Tribological Performance of Textured Surfaces Created by Modulation-Assisted Machining

Scalable surface texturing at the micro- and mesolength scales continues to receive significant attention from the tribology community, as many studies have pointed to various types of controlled textures that can be introduced to improve friction and lubrication ability of mechanical components [1–7]. For example, microsized depressions (e.g., grooves, dimples) intentionally created on a bearing surface are known to act as fluid reservoirs, promoting thin film lubrication between components. Surface dimples have been shown to boost hydrodynamic pressure, causing enhanced separation of surfaces [8]. Further, these depressions also simultaneously function as receptacles for wear particles, thereby reducing scratching of the surfaces that can be caused by the relative motion occurring at sliding interfaces.

While many fabrication methods can be used to create textured surfaces (e.g., sand-blasting, electron-beam texturing, photolithography methods, ion beam texturing, and laser ablation), most are limited in practical application due to either lack of control or scalability [9]. In comparison, mechanical surface texturing methods based on controlled microscale material removal provide a cost-effective approach for creating textured surfaces with control over the geometry of the textured surface. The present study is focused on exploiting this capability to control surface (CS) texture morphology to explore the effects of varying dimple morphology on surface wear performance.

Microscale material removal can be facilitated using positioning systems for modulation-assisted machining (MAM) [10–13], wherein piezoelectric-driven tooling platforms are used to rapidly control tool motion. In MAM processing, wherein a controlled low-frequency modulation (<1000 Hz) is superimposed on the linear motion of the cutting tool, typically to enable tool-work separation. If the controlled modulation is applied in the direction of tool feed, it can enable both the chip formation and resulting surface generation to occur in a discrete manner [10–13]. This has been exploited in prior work [10] to produce both particulate and surfaces with controlled microscale morphology. In this regard, the relative motion of the cutting tool over the work surface and the resulting surface texture are determined by controllable machining and modulation parameters.

There are two types of texturing configuration [6,7] based on the turning orientation. “Sliding type” configuration involves texturing of an inner or outer diameter of a workpiece by repeated sliding of a hard tool along the axis of the work. The second configuration, “plunging type,” involves repeated plunging of tool into the face of the workpiece in the axial direction to create the textured surface. In this work, microdimple surface patterns were created using the plunging-type texturing configuration (Fig. 1).

The effect of surface texturing on the tribological properties of materials has been previously studied both experimentally [14–16] and theoretically [17,18], showing that textured surfaces have a great potential to reduce friction and improve load-carrying capacity compared to untextured surfaces. Dimple shape and density [5,6] can be important factors affecting the lubrication performance of textured surfaces. Other researchers [19,20] have shown that dimples with elliptical/circular shape have the best load-carrying and friction reduction capacity. Further, mechanical texturing-based processes are best amenable for these types of surface geometries. In this paper, the tribological behavior of several textured surfaces with elliptical/circular dimple geometry created by MAM is investigated under lubricated ball-on-flat reciprocating configuration and compared with that of flat specimens.

Cylinders of brass 360 were machined using a four-axis computer numerical control lathe (HAAS ST-20Y, Oxnard, CA). To prepare the initial work surface, a polycrystalline diamond coated tool (Hertel HT-CNMA 432 HT420CD, Latrobe, PA) with a 0.8 mm corner radius was used. The faces of the cylinder were machined using a feed rate of 0.05 mm/rev and 1000 rpm; this produced a mirror-like polish surface finish with an Ra ∼ 0.081 μm. A custom-built, piezo-electric tool holder was designed to enable plunging-type texturing and is displayed in Fig. 1.

Surface textures were produced on the faces of the brass cylinders according to the machining and modulation conditions in Table 1 using a custom carbide-engraving tool. The resulting surfaces were tested against 6 mm diameter balls of AISI 440C stainless steel under lubricated conditions using a ball-on-flat (nonconformal contact) reciprocating tribometer. The normal load of 4.9 N (mean Hertzian contact pressure = 1.16 GPa, maximum Hertzian contact pressure = 1.75 GPa), frequency of 3 Hz, and amplitude of 10.5 mm were kept constant. Variable sliding distance tests were carried out under sliding distances of 19 m, 38 m, and 76 m. Brass surfaces were cover with 2 mL of a synthetic base oil (Synton PAO 40, Middlebury, CT) at the beginning of the tests. The physical properties of the lubricant oil used are shown in Table 2.

Optical micrographs were obtained using a Zeiss optical stereoscope (Oberkochen, Germany). Scanning electron microscopy (SEM) images and energy dispersive spectroscopy (EDS) measurements were obtained using an AMRAY 1830 (Bedford, MA) scanning electron microscope.

Although other parameters such as sliding speed, dimple density, normal load, or even lubricant viscosity would influence the tribology behavior of textured samples [6], in this paper, only the influence of sliding distance and dimple size (Effect of Dimple Length and Depth section) is discussed. The effect of sliding distance in the wear test was studied on three different surfaces, an untextured CS and two textured surfaces with differing dimple geometries (#1A and #1B). Textured samples #1A and #1B were created under similar modulation conditions but differing machining conditions (Table 1). In this regard, sample #1A was created under constant surface speed and sample #1B was created under constant spindle speed. From images of the samples in Fig. 2, it is clear that maintaining either constant surface or constant spindle speed in the plunging-type texturing process will result in different dimple geometry on the textured surfaces. Specifically, it is shown that the dimples are significantly longer in the case of the constant spindle speed, while they are shorter and more uniform in the case of constant surface speed.

Figure 3 compares wear volumes of the three samples obtained under increasing sliding distance using both wear measurement methods employed. As expected [23,24], wear volumes determined by image analysis according to Eq. (1) are always higher than those determined from profilometer-based measurements. This can be explained from the observation of the wear track profile in Fig. 4, where the area of material deformed plastically (A2 + A3) is sometimes greater than the wear loss area A1. With both methods, however, that same general trend was observed. Under the shortest and middle sliding distances, no important differences were found between textured and untextured surfaces. This was presumably due to the mildness of the contact conditions. When the sliding distance is short, the amount of material worn is too small in all cases, and the effect of the dimples on the wear behavior of the surface is not clear. As wear volume increases (76 m), the dimples act as fluid reserves and help to form a lubricating thin film between surfaces [5,8], achieving a wear reduction of around 75% for both textured surfaces with respect flat specimen.

As can be seen in Fig. 5, while wear scars of the textured surfaces remain almost constant when the sliding distance is increased from 38 m to 76 m, a much wider wear width is observed for the untextured surface after a sliding distance of 76 m. Under the experimental conditions studied, AISI 440C steel balls did present wear, but presented adhesion of brass wear particles (Fig. 6) that increases in size with sliding distance for the untextured surface and remains invariable for both textured surfaces.

Previous studies [5,6] have shown that dimple shape and their distribution on the work surface can be important factors affecting the lubrication performance of textured surfaces. Effect of dimple density has also been studied in past work [6]. In order to establish the effect of dimple size on the wear performance of the textured samples, four differently textured samples with variable dimple length and depth and constant dimple density were created by modifying changing the modulation conditions per Table 1 (#2A, #2B, #2C, #2D). Table 3 summarizes the dimple lengths and depths of the samples used in this section. The selection of the dimple depths (0.045 and 0.015 mm) and lengths (1.4 and 2.4 mm) was based on previous results [3,19] and some limitations of the texturing process. Optical and white light interferometric images of the surface dimples created on the brass samples are shown in Fig. 7.

The wear performance of the differently textured surfaces is summarized in Fig. 8. From the figure, increasing the dimple length from 1.4 mm to 2.4 mm for the samples with deeper dimples (#2A and #2B) increased the wear volume under the experimental conditions studied. This is also seen in the shallower dimples (#2C and #2D), though to a much lesser degree than in the longer dimples. Also evident in the figure is the effect of dimple depth on wear performance. For the samples with longer dimples (#2B and #2D), reducing the dimple depth reduced the wear volume of the sample. This is also seen to a lesser degree in the samples with shorter dimples (#2A and #2C). From the entire set of conditions, the highest wear resistance was observed in the case of sample #2C, which consisted of relatively short and shallow dimples. Figure 9 shows the optical images of wear tracks on surfaces with dimple depth of 0.045 mm (#2A and #2B) and dimple depth of 0.015 mm (#2C). It is interesting to note that the wear scars were wider at the entrance and exit of the pin in the vicinity of the deeper dimples (Figs. 9(a) and 9(b)), while the wear scar remained constant for the sample with shallower dimples (Fig. 9(c)).

In this paper, wear mechanisms are discussed in the context of optical and SEM micrographs, as well as EDS-based analysis of wear tracks, which is an accepted method to discuss wear mechanisms of materials in the area of tribology. It is important to note that the optical and SEM micrographs selected for the paper are representative of the worn surface.

In all cases, but particularly at low and medium sliding distances, brass surfaces showed a strong component of abrasive wear (Figs. 10(a), 10(c), and 11). Under the longer sliding distance, a component of adhesive wear due to plastic deformation was observed, with accumulation of material at the edges of the wear track (Figs. 10(b) and 10(d)). Finally, an adhesive wear component was also observed for all samples with transfer of material from the brass disk to the steel pin (Fig. 12). This transfer layer was larger on untextured samples. Further, the EDS elemental map (Fig. 13) of the worn surfaces showed no transfer material from the steel pin to the brass surfaces.

The reciprocating wear resistance of brass dimpled surfaces created by MAM was investigated under lubricated conditions and compared with the wear performance of a flat untextured specimen. The effect of dimple length and depth on wear performance was studied. For textured surfaces, a wear reduction of about 75% with respect to the untextured surface was observed for a relatively longer sliding distance. Increasing dimple length caused wear volume to increase, particularly for samples with relatively deeper dimples. Similarly, increasing dimple depth also caused an increase in the wear volume. This was further evident in the observation that highest wear resistance was resultant from samples with short length and shallow dimples. From the SEM micrographs, all surfaces showed a strong component of abrasive wear. Under longer sliding distances, plastic deformation and material adhesion to the steel ball were observed.

From these results, it is clear that MAM has the potential to be used as a method to produce surfaces with controlled dimple geometry in a wide range of metals and alloys that can be processed by conventional machining routes. Further, these results show that dimple geometry can be modified using the control inherent in MAM so to realize controllable improvement in surface wear performance. Work is currently ongoing to better establish the geometric limits of MAM as a mechanical surface texturing route, as well as other potential limits related to material system compatibility.

The authors would like to thank the financial support from the Mechanical Engineering Department at the Rochester Institute of Technology.

• Division of Civil, Mechanical and Manufacturing Innovation (NSF CMMI 1130852 and NSF CMMI 1254818).

A =

modulation amplitude

f_m =

modulation frequency

h₀ =

feed rate

rt =

tool tip radius