Performance of Lubricated Sliding Contact in Magnetic Field

Lubricated sliding contact experiments were carried out using ball-on-plate contact configuration. Two permanent magnets were installed in the direction same as the ball sliding over the plate specimen. The apparatus is shown schematically in Fig. 1. The installation angles of the magnets relative to the sliding direction could be adjusted to 0°, 45° and 90° as shown in Fig. 2.

Arrangements of magnets creating magnetic field perpendicular to the sliding direction was denoted 0° and was exclusively used in the experiments reported in this article.

Test plate was made of a mild steel and was 65-mm long, 20-mm wide, and had thickness of 4 mm. Two holes at both ends of the plate were used to attach it to the moving stage of the test apparatus. Surface of the plate was polished with abrasive papers to reach roughness of about 0.32 μm. The plate was in sliding contact with a silicon nitride ball of 6.35-mm diameter and had hardness of 1400 HV. Before each test, both the plate and the ball were thoroughly cleaned with acetone.

In the lubricated sliding wear test, commercially available engine oil was used. Its kinematic viscosity was 100 and 14.5 mm²/s at 40 and 100 °C, respectively.

The horizontal magnetic field was produced by two types of permanent magnets. One was made of rare earth and its flux density was 1.1 T. The other one was a ferrite magnet with flux density 0.4 T. Magnets were located near the ends of the contact area and were attached to the ball holder in the sliding wear test machine. Therefore, the total contact load P is a sum of attractive force P _m due to magnets and the normal load P _l produced by dead weight.

All tests were carried out under at the initial contact pressure of 925 MPa estimated from Hertz’s equation for a point contact (used contact configuration: ball-on-plate). Average sliding speed (because of the reciprocating motion) was 8.6 mm/s. The parameter λ, commonly used to decide on the lubrication regime [10], was equal to 0.03 therefore it can be said that lubrication within the contact region was of a boundary type.

Both plate and ball specimens were dried naturally after they were cleaned by acetone using an ultrasonic washing machine. Before the start of wear testing, weight and surface roughness were recorded and optical microscope observations of the surface were carried out. The plate was attached to the movable stage of the test apparatus and the ball was inserted into its holder. Permanent magnets were attached to a bracket with adhesive and cellophane tapes. The bracket was fastened to the ball holder with bolts. The bracket had a gap of 1 mm at the interface between the frame and permanent magnet. The arm of the testing machine was made horizontal with a spirit level. An enclosure was made on the stage in order to prevent leakage of lubricant from the contact area. The enclosure was made by 0.5-mm-thick plate and was 30 mm in length and 10 mm in width. The test plate was flooded with 0.2 ml of lubricant. Contact between the plate and the ball was made when the plate was moving in a reciprocated motion.

The total load on the contact was the sum of the magnetic force and dead load because permanent magnets attracted the plate specimen. Resulting initial contact stresses were 925 MPa. The average sliding speed was 8.6 mm/s, and amplitude was 15 mm hence the corresponding frequency was 0.29 Hz. The total number of sliding strokes was approximately 213 × 10³. In order to examine the magnetic field effect on wear process, the tests were stopped every 24 × 10³ strokes and weight measurements, the surface roughness measurements and the optical microscope observations were carried out.

Additional post-test examinations were carried out to evaluate the effect of magnetic field. They induced hardness measurements, observations of wear track on test specimens with scanning electron microscope (SEM) and X-ray diffraction (XRD) analyses. The distribution of wear debris on the plate specimen was investigated together with SEM observations and SEM analyses of wear particles. Moreover, the XRD analysis was undertaken. After cleaning test specimens, weight measurements, optical microscope observation and surface roughness measurement were carried out. Then, the plate specimen was cut and embedded in a resin and the cross-section polished. The cross-section of the wear track was used for an optical microscope and SEM observations.

Figure 3 illustrates accumulated mass loss of the plate versus the number of sliding strokes under different magnetic field densities. The accumulated mass loss is proportional to the number of sliding strokes. In addition, the accumulated mass loss at 0 T (no magnetic field present) increases sharply from 8 to 184 mg approximately, whilst the accumulated mass losses at both 0.4 T (magnetic flux density equal to 0.4 T) and 1.1 T (magnetic flux density equal to 1.1 T) increase steadily from 5 to 118 mg approximately. In detailed comparison between 0.4 and 1.1 T, the mass loss at 1.1 T is only greater by 0.4 mg. However, comparing that to mass loss at 0 T, it is clear that the presence of the magnetic fields decrease the wear amount of the plate.

Figure 4 illustrates changes in surface roughness of the wear track for different magnetic field densities. The surface roughness at 0 T changes sharply from 0.17 to 0.32 μm during 72 × 10³ strokes. In the case of 1.1 T, the surface roughness increases gradually and is fluctuating between 0.25 and 0.41 μm during 192 × 10³ strokes, whilst the surface roughness at 0.4 T fluctuates approximately between 0.25 and 0.32 μm during 72 × 10³ strokes. After 96 × 10³ strokes, all Ra values range from 0.25 to 0.42 μm, and increase slightly until 168 × 10³ strokes are attained. Unexpectedly, surface roughness at 0 and 0.4 T decreases after 216 × 10³ strokes, whilst the surface roughness at 1.1 T considerably decreased after 144 × 10³ strokes. Thus, it can be said that the range of the surface roughness changes somewhat narrows in the presence of the magnetic field.

Optical microscope images of wear tracks generated after 24 and 184 × 10³ strokes were taken. They reveal that wear surface generated under the influence of the horizontal magnetic field is a smooth with arc-like shape similar to the wear track produced in the absence of the magnetic field. However, the magnetic field additionally produced a small jagged shape after 24 × 10³ strokes.

The wear track at 0 T was 1.81 mm in width and 62.0 μm in depth after 24 × 10³ strokes and grew to 5.1 mm in width and 510 μm in depth after 184 × 10³ strokes. Whilst, in comparison, both the depth and width of the wear track at 0.4 T are almost of same sizes as the depth and width at 1.1 T0 and the same number of sliding strokes. Therefore, it appears that the influence of even relatively weak magnetic field (0.4 T) is significant and results in slight decrease in the wear of the plate.

Fine scratches are the main feature of wear surface produced at 1.1 T and 184 × 10³ strokes. The observation of the surface was carried out at 50× magnification. At magnification of 50×, signs of delamination and spalling could be seen on the wear track produced at 1.1 T.

Figure 5 shows the wear track at high magnification. Symptoms of delamination can be observed on the wear track created in the presence of magnetic field.

Secondary electron images, shown in Fig. 6, show a variety of shapes and sizes of wear debris at a magnification of 500×. After 24 × 10³ strokes, a common features for all of them regardless the density of magnetic field is that the particles are fine and less than 10 μm in size. Besides, flaky particles with irregular shapes also exist as shown in Fig. 6a, c, e. However, the wear particles for both 0.4 and 1.1 T conditions feature split flaky marks which are seen in Fig. 6c, e. In addition, there are aggregated particles, absent at 0 T, scattered and aligned with the wear track. Hence, it can be said that the magnetic field produces split flaky particles together with fine particles. Finally, wear debris after 120 × 10³ strokes include differently shaped particles in comparison with wear debris generated after 24 × 10³ strokes. They are predominantly flat particles and at 1.1 T their size is less than 40 μm.

SEM analysis was carried out for ferroparticles using white colour in the backscatter electron images (see Figs. 7, 8). The element spectrum shows that wear debris consist of C, Ca, Fe, O, S and Si. The Ca and S are lubricant additives. Oxygen peaks for both 0.4 T0 and 1.1 T conditions are high comparing to the oxygen peak at 0 T.

The XRD analysis shown in Fig. 9 show that wear debris under different magnetic flux densities consist of α-iron and CaCO₃ (calcium carbonate). Besides, other weak peaks, which are CaCO₃, are present at several points. Wear debris contain clear peak of CaCO₃ which increases with magnetic field density. Highest peak was recorded for 1.1 T0 test condition. Therefore, SEM analyses suggest that the magnetic field enhances the oxidation of ferro-wear debris. Furthermore, the peaks of CaCO₃ are increased due to the influence of the magnetic field.

Generally speaking, adsorption of surface-active substances changes the deformation behaviour of nonferrous metals. In addition, if surface electrical charge on the wear surface of Si₃N₄ ball is absent, Macmillan et al. [11] showed that the fluidity of lubricant is improved due to higher surface tension. Also, there is a reduction of the friction coefficient associated with high hardness. Therefore, it is postulated that the reduction of mass loss of the plate specimen is caused by the influence of magnetic field on the electrical charge of the wear surface of Si₃N₄ ball.

The calcium sulphonate surfactants covering the surface of the plate specimen and forming a protective film can be reinserted and reformed in regions where the protective films are damaged due to abrasive action of oxidised wear particles. However, the jagged wear surface regions are not protected from oxidation because the calcium sulphonate surfactants are not inserted into that regions [12]. Therefore, the jagged delamination regions are domains easily adsorbing oxygen.

It is postulated that generation of shallow delamination area is caused by the lack of the calcium sulphonate surfactant at the contact interface. Calcium sulphonate surfactants prevent adsorption of oxygen to the wear surface and wear particles. If the calcium sulphonate surfactants do not cover the wear particles and the surface of contact zone, it is expected that iron wear particles are attracted to the wear surface of the plate specimen. Then, the tangential load creates the shallow delamination area and scratches on the wear track of the plate specimen. In particular, the delamination region may be easily oxidised indicating increased surface reactivity of the plate due to magnetisation [6].

Whilst the influence of magnetic field reduces the mass loss of the plate under lubricated sliding, however, high magnetic flux density diminishes the effectiveness of wear reduction compared with low magnetic flux density and produces the delamination area on the wear track as a whole. These evidences suggest that the strength of magnetic flux density relates to the trapping of iron wear particles. In the wear particle formation, lubricated sliding wear under the influence at the magnetic field produces the split flaky particles because the iron wear particles at the interface between contacting specimens are compressed.

In lubricated sliding contact without magnetic field, wear debris produced are not attracted by the surface of specimens in order to be covered individually by the calcium sulphonate surfactant. In addition, the production of wear debris is continuously accumulating at the interface, and thus accelerates the abrasive action. Also, oxygen is not adsorbed to the iron wear particles and the wear surface of the plate specimen.

When magnetic field is present, it removes the iron wear particles from the interface region between specimens and increases the fluidity of lubricant. Lubricated sliding wear polishes the interface and produces fine wear particles. Besides, iron wear particles produced are moved to the outside of wear track by the influence of magnetic field (see Fig. 10a). Scratches on the plate specimen are mainly created by silicon wear particles. Iron wear particles are mainly responsible for delamination. In addition, the wear surface created by magnetisation attracts the iron wear particles and induces the transfer of the wear particles (see Fig. 10b). However, the calcium sulphonate surfactant (Ca⁺⁺ and CaCO₃) does not reinsert itself to the delamination region, hence, the region is open to oxidisation. The wear with magnetic field is decreased comparing with those without magnetic field because wear debris are removed from interface region by magnetic field. High magnetic flux density increases the transfer of particles on the wear surface of the plate specimen, therefore, the surface mainly consists of jagged delamination regions leading to wear increase.