The prediction of lubrication behavior during its application is a major challenge. Grease formulations have a lot of necessary properties that are essential for a well-lubricated system. For example, the fixation of the oil in the thickener matrix and its release during application, as well as the resistance against mechanical or thermo-oxidative stress, characterize the lubrication behavior of a grease. In this article, these properties are summed up as grease stability.

During OWIs research, the determination of activation energy of different grease formulations via non-isothermal TGA measurements was calculated. TGA is able to characterize the sample regarding the change in mass during heating, utilizing a defined temperature profile. It enables the identification of effects like coking, the formation of oxide layers and evaporation properties in general. Depending on the samples reactivity and the applied gas atmosphere (nitrogen, air or oxygen) different reactions are favored. Here, the reaction kinetic of thermo-oxidative degradation of various grease compositions and additives were investigated in order to characterize the grease stability.

The temperature profile for TGA was set to between 100 and 300 degrees Celsius with different heating rates. A 40 microliter aluminum crucible without a lid and a gas volume flow of 25 milliliters per minute were used. During TGA, the grease shows a mass loss over time.

The calculation of activation energy was based on the data obtained by TGA. Isothermal measurements are typically used for the determination of activation energy. In contrast, an Ozawa-Flynn-Wall analysis describes a method using dynamic temperature profiles.

Therefore, an abortion criterion during the non-isothermal reaction had to be defined. Here, the degradation up to a mass loss of 5 or 10 percent, respectively, were investigated, which was then used as an indicator for a fictive starting point of the oxidation reaction. Due to the temperature profile (initial temperature, heating rate and time of defined point of reaction), this point could be correlated to a specific temperature

This specific temperature differed with various heating rates caused by alternating residential time and was used for the analysis following Arrhenius equation: k = a exp^(EA/RT), with k as the rate constant, a as the re-exponential factor, E_A as activation energy, R as the universal gas constant (8.314 joules per mole*K) and T as temperature.

In contrast with the Arrhenius equation, the log of the heating rate was plotted instead of ln(k) of the isothermal measurement. Analogous to Arrhenius, the resulting graph displays slope m for further calculation.

Due to the dynamic temperature profile according to Ozawa-Flynn-Wall:

log() + a(EA/RT) = const. With as the heating rate.

The resulting graph enables a linear fitting with a slope m. This slope was defined as: m = a(E_A/RT)

For the calculation of E_A, the pre-exponential factor was initially taken from the standard ASTM E1641. Via iterative computation, the factor was adjusted until there was no significant change in the activation energy.

Test Results

Three base oils – polyalphaolefin, naphthenic and paraffinic – were used. The thickener system was mainly lithium 12-hydroxystearate thickener. A urea thickener was also tested. A combination of aromatic and phenolic compounds (each 0.5 percent concentration), which represented the antioxidants mixture. A zinc dithiophosphate (1 percent concentration) was used as wear protection.

Figure 1 shows the mass losses of the examined grease formulations at a temperature ramp of 100C to 300C with 2 Kelvin/min and 25 ml/min of oxygen or nitrogen. The base oils that were thickened with lithium 12-hydroxystearate showed discernable evaporation behavior. The naphthenic grease showed the fastest mass loss in the test series. This was followed by the paraffinic and PAO formulations. The change to an oxygen-free atmosphere during the test series was intended to ensure a purely thermal degradation of the grease formulation.

When in nitrogen, all tested greases showed both a smaller mass loss during test period and later onset of mass loss, i.e., the onset of mass loss at higher temperatures was observed. This indicated the formation of highly volatile autoxidation products in an oxygen atmosphere.

In Figure 2, TGA curves are broken down in more detail into the different atmospheres using PAO grease as the example. While the earlier onset of mass loss in oxygen indicates the formation of volatile reaction products, the purely thermal degradation begins later. In the following (at approximately 85 minutes), a uniform decrease in mass loss under both atmospheres was detected. It is assumed that thermal stress under this conditions and with this grease influences the mass loss much faster than the autoxidation reaction, which would lead to the formation of volatile components and thus to an increased loss of mass.

The use of polyurea-thickened formulations of PAO lubricating grease showed a similar pattern as PAO thickened with lithium 12-hydroxystearate. However, the polyurea thickener showed greater mass retention at higher temperatures.

The PAO-based grease showed lower activation energy compared to the naphthenic or paraffinic formulation, which was not expected at first. Due to the production route from mineral oil, natural antioxidants such as sulfur compounds may be present. As expected, the addition of antioxidants to the PAO-based formulation led to an increase in activation energy, i.e., to increased resistance to thermo-oxidative degradation. The addition of the ZDDP wear protection additive led to a slight increase in the activation energy, which can be explained by the thiophosphate group of the wear protection and its antioxidative function.

Both the naphthenic and paraffinic grease formulations showed unexpected behavior with regard to the activation energy. A reduction of activation energy was determined in both lubricating greases as soon as antioxidants were contained in the formulation. However, differences in the calculated activation energies are so small that no deterioration in stability can be interpreted. A cause for this anomaly has not yet been determined.

Figure 3 shows the influence of the forced thermo-oxidative stress using the PetroOxy process on the chemical composition of the grease formulation resulting in mass loss in the TGA. The forced degradation of the greases took place up to a pressure drop of 10 percent in the PetroOxy reactor. As a result, not only the temperature but also the load times due to different residence times varied.

Induction times were achieved during the PetroOxy measurement until a pressure loss of 10 percent.While ameasuringtemperature of 110C led to the samples residence time of about 2,100 minutes, the pressure drop at 140C was already reached after 280 minutes.

Figure 3 also shows the TGA of the stressed greases. It was shown that despite a short residence time, the peak temperature had a greater influence on the grease stability. A long residence time in the 110C measurement series, compared with the other temperatures, led to a slight acceleration of the mass loss and a reduced final mass in the TGA. An increase to 120C or 130C led to a shift of the initial mass loss to lower temperatures and a higher final mass loss. However, the difference between 120C and 130C was small. The short residence times at 140C led to a direct mass loss at the beginning of TGA investigations.

Due to the low thermal stress, the grease also exhibited the highest activation energy of the measurement series when subjected to a stress at 110C. This means that the grease possessed the highest activation energy of the measurement series and therefore showed the highest resistance to thermo-oxidative degradation.

The determined activation energies of the greases stressed at 120C and 130C were of similar magnitude, at 77 kilojoules per mole and 78 kJ/mol, respectively. It appears that the increased stress-temperature was compensated by the longer residence time. At 55 kJ/mol, the formulation stressed at 140C was only slightly above half of the original activation energy.

The measurements via TGA were able to differentiate the tested greases on the basis of their base oils, as well as their thermo-oxidative aging state. The calculation of activation energy refers to a consideration of an overall reaction. There is no characterization of single reactions – it is a simplified method for the differentiation of various lubricants and specific compositions.

Via the determination of activation energies, various lubricants could be differentiated. PAO-based grease showed the highest resistance to thermo-oxidative degradation, followed by paraffinic- and naphthenic-based greases. The influence of additivation with antioxidants was observed, while additivation with an antiwear agent showed no significant effect. The polyurea thickener system showed higher activation energies in pure greases compared with the lithium 12-hydroxystearate thickener. It is assumed that the aminic compounds may have an effect on the thermo-oxidative stability of the grease.

Simon J. Eiden is the deputy team leader of fuels and lubricants at Oel-Waerme-Institut, an independent German non-profit research institute that develops technologies for the efficient use of conventional and alternative fuels.