    ^(A.5)

 _bl  _EHL

bl bl

sl    

    1   (A.6)



n



d_m bl

e

4 , 8 1

10 . 6 , 2

1

 

_

 (A.7)

B – Nomenclature

d

_m

- bearing mean diameter [mm]

F

_a

- axial load [N]

G - material parameter [-];

G

_cr

- critical stress [Pa];

G

_rr

, G

_sl

- factor that depends on the bearing type, bearing mean diameter and applied load [-];

H

_0c

- centre film thickness [µm];

K

_rs

- replenishment/starvation constant [-];

K

_z

- bearing type related geometry constant [-];

M

exp

- friction torque measured [N.mm];

M

_rr

- rolling friction torque [N.mm];

M

_sl

- sliding friction torque [N.mm];

M

_t

- total friction torque in a rolling bearing [N.mm];

n - flow index (Eq.1 and 2) [-];

n - rotational speed (Eq.3) [rpm];

R - shear-thinning reduction factor [-];

R

x

- equivalent radius [m];

R

₁

- geometry constant for rolling frictional moment [-];

S

₁

- geometry constant for sliding frictional moment [-];

U - speed parameter [-];

W - load parameter [-];

φ

bl

- Weighting factor for the sliding friction torque [-];

φ

ish

- inlet shear heating reduction factor [-];

φ

rs

- kinematic replen./starvation reduction factor [-];

µ

_EHL

- Friction coefficient in full film conditions [-];

µ

_sl

- Sliding friction coefficient [-];

ν - kinematic viscosity at the operating. temp. [mm

²

/s];