Effects of surface-topography directionality and lubrication condition on frictional behaviour during plastic deformation

Abstract

Surface-topography, friction and plasticity form a complex system at the interface of a plastic deformation operation. Surface-topography before deformation affects the frictional behaviour at the interface, which in turn affects the level and distribution of plastic deformation. The surface-topography after deformation operations is determined by a combination of the initial surface-topography, the frictional behaviour at the interface and the level and distribution of the plastic deformation. Friction testing with the asymmetric friction upsetting (AFU) test machine and subsequent surface-topography analysis help to further clarify the relationship between surface-topography, interface friction and plasticity.

The frictional conditions during testing are controlled via lubricant viscosity, film thickness, surface roughness, and deformation velocity. High and low viscosity lubricants are used in conjunction with a range of test speeds to produce frictional conditions in the boundary, mixed and hydrodynamic lubrication regimes.

The effect of initial surface-topography was examined by preparing five different initial specimen surfaces and recording the surface-topographies before testing. Five surface conditions were used: as received, etched, coarse ground perpendicular to test direction, coarse ground parallel to test direction and polished.

A correlation between surface-topography directionality and frictional resistance has previously been observed by testing specimens with grooves machined into the surface at a range of angles [A newly developed test method for characterization of frictional conditions in metal forming, in: Proceedings of the Eighth International Conference on Metal Forming, Krakow, 2000, pp. 91–97; Steel Res. 69 (1998) (4–5) 154–160; Beurteilung des Schmierungsverhaltens unterschiedlich texturierter Oberflächen mit Hilfe des Streifenziehversuches, Diplomarbeit Universität, Gesamthochschule Duisburg, 1996]. In the current work coarse grinding is used in place of machined grooves. The scale of the effect from the surface-topography directionality is compared to the scale of the effect of lubricants, arithmetic roughness value and friction regime.

Results indicate that arithmetic roughness value and lubrication regime have greater influence than directionality. These results can be explained via the application of lubrication regime theory and the importance of each component in determining the lubrication condition.

Introduction

Surface-topography, friction and plasticity form a complex system at the interface of a plastic deformation operation. This interaction is particularly critical in rolling operations. Friction and lubrication behaviour in the roll gap are widely accepted to have a significant influence on thickness reduction, rolling efficiency and finished product surface appearance [2], [4], [5], [6], [7]. An understanding of the effect whose initial surface-topography has the lubrication regime within the roll gap and the subsequent effect of the lubrication regime on the rolling operation and the final surface of the finished product is naturally desirable. In order to aid the understanding of roll gap behaviour mechanical simulations, which isolate individual aspects of the roll gap, must be employed. Subsequently, individual parameters within these individual aspects of the roll gap can be evaluated.

Previously several studies of the effects of surface-topography directionality, otherwise known as surface lay, on frictional behaviour have been made at the Max Planck Institute for Steel Research. These studies have been performed using flat drawing tests, rolling tests and plane-strain upsetting tests. The flat drawing and upsetting tests simulate specific aspects of the roll gap while the rolling test provides an overall view of roll gap behaviour. In all of these tests surface lay perpendicular to the test specimen flow-direction produced the lowest friction for lubricated surfaces.

Wolff et al. [1] studied the influence of surface lay on frictional behaviour using a flat drawing test. Lubricated specimens were tested with grooves machined into the surface at angles of 0 (parallel to drawing direction), 30, 45, 60 and 90°. Friction was significantly lower at all speeds for the specimens with a surface lay perpendicular to the drawing direction. Another interesting result from Lordan’s work was the appearance of the machined grooves subsequent to testing. The grooves at angles 30, 45 and 60° showed decreasing levels of deformation, respectively. The decreased level of deformation is a result of the inability of the lubricant to flow out of the work zone of specimens with larger groove angles. The pressurized lubricant trapped in the grooves prevents the flow of material into grooves.

Rasp and Häfele [2] studied the influence of surface lay on lubricant film thickness and workpiece reduction during cold rolling of sheets. Directional surfaces were cold rolled at a range of speeds between 0.33 and 2.83 m s⁻¹. The results showed that surfaces with a lay perpendicular to the rolling direction developed a thicker lubricant film leading to lower friction and greater thickness reduction. Velocity changes between 0.33 and 2.83 m s⁻¹ had very limited effects on the film thickness developed in the roll gap during testing of the directional surfaces. However, reduction of the workpiece during rolling increased with increasing velocity and lay angle.

Lordan [3] studied the influence of surface lay on frictional behaviour using plane-strain upsetting. It was observed that surface lay perpendicular to the direction of material flow in the test specimen provided the lowest friction. The friction decreased continuously as the surface lay angle was changed from parallel to flow-direction to perpendicular to flow-direction. Additionally, Lordan [3] observed that surfaces with lay perpendicular to flow-direction retained their lay better than surfaces with lay parallel to flow-direction.

The coefficient of friction, μ, and the friction factor, m, provide a quantification of the frictional behaviour at the interface, however other means are necessary for qualifying and quantifying the surface-topography changes of the surfaces at the interface. Three-dimensional surface-topography analysis provides means of qualitatively describing a surface via surface contour plots. However, a quantitative method of surface description for the comparison of surfaces in an objective manner is also desirable. The three-dimensional arithmetic mean surface roughness is given by Stout et al. [8] as:

$$\text{S}{}_{\mathrm{<mtext>a</mtext>}}\text{=}\text{1}\text{MN}\text{∑}\text{j=1}\text{N}\text{∑}\text{i=1}\text{M}\text{η(x}{}_{\mathrm{i}}\text{,y}{}_{\mathrm{j}}\text{)}$$

where M and N are the number of points measured in the x- and y-directions respectively, η(x, y) the surface height at coordinate (x, y), S_a the three-dimensional analogue of the standardized two-dimensional surface roughness [9]. Both R_a and S_a are subject to some variation resulting from data filtering and differing scan-area sizes.

Whereas S_a describes only the vertical properties of the surface, other information is available which may provide more useful information about the surface with regard to lubrication and friction. Quantifying the volume below a given surface height, the number of isolated volumes (non-connecting volumes) below a given surface height and the average size of isolated volumes below a given surface height is particularly useful for understanding the interaction between roll gap lubricant film thickness and surface roughness.

The height at which the volume is calculated will have a large effect on the resulting volume parameters. Fig. 1 shows an illustration of a surface profile, which shows the effect of the height at which volume is calculated. Fig. 1(a) shows the original surface profile. In Fig. 1(b) a line has been drawn through at some given height, h^∗, and the six volumes below h^∗ have been shaded in. When the height at which the volume is calculated is increased to h^∗∗ as in Fig. 1(c), the volume of interest changes significantly. Although there are still six individual volumes, the size and nature of the volumes are different. The total volume and the average volume have both increased. Additionally, some volumes that were isolated in Fig. 1(b) are now connected in Fig. 1(c). Specifically, volumes 2 and 3 in Fig. 1(b) have combined to create volume 3 in Fig. 1(c). Finally new volumes, specifically volume 1 in Fig. 1(c), are created as the height is increased.

One way to ensure that volume parameters are calculated at a consistent height is to use some quantifiable height parameter. The Abbot curve offers a solution to this problem. Fig. 2 shows an example of the Abbot curve and the compensating gradient. The compensating gradient is a straight line of any slope, which is placed on the Abbot curve at the position where it gives the best fit over 40% of the data.

The height at which the compensating gradient intersects the surface height axis at data-cut values of 0 and 100% are η₁ and η₂, respectively. The surface heights between η₁ and η₂ are the heights of the surface that bear most of the load during tribological contact. Thus in these areas the die and the workpiece are considered to be in intimate contact and any lubricant at the interface other than a monolayer will either be forced out of the interface or into the valleys below η₂. In accord with these assumptions the current work uses the volume below η₂ to describe surface void volume before and after testing.