Tribological studies have always focused on the frictional properties of ball bearings and their components due to the fact that lower friction results in lower energy consumption. However, it is necessary to protect the bearings from penetration of impurities in order to keep this low coefficient of friction (CoF) To achieve this, rubber seals are employed to prevent the entry of impurities into the ball bearings as well as to maintain lubricants inside the bearing.
Many studies have been exclusively performed to characterize the CoF of the ball bearings themselves but only a little attention has been paid to study the frictional properties of these rubber seals. The CoF of rubber seals influences the overall friction of a ball bearing, but high CoF values can cause premature damage of the seal due to frictional heat and rubber degradation. Hence, considerable effort has been taken to improve the frictional properties of rubber seals by applying surface coatings with low CoF.
Diamond-like carbon (DLC) thin films are considered to provide an optimal solution for lowering the CoF of rubber seals. DLC thin films have to show good adhesion to the rubber material used in addition to being very flexible. Using appropriate deposition techniques such as optimized plasma-assisted chemical vapor deposition (PACVD)  can help impart good flexibility to DLC films, while tribological testing has been used to characterize the friction properties. Also, tribological tests should be performed at different loads to gain insights into the mechanisms of friction/adhesion in respect to various loads and DLC structures.
This article discusses a study of frictional properties of 300 nm thin DLC coatings applied over alkyl acrylate copolymer (ACM rubber) under different loads. Thanks to the ability of the NTR2 Nanotribometer to apply a wide range of normal loads and pressures, the study results revealed the variations between friction and deformation mechanisms during the tribological contact.
Deposition of DLC Thin Films
Here, the PACVD method was used to deposit the DLC thin films on the ACM rubber. A pulsed DC power unit was employed as a substrate bias source, operating at 250 kHz with a pulse off time of 500 ns and voltages in the range of 300-600 V. In each batch, two pieces of ACM rubber with dimensions of 50×50×2 mm were coated. Before depositing the DLC thin films on the rubber substrates, they were cleaned by two subsequent wash procedures to achieve a good film adhesion. The first procedure involved five cycles of ultrasonic washing in a 10 vol. % solution of detergent in demineralized water at a temperature of 60 °C for 15 minutes. The second wash procedure involved five cycles of ultrasonic washing in boiling demineralized water for 15 minutes in each cycle.
The deposition method was a two-step process, where the ACM samples were first etched for ~30 minutes in argon plasma to further clean the contaminations on the surface and then in a plasma mixture of argon and hydrogen for ~10 minutes to further improve the adhesion of the subsequently deposited DLC thin films. The second treatment involved replacement of hydrogen with acetylene, followed by deposition of the DLC films.
This resulted in the preparation of two types of samples:
- Uncoated ACM Rubber with a thickness of 2 mm
- ACM rubber with homogeneous DLC coating with a thickness of 300 nm
Figure 1 shows the surface and the cross-section of the DLC coating.
Figure 1. The morphology of the DLC coated ACM rubber on top surface (a) and on cross-section (b). The scale bars represent 50 µm and 5 µm respectively.
Anton Paar Pin-on-disk Nanotribometer (NTR2), shown in Figure 2, was used to perform the nanotribological experiments at four different loads: 1 mN, 10 mN, 100 mN and 1000 mN. The NTR2 is a unique instrument that can perform these tests at a wide range of loads and contact pressures. Using active force feedback, the NTR2 ensures precise control of normal load under different conditions. Its concept with easily exchangeable double cantilevers enables the maintenance of outstanding force and displacement resolution in the range of 0.005-1000 mN.
In all tribological experiments, the counterbody used was a 2 mm diameter stainless steel ball. The stainless steel ball can easily be placed on a specially designed support shaft, which also enables attachment of other customer made counterbodies. The tests at each load were carried out with wear track radii of 4 mm, 6 mm, 8 mm and 10 mm so that sufficient distance is ensured between individual wear tracks. The linear speed was set to 5cm/s and the total number of laps was set to 10000 for all tests. Then, the CoF was recorded.
Figure 2. The Anton Paar Nanotribometer. System of easily interchangeable cantilever allows to span loads from 0.005 mN up to 1000 mN.
Results and Discussion
For all loads, the average value of CoF for the uncoated ACM rubber varied between 0.7 and 1.3 due to the existence of strong adhesive interactions between the counterpart and the uncoated rubber (Figure 3). During each test, the CoF value was also varying, indicating the severe damage caused to the ACM rubber (Figure 4). On the other hand, the DLC coated samples showed a lower CoF compared to the uncoated ACM rubber, varying between 0.05 and 0.5 based on the applied load. The lowest CoF value was obtained on the DLC film with a load of 10 mN (average CoF value of 0.05). Unlike the uncoated ACM rubber, the DLC coated samples showed a very stable CoF during the whole duration of the tests (10,000 laps). Such a behavior shows not only good stability but also outstanding durability of the DLC coating on the rubber sample during repeated spherical contact even at high loads (Figure 4). However, a slight increase in the CoF value during the tests can be noticed, especially at higher loads (100 mN and 1000 mN). This is because the contact area gradually increases during the tests due to non-recovery of the underlying ACM rubber substrate , which, in turn, increases the CoF.
Figure 3. Comparison of coefficient of friction at various loads for the two tested sample: uncoated rubber (a) and DLC coated rubber (b).
The evolution of the CoF with the applied load (Figure 3) can be described as a combination of two effects: adhesion of the steel ball to the DLC coating and viscoelastic hysteresis of the underlying ACM rubber substrate caused by repeated mechanical loading.
Figure 4. Comparison of the wear track after the 1000 mN load pin on disk test on ACM rubber (a) and DLC coated rubber (b). Note deep track on the ACM rubber indicating severe damage.
The effect of adhesion between the DLC surface and the steel ball is important at low loads (1 mN), resulting in high CoF values. The contribution of adhesion reduces at the 10 mN load and the viscoelasctic hysteresis is not as pronounced as at high loads (100 mN and 1000 mN), resulting in a very low CoF value (~0.05). The evolution of CoF corresponding to the application of the normal load during the nanotribological experiments is shown in Figure 5. For the DLC coated rubber, the lowest average CoF values were recorded at the 10 mN load.
Figure 5. Evolution of coefficient of friction for the ACM uncoated rubber and DLC coated rubber as a function of normal load.
The ability of the NTR2 Nanotribometer to apply a wide range of loads enabled the study of the contact conditions on the transition between the adhesive and hysteresis contributions to the CoF values on the coated and uncoated ACM rubber samples. Furthermore, the application of the DLC coating results in a significant decrease of CoF in relative to uncoated ACM rubber, which is vital for application of rubber seals in ball bearings.
The study results clearly demonstrate that application of the DLC coating decreases CoF considerably compared to uncoated ACM rubber. Also, the CoF remains very stable all through the whole tribological test, demonstrating the high durability of the DLC coating. On the other hand, the severe damage is observed for uncoated ACM rubber at 100 mN and 1000 mN, further demonstrating the protective function of the DLC coating. The wide range of applied loads demonstrated that the combination of adhesion (predominant at low loads) and viscoelastic hysteresis of rubber (predominant at high loads) governs the frictional behavior of this kind of elastic-plastic materials.
The ability to provide a large range of applied forces and superior force range and resolution make the Anton Paar NTR2 Nanotribometer an ideal instrument for such focused research.
 Y.T. Pei, X.L. Bui, J.P. Van der Pal, D. Martinez-Martinez, X.B. Zhou, and J.Th.M. De Hosson, ‘Flexible diamond-like carbon films on rubber: on the origin of self-acting segmentation and film flexibility’, ActaMaterialia 60 5526 (2012).
 D. Martinez-Martinez, J.P. Van der Pal, Y.T. Pei, J.Th.M. De Hosson, ‘Performance of diamond-like carbon-protected rubber under cyclic friction. I. Influence of substrate viscoelasticity on the depth evolution’, Journal of Applied Physics 110, 124906 (2011).
 D. Martinez-Martinez, J.P. Van der Pal, M. Schenkel, K.P. Shaha, Y.T. Pei, and J.Th.M. De Hosson, ‘On the nature of the coefficient of friction of diamond-like carbon films deposited on rubber’, Journal of Applied Physics 111 114902 (2012).
This information has been sourced, reviewed and adapted from materials provided by Anton Paar TriTec SA.
For more information on this source, please visit Anton Paar TriTec SA.