fabrication of micro-dimpled surfaces through micro ball end milling

INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING Vol. 14, No. 9, pp. 1637-1646 SEPTEMBER 2013 / 1637

© KSPE and Springer 2013

Fabrication of Micro-Dimpled Surfaces through Micro

Ball End Milling

Eldon Graham1, Chaneel I. Park1, and Simon S. Park1,#

1 Department of Mechanical and Manufacturing Engineering, University of Calgary, Calgary, AB, Canada# Corresponding Author / E-mail: [email protected], TEL: +403-220-4175

KEYWORDS: Micro patterns, Functional surfaces, Dimples, Micro ball end milling

Industry and researchers have begun to shift their focus towards more sustainable and eco-friendly manufacturing processes in recent

years. They have recognized the vital role of functional micro surfaces for a wide range of advanced applications to address this issue.

By physically altering the surface structure of a material in micrometric scale, tribological, optical, fluidic properties and many other

surface characteristics can be altered. Several researchers have reported that micro surface patterns can reduce friction between

lubricated sliding surfaces, which in turn alleviate energy consumption and increase service life of components. Micro ball end milling

is another viable technique for creating patterned surfaces, especially for metallic parts. By tilting the spindle and tool at an inclined

angle, the spindle speed and feed rate can be adjusted so that the flutes of the cutter create periodic patterns in a workpiece surface.

Machining is an efficient and versatile manufacturing technique, making the micro dimple machining technique an ideal method to

fabricate dimpled surfaces. In this study, the fabricated surfaces are evaluated at a tribological level to illustrate their effectiveness

at reducing friction. The development of efficient methods to produce micro patterns onto large surface areas can promote a

sustainable future for a variety of novel products. The development of efficient surface pattern algorithms for generating different

dimple geometries is also a focus; and, trends in cutting forces are identified by changing different machining parameters. Depth of

cut and dimple shape, spacing and arrangement are crucial parameters, all of which factor into the performance of a functional

surface. The results of this study strongly indicate micro dimple machining as an environmentally sustainable method of producing

functional surfaces for advanced technological applications.

Manuscript received: October 11, 2012 / Accepted: February 15, 2013

1. Introduction

In recent years, there has been an increase in the understanding of

surface phenomena, which has lead to a growing interest in the

fabrication of functional surfaces. By artificially altering the

topography of surfaces at the micro level, different surface and material

properties can be influenced and manipulated. Not only are new

applications being developed, but researchers and engineers are also

finding new manufacturing techniques with which to fabricate such

surfaces. The surfaces of different products can be tailored to suit many

potential applications.

Bruzzone et al. summarized many aspects of functional surfaces,

including important surface properties that can be influenced, and

current uses for engineered surfaces.1 Many more important benefits of

functional surfaces will continue to be discovered as more research is

performed. As with any new advancement, the commercial application

of the technology can only be realized with efficient manufacturing

NOMENCLATURE

Aw = workpiece area

D = dimple density

F = feed rate

N = number of flutes

R = nose radius of ball end mill

Td = dimple period

d = cutting depth

n = spindle speed

sd = dimple spacing

t = machining time

xpitch = number of dimples per mm in feed direction

γ = spindle inclination angle

κ = circumferential angle along axis of end mill

ϕ = tool rotation angle

θ = angle between feed direction and x axis

DOI: 10.1007/s12541-013-0221-9

1638 / SEPTEMBER 2013 INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING Vol. 14, No. 9

techniques.

Most notable use of functional surfaces by researchers include

tribological applications, especially used for friction reduction between

two sliding surfaces.2,3 Wear debris generated during the friction can be

imbedded between two surfaces, forming a wedge to scratch the

surface and increasing friction. The presence of micro dimples act as

traps for wear debris, reducing the chance of wedge formation.4

Dimples also change lubrication conditions by acting as additional

lubricant reservoirs and influencing the hydrodynamic pressure

distribution.3 By reducing friction for mechanical components, dimpled

surfaces can have a direct effect on efficiency and reduced energy

consumption. Reduced friction provides the additional benefit of

decreasing wear, which can prolong the life of mechanical parts.

Several researchers have tried to characterize how shape, size

and density of dimpled surfaces affect the lubrication film thickness

between sliding surfaces and the overall tribological performance.5,6

Even the sliding velocity between the surfaces has been shown to

have an effect on the coefficient of friction.7 In general, many have

concluded that the presence of micro dimples can reduce friction.

Ryk and Etsion8 reported that under certain conditions, micro

dimples fabricated onto the face of piston rings can result in up to

a 25% friction reduction with lubricant. For automotive engines the

majority of frictional losses occur inside the engine and this result

translates directly into energy efficiency and fuel savings. With such

significant improvements, surface geometry and manufacturing

technique have become important manufacturing concerns that need

to be addressed.

Currently, micro patterns can be fabricated through a number of

different manufacturing methods. Many variants of lithography

including ultraviolet, X-ray, electron beam and soft lithography are

typically restricted to manufacturing microelectronic and other related

industries due to disadvantages, such as expensive equipment,

requirement of clean laboratory facilities, and limitations on workpiece

geometry and material. Micro patterns applied through surface coatings

have been achieved, although precise control of surface topography can

become difficult.9

The move towards laser machining is becoming a popular trend in

manufacturing, due to the speed at which lasers can cut complex micro

surface structures and the ability to be used on materials difficult to

cut.10 Despite the technology's success, lasers are not viewed as a

sustainable form of manufacturing since the power consumption and

efficiency can vary. One trend in this area has diverged towards the use

of piezo actuated fast tool servos (FTS) to deliver highly precise

oscillating cuts or indents in the shaping of a surface.11,12 Performance

and consequently the types of surfaces that can be created in FTS

highly depends on the accuracy of the device, and the quality of the

control systems used.

Micro ball end milling, with the spindle axis of rotation tilted at an

angle, has become a viable technique for fabricating dimpled surfaces.

By adjusting the spindle speed and the feed rate of the workpiece, the

periodic cutting of the workpiece by each flute can create micro-sized

dimples into the workpiece surface. The use of micro milling to

fabricate micro dimples provides many of the same benefits that

conventional machining processes have. In comparison to other

manufacturing processes, machining is much more flexible and is able

to create dimpled surfaces not only on flat plates12 but even on

workpieces with more complex shapes, like cylindrical components.2,14

While functional surfaces can be fabricated through a variety of

different means, many methods are ill suited for commercial

implementation when compared to micro ball end milling, due to lower

material removal rates.15 By tilting the spindle, less of the cutting tool

is immersed in the material, possibly allowing for even higher feed

rates than conventional cutting process. The inclined ball end milling

technique allows large areas of dimpled surfaces to be fabricated,

making it well suited for large-scale commercial manufacturing. With

modern technology, such as precision stages and high speed spindles,

becoming more widely available, accurate dimpled patterns can be

produced quickly and efficiently, saving manufacturing time and

energy costs.

Micro ball end milling offers many more inherent advantages over

other methods such as positional accuracy due to computer control of

the cutting tool and the ability to cut a variety of different materials.

Dimpled surfaces have not only been created on aluminum, but also

attempted on titanium and glass.13,16,17 Depending upon the cutting

technique or tool, several other surface topographies can also be

machined. Using a single crystalline diamond ball end mill, Yan et al.

created micro-sized dimples and micro grooves with two different

cutting methods, and even micro pyramids by cross grooving.18

Precision control of pattern geometry is important since the

performance of a functional surface in a particular application is highly

dependent upon the surface structure. Thus, the primary focus is the

development of algorithms and modeling for inclined micro ball end

milling for use in fabrication and characterization of dimpled surfaces.

Understanding of the effects of basic parameters, such as feed rate,

depth of cut and shape of the cutting edge, is vitally important in the

fabrication of the desired pattern and shape of the surfaces.

The secondary objective is the friction testing of the fabricated

surfaces to study the tribological aspects of functional surfaces.

Dimpled surfaces can be advantageous in some applications by

reducing friction between sliding surfaces. For this research, a

comparative study was done: under dry and lubricated friction

conditions, on parallel and staggered dimple patterns, and at three

different at sliding velocities.

Micro ball end milling has been shown to be a promising method

to create functional surfaces, but there are still many aspects of the

technique that need to be studied in order to further refine the

manufacturing process. With recent investigations in the area of micro

mechanical machining, it has been realized that the ability to fabricate

any smaller features is limited by factors that were often considered

negligible in the conventional macro machining process. Some of these

unknown factors, such as plowing effects, process damping related to

chatter, size effects and resulting periodic forces, were discussed briefly

in the conclusion.

The organization of this manuscript is as following: the

experimental apparatus and conditions are explained in Section 2,

followed by the description of the dimple surface generation algorithm

in Section 3. In Section 4, resulting cutting forces and surface profile

are discussed along with the results of friction tests on patterned

surfaces. The manuscript was concluded with the brief summary and

discussions on future studies.

INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING Vol. 14, No. 9 SEPTEMBER 2013 / 1639

2. Experimental setup

In the present research, a micro-computer numeric controlled

(CNC) machine tool system was used to fabricate dimpled surfaces.

The frictional characteristics of these dimpled surfaces were then

analyzed and compared.

The machining tests were performed on a micro CNC, as shown in

Fig. 1. An electric spindle (NSK Astro-E 800Z) with ceramic ball

bearings was bolted onto a bracket with mounting holes that allow the

spindle to be adjusted to any inclination angle from 0 to 90 degrees, in

15 degree increments. For the present study, the spindle was tilted at a

45 degree angle and 0.508 mm (0.02”) diameter, 2 fluted micro ball

end mills (PMT® TS-2-0200-BN) were used. Stepper motors drove the

system's three linear stages (Parker Daedal 10600), which were

controlled by a motion control system (National Instrument PXI-

1042Q). The entire milling system was secured to a vibration isolation

air table to eliminate possibility of capturing unwanted ground

vibration signals.

Initially, a 6061-T6 aluminum workpiece was prepared by ensuring

that both sides were flat and parallel. It was then mounted onto a piezo

electric table dynamometer (Kistler 9256C) to measure the cutting

force data in the x, y and z directions, which were acquired through an

anti-aliasing filter (Krohn Hite 3364) and a data acquisition system (NI

cDAQ-9172). The dynamometer was calibrated with an impact

hammer (PCB 2222) to ensure an accurate force measurement.

The inclined spindle was lowered until the rotating cutting tool just

reached the workpiece surface. An acoustic emission (AE) sensor

(Physical Acoustics Nano30) was utilized to help detect when the

cutting tool touched the workpiece surface to provide a zero point of

reference. Different lines of dimples were machined for a number of

different cases by varying the cutting depth, feed rate and spindle

speed. A surface profilometer (Surftest SJ-201P) was used to

characterize the depth profiles of the resulting patterned surface. The

profilometer measured typical surface roughness parameters, of which

the maximum profile height was of particular interest since it

represented the dimple depth.

Dimpled surfaces were machined on an area of 50 mm × 5 mm on

the surface of an aluminum workpiece in two different pattern styles,

parallel and staggered, as shown in Fig. 2. The cutting process was

carried out at a spindle speed of 500 rpm, a feed rate of 2.5 mm/s, a

cutting depth of 7 μm and an offset of 75 μm for the staggered surface

patterns. Additional parallel patterns were also machined at feed rates

of 3.3 mm/s and 4.2 mm/s.

A flat aluminum sliding surface was machined and made parallel to

the workpiece surface to perform friction testing. Steel blocks were

secured to the sliding block to give a total mass of 1.5 kg (14.7 N). The

feed and normal forces are measured using the table dynamometer. The

stages of a precision micro milling machine (Kern Micro) were used to

provide controlled movement for the friction testing.

3. Dimple surface generation algorithm

Successful generation of functional dimpled surfaces depends on the

ability to successfully predict geometric aspects. Dimple shape,

periodicity and pattern are all important parameters. Each of these

highly depends upon the inclination angle of the tool, spindle speed,

and feed rate. As shown in Fig. 3, a ball end mill was assumed to be

tilted at an inclination angle, γ, in the x direction about the y axis. The

end mill rotates with angular speed, n, and has N number of cutting

edges and a radius of R. For a given spindle speed, each tooth can

successively remove material to create adjacent dimples into the

workpiece surface and provided the linear feed rate is high enough, the

tool will travel a far enough distance to avoid overlap of the tool path.

As each tooth is immersed into the material, the maximum vertical

depth of cut, d, will be less than the cutter radius.

Dimples were created when each tooth successively cuts into the

material without overlap. The rate at which dimples can be cut into a

workpiece must be equal to the tooth passing frequency of the cutting

tool. Therefore, the time between each dimple is equal to the tooth

passing period, and this can be shown to be inversely related to the

spindle speed. The dimple spacing in the feed direction, sd, represents

the distance between two adjacent dimples, measured at the center of

each sample. It can be calculated for a given feed rate using the

following equation:

(1)

where n is the spindle speed in rev/min, N is the number of flutes, F

is the feed rate [mm/s], and the dimple period [sec] is Td= 60/nN,

which gives the time between each dimple machined. By knowing two

of the desired parameters (spindle speed, feed rate, or dimple spacing),

Sd

60F

nN---------=

Fig. 1 Micro CNC milling machine with micro ball end mill

Fig. 2 Parallel and staggered configurations (Top view)

Fig. 3 Inclined ball end milling model


the respective third parameter can be easily calculated.

Utilizing the hemispherical geometry of the ball end mill, the

approximate shape of the fabricated dimples can be predicted through

geometrical relations. Matsumura et al. derived equations that describe

the position of the cutting edge in x-y-z coordinates:17

(2)

where ft is the feed rate (mm/flute), κ is the axial angle in the vertical

plane defined from the bottom of the end mill, φ is the tool rotation

angle adjusted for the number of flutes and position along the cutting

edge, and θ describes the direction of the tool feed defined from the x

axis in the xy plane. For this study, the inclination of the spindle was

set to be in the same direction as the feed or x direction (θ = 0). For a

given rotation angle, if the axial angle was such that the z coordinate

of points along the cutting edge satisfied Z < 0, the cutting edge was

engaged into the workpiece surface, and dimples are created.

Once the positions of the cutting edges creating the dimple shapes

are predicted using Eq. 2, other geometric parameters, such as dimple

length and width, can be calculated by knowing the positions of two

adjacent dimples. For a known spindle speed, tool geometry, desired

depth, d, dimple spacing, sd, and row spacing in the y direction, the

resulting surface patterns and dimple geometry can be plotted, and the

required feed rate can be found using Eq. 1.

If the dimensions of a flat workpiece are known, the dimple pitch

(dimples per unit length) and the density of dimples can be calculated

in each direction. This provides the framework for a basic algorithm to

simulate and predict dimple surface geometry. The relationship

between machining parameters and the dimple spacing and geometry is

an important aspect in understanding how to simulate and control the

resulting surface patterns.

In general, the pitch in each direction can be found by dividing the

number of dimples in one row by the corresponding dimension of the

workpiece surface. If the dimensions and spacing of the dimpled

surface are known, the number of dimples in one row can be easily

found. The pitch in the feed direction can be shown to be inversely

related to the specified dimple spacing:

(3)

A higher rotational speed means a higher tooth passing or dimple

frequency; however, increasing the feed rate of the stage has the reverse

effect. By calculating the total number of dimples on a given area, the

approximate machining time can be estimated by multiplying by

dimple period.

It becomes important to understand the effects of these different

geometric parameters on the resulting surface pattern. For many

applications, optimization of the surface topography is of the highest

concern to achieve the best performance. The first step in

accomplishing this goal is the development of algorithms to simulate

surface geometry and their subsequent verification through research

and experiments.

4. Results and discussions

Dimple machining with an inclined ball end mill is an efficient way

to fabricate functional surfaces, especially metallic surfaces; however,

more work is required to investigate the resulting forces and surface

geometries. Having an understating of cutting forces gives major

insight into practical manufacturing considerations, and are especially

critical to maintaining the longevity of micro cutting tools. Excessive

forces can cause severe wear, chipping, and eventual breakage of the

cutting tool. These factors in turn can affect the quality of the

workpiece. Wearing of the cutting edge will reduce dimensional

accuracy, and tool breakage can possibly damage the workpiece

resulting in decreased productivity.

Cutting forces are an even greater issue when it comes to micro

cutting tools, due to their fragile nature. The speed of the inclined ball

end milling technique is due to the single pass used to cut a row of

dimples. Unlike other dimpling methods, the inclined technique

drastically reduces the number to time the machine needs to raise and

lower the tool to the workpiece surface. There may, however, be limits

to the depth of cut that single passing can achieve. In order to optimize

the process, an understanding of how different machining parameters

can affect the involved cutting forces and the resulting surface

geometry is vital.

X Rsin(κ)cos(φ)cos(γ)R d–

cos(γ)--------------- Rcos(κ)–⎝ ⎠⎛ ⎞sin(γ) ftcos(θ)+ +=

Y sin(κ)sin(φ) ftsin(θ)+=

Z R– sin(κ)cos(φ)cos(γ)R d–

cos(γ)--------------- Rcos(κ)–⎝ ⎠⎛ ⎞sin(γ)+=

xpitch1

sd----=

Fig. 4 Forces at varying depth with a constant n of 500 rpm,

F = 2.5 mm/s and ft = 0.15 mm/flute


4.1 Cutting force measurements and surface profiles

For an effective functional surface, it is important to try and achieve

the maximum possible use of the surface. As such, the parameters were

chosen so as to maximize the number of dimples on the workpiece

surface. In conventional machining, a larger cutting depth (d) yields a

higher cutting force for a given spindle speed and feed rate. Fig. 4

shows cutting forces measured during machining of dimpled patterns at

a constant feed and spindle speed with varying cutting depths. The

figure specifies the nominal cutting parameters used and shows the

forces plotted against the rotation angle of the end mill for two

revolutions.

With the specified constant spindle speed and feed rate, the distance

between the same points between adjacent dimples is constant

according to Eq. 1. The actual cutting depth was measured with the

profilometer and is shown in Fig. 5, which outlines the measured

profiles over the 1.25 mm distance and the side view (xz plane) of the

analytical profiles predicted.

The predicted dimple pitch (xpitch) in the feed direction and the

number of dimples over the 1.25 mm detected by the profilometer are

also specified. For each case, both profiles match closely with minor

deviations. We can observe that, for varying dimple depth with constant

spindle speed and feed, the size of each individual dimple shape

decreased, while the specified spacing between each dimple (sd) and

the dimple pitch (xpitch) in the feed direction both remained the same.

As outlined in Eq. 1, the feed rate and dimple spacing are directly

proportional to one another. Fig. 6 outlines the cutting forces while

varying the feed rate and holding a constant spindle speed and cutting

depth. Fig. 7 outlines the depth profiles in both the xz plane and the

actual measured dimple depth.

While the cutting depth indirectly affected the spacing between

dimples through changing the dimple size, change of the feed rate

directly affected the spacing. Variation of the desired dimple spacing or

the feed rate had equivalent effects, since a higher feed rate at a

constant rotational speed yields a larger spacing; and, a larger specified

spacing requires a higher feed rate.

The number of dimples in each scenario in Fig. 7 approximately

matches the number calculated. With wider spacing, there are fewer

dimples per unit length. A constant cutting depth was reflected in the

measured forces, since all three force sets were relatively the same in

magnitude and the dimple shapes also appeared to be unchanged.

Fig. 5 Profiles at varying depths with a constant n of 500 rpm,

F = 2.5 mm/s, ft= 0.15 mm/flute, xpitch= 6.7 dimples/mm and 8.4 dimples

Fig. 6 Forces with varying feed rate at a constant n of 500 rpm and

d = 10 μm

Fig. 7 Depth profiles at varying feed rate with a constant n of 500 rpm

and d = 10 μm


Compared to a changing feed rate, the spindle and dimple spacing

are inversely related through Eq. 1. Figs. 8 and 9 show the forces and

depth profiles for varying spindle speed, while holding a constant

cutting depth and feed rate.

For a constant feed rate, increased spindle speed yielded decreased

dimple spacing, and a smaller spacing required a higher speed. The

cutting forces shown in Fig. 8 were all relatively close; however, the

forces in Figs. 8(b) and 8(c) appear to be slightly larger. This may be

due to close dimple spacing that resulted in overlapped dimples, as

shown in Figs. 9(b) and 9(c). When dimples overlap, part of the end

mill undergoes continuous cutting creating a channel with a dimpled

bottom surface. Measurements with the profilometer must have been

made while the stylus tip was within the channel, since the measured

profiles closely matched the bottom of the analytically predicted depth

profiles. To find the actual measured depth, profilometer measurements

needed to be taken across the dimpled profile. In contrast to varying the

feed rate, a lower spacing means more dimples per unit length.

In Fig. 4(c) and Fig. 8 where the cutting depth was about 7 μm,

there is clearly less oscillation between experimental force peaks, and

significant drop in magnitude is observed in comparison to the other

force measurements where the cutting depths were 8 and 9 μm. With

decrease in cutting depth and feed rate, tool deflection and impact

forces will be decreased as well. In addition, the instantaneous area of

cut, or the volume of material removal at each instance, is nonlinear

with respect to the cutting depth due to the geometry of tool edge.

Cutting forces and chip formation in micro milling are often nonlinear

processes. The effects of material properties and ploughing may also

contribute to this difference.

From the plots of the experimental forces, there is some oscillation

in the heights of the peaks of the forces in the z direction. Two common

issues associated with micro tools that contribute to this is tool

eccentricity and tool run out. Eccentricity problems occur when the

cutting tool does not rotate perfectly about the center of the rotational

axis of the spindle. As the tool displaces, the cutting edges of each flute

may not evenly engage the workpiece surface. Adjustments of

clamping conditions of the collet and tool overhang length can help

reduce the eccentric rotational effects. Runout occurs when one flute

engages deeper into the workpiece than the other. A simple way to

avoid tool runout is to utilize a single flute cutter, since this issue arises

when there are inconsistencies in the geometry of each cutting flute,

due to uneven wearing or small manufacturing errors. These machining

issues may explain the alternating differences between adjacent peaks

in the force data and will need to be further addressed.

In a commercial setting, oscillations in cutting forces between flutes

can lead to machining inaccuracies, inconsistent surface topography

and potential uneven wearing of the cutting tool. Matsumura et al.

presented an initial discussion on correcting effects of cutter error on

dimple forces and utilized a simple feedback system in their

experiments to adjust tool clamping conditions to compensate for their

effect.17 Cutting forces gives greater insight into potential machining

issues and therefore it becomes important to understand how different

parameters can affect these forces.

Scanning electron microscopy (SEM) images were taken of the

dimpled surfaces to be used for friction testing. Surface profiles (top

view) and the dimple geometry in the xy plane were predicted and are

shown in Fig. 10.

Using the developed algorithm procedure, it can be estimated that

Fig. 8 Forces with varying spindle speed at a constant F of 2.5 mm/s

and d = 7 μm

Fig. 9 Depth profiles at varying spindle speeds with F = 2.5 mm/s feed

and d = 7 μm


each 50 mm × 5 mm surface patterned area, shown in Fig. 10, will take

11 minutes of machining time, not counting for the time needed for

setup and the toolpaths not directly relevant to the machining

processes. The patterns shown were machined at a spindle speed of 500

rpm, a feed rate of 2.5 mm/s, and a cutting depth of 7 μm. The rows

were spaced at 150 μm with an offset of 75 μm for the staggered

surface patterns. The approximate machining time [min] is related to

both the dimple spacing and the total workpiece surface area, and can

be estimated using the following formula:

(4)

where Td is the dimple period [sec] from Eq. 1, Aw is the flat surface

area of the workpiece to be machined, and D is the density of dimples

[dimples/mm2]. While pitch describes the linear spacing of fabricated

dimples, the density describes the spacing per unit area. The density of

dimples can be calculated using Eq. 5:

(5)

where Nd is the total number of dimples cut into the workpiece surface,

and can be calculated knowing dimple pitch and workpiece dimensions

through the algorithm described in Section 3.

The inclined ball end milling technique is comparatively faster than

other mechanical manufacturing methods. A ball end mill could be

used to cut dimples if it were plunge upright into the workpiece.14 In

comparison to the inclined technique, repetitive vertical and horizontal

movements of the tool to individually machine dimples would require

significantly more time. This is also the main drawback of using FTS,

where the indentation of individual dimples is performed multiple

times at a high frequency. FTS technology presents other challenges as

well, such as material deformation around dimple edges, control system

quality, and cost.12

While the inclined method shortens machining time, there is a

tradeoff between dimple shape. At higher feeds, the dimples will be

more elongated and elliptical in shape, while lower feed rates will

generate shorter, more circular dimples. In Fig. 10, it can be observed

that there is some elongation in the feed direction. Depending upon the

application, this may or may not be desirable. Using the inclined ball

end milling technique requires that a balance be made between

machining time and desired dimple shape.

Though there are small errors in the shape of dimples and their

positions found at the machined surface compared to the predicted

patterns, machined surfaces have clearly distinguished parallel and

staggered configurations, viable for comparative analysis in the

friction test.

4.2 Friction testing of patterned workpiece surface

Friction tests were performed to study the advantageous tribological

characteristics of the dimpled surfaces. A sled-type friction test setup

was used where the coefficient of friction was measured between two

flat sliding surfaces as shown in Fig 11. A sliding block made of

aluminum was used for the experiments. On top of the aluminum

sliding block, steel blocks were attached to exert a total load of 14.7 N

by means of its weight. The samples were rigidly attached to the stage,

which was controlled using a computer, whereas the sliding block was

held in place using clamps. It was ensured that the horizontal forces

exerted by these supporting clamps were negligible. The sliding surface

was placed on top of the patterned workpiece so that only the weight

contributes to the normal force. The table dynamometer was used to

detect the sliding forces at different sliding velocities, with and without

hydraulic oil (CARVER Inc. AW 32) for lubrication on both the

patterned surfaces and a blank workpiece for comparison.

The sliding coefficient of friction, μsliding, is calculated by taking the

ratio between measured sliding force and the normal force:

(6)

where Fsliding is the feed force measured by the table dynamometer, and

FN is the normal force exerted by the sliding block. The performance

of a surface can be characterized by the coefficient of sliding friction.

A comparison of a non-dimpled surface with a parallel and staggered

patterned surface is shown in Fig. 12. The samples used in this

comparative analysis were the same samples shown in Figs. 2 and 10.

t DAw

Td

60------=

DNd

Aw

------=

μsliding

Fsliding

FN

---------------=

Fig. 10 Predicted and machined surface patterns

Fig. 11 Friction testing schematic


In general, the experimental data showed that the value of the

coefficient of friction for the non-dimpled surface was higher than that

for the dimpled patterns for both the non-lubricated and lubricated

cases. Additionally, the staggered samples showed slightly smaller

coefficient of friction than the parallel ones. The difference was small

in both testing conditions, and this was the expected result as the

density of dimples in given surface area was equal in both

configurations.19

Interestingly, the coefficient of friction was clearly distinguishable

at the lubricated condition especially at the sliding velocity of 90 mm/

min. Many researchers have discussed how lubrication may only be

beneficial under certain conditions. Geometric factors, such as density

and shape, can affect the fluidic behaviour of lubrication between

sliding surfaces.6 This may, however, be limited to an extent, depending

upon how thick the film of lubricant develops.20 Some researchers have

described how dimples can act as additional lubricant reservoirs and

change the pressure distribution, resulting in increased load carrying

capacity.2,3 Again, depending on the geometry, a dimple size that is too

small may limit this behavior.5 Further studies are required to

investigate the mechanism behind how configurations of dimple

patterns affect the frictional characteristics.

Frictional behavior was also sensitive to sliding velocity. The

coefficient of sliding friction was observed to increase with the sliding

velocity for all conditions. For relatively ductile metals, such as

aluminum, this may be due to increased rubbing between the soft

aluminum surfaces.21 For lubricated conditions without the proper

speed, a suitable hydrodynamic pressure and lubricant thickness

between the surfaces may not develop, resulting in little reduction of

friction.6

In several studies the well-known Stribeck curves had been used to

illustrate frictional behavior between sliding surfaces for a given

surface geometry.2,20 For low sliding velocities, these curves typically

illustrate that the coefficient of friction decreases, then increases after

a certain velocity. The sliding velocities used would place the

experiments at the higher end of the curve if plotted on a Stribeck

curve, where coefficients of friction increase with increasing velocity.

Different patterns can exhibit different tribological behaviors; however,

more study is needed to verify these results.

An estimate of the percentage of the workpiece area to total dimpled

area, can be made by approximating each dimple as an ellipse. As

outlined in Section 3, the width, wd, and length, ld, of dimple can be

calculated by knowing the positions of two adjacent dimples through

Matsumura’s equations, shown in Eq. 2. The contact area ratio can be

calculated by:

(7)

This ratio gives the percentage of the surface area of the workpiece

that comes into contact with the sliding block used in friction testing.

Surface patterns were also machined at varying feed rates identical

to the parameters shown from Figs. 9(a) to 9(c), and parallel

configuration was maintained throughout machining. Fig. 13 shows a

comparison of the frictional performance without lubrication of three

surfaces machined at feed rates of 2.5 mm/s, 3.3 mm/s, and 4.2 mm/s.

For each of the feed rates used in the friction testing, the density and

contact area ratio, calculated using Eqs. 5 and 7 respectively, are

summarized in Table 1.

A higher feed rate resulted in a wider spacing between dimples

according to Eq. 1, which leads to lower dimple density and exposing

more workpiece surface to contact the sliding surface as shown in

Table 1. However, the coefficient of friction was not directly

proportional to the dimple density. Surfaces with the dimple density of

33 dimples/mm2 showed the highest coefficient of friction, even higher

than the surfaces with dimple density of 27 dimples/mm2.

Other literatures also suggest that the coefficient of friction has a

nonlinear relationship to the dimple density. Wakuda et al.5 presented

that textured surfaces with dimple density of 15% have lower

coefficient of friction than surfaces with dimple density of 7.5%, but

showing significantly higher coefficient of friction if dimple density

increases to 30%. They also reported that both dimple geometry and

Acontact 1

1

4---πwdldNd

Aw

-----------------------– 11

4---πwdldD–= =

Fig. 12 Friction performance of dimpled and non-dimpled surfaces

Fig. 13 Comparison of frictional performance with varying Dimple

Density, D (Parallel dimples)

Table 1 Dimple spacing for feeds used in friction testing

Feed Rate , F,

[mm/s]

Density, D,

[dimples/mm2]

Contact Area, Acontact,

[%]

2.5 44 46

3.3 33 60

4.2 27 67


sliding velocity can affect this relationship. Surfaces with a dimple

density of 7.5% showed the lowest coefficient of friction when high

sliding velocities with large dimple sizes were applied.5 Similarly, Li et

al. also reported similar trend, except that the highest coefficient of

friction was observed at the lowest dimple density under their

experimental conditions.22 The beneficial effects of dimples acting as

reservoirs for lubricants, enhancing the lubrication effects by generating

micro fluidic pressures3 as well as acting as traps for wear debris,4

might be overwhelmed by the negative effects of concentrated pressure

(i.e. friction adhesion) on the reduced contact area under same load,

leading to the increased coefficient of friction.

The experimental results in this study illustrates that dimple

machining can be applied to improve frictional characteristics of

surfaces depending on the condition of friction. The dimple geometries

and pattern configurations can be manipulated to control the magnitude

of improvement, but further investigations are required to identify the

optimal dimple geometries and configurations.

5. Conclusions

This study involves developing a dimple surface algorithm which

can be used to analytically determine necessary parameters in tilted

micro ball end milling. The efficient method to produce micro dimpled

surfaces would lead to sustainable manufacturing of functional surfaces

with improved tribological performances. A basic algorithm for

predicting surface geometry and dimple pitch was developed, based on

the relationship between spindle speed, feed rate, the distance between

two adjacent dimples, and equations developed by Matsumura et al.17

The algorithm was shown to produce acceptable results compared to

the experimental tests.

Cutting forces are a fundamental part of understanding machining

operations, different force and geometry trends were characterized by

varying different parameters. It was observed that dimple size and

cutting force magnitudes directly changed but the dimple pitch

remained the same by varying the cutting depth. A varying feed rate

and a varying spindle speed were respectively shown to directly exhibit

an opposite relationship with the dimple spacing. Overlapping dimples

resulted in the case with increasing spindle speed and an insufficient

feed rate. Varying the feed or the rotational spindle speed did not show

any significant trends in the cutting forces, suggesting that cutting

depth is a dominating factor in the magnitude of the forces.

A study was also performed to compare the friction characteristics

of dimpled surfaces. The coefficient of friction for dimpled surfaces

was smaller in comparison to the non-dimpled surface for both dry and

lubricated conditions. When parallel and staggered dimple

configurations were compared, the frictional behavior showed similar

trends with coefficient of friction increasing for higher sliding velocities.

However, under lubricated condition, the staggered configuration

showed slightly smaller coefficient of friction compared to the parallel

configuration at a certain sliding velocity, which was 90 mm/min. With

a wider dimple spacing resulting from a higher feed rate, in general, the

frictional performance was reduced.

Extended studies are required in both improving the manufacturing

technique and understanding the functionality of dimpled surfaces. Our

most immediate future work would include developing a novel force

model for tilted micro ball end milling, considering the plowing effects,

the process damping due to chatter and the size effects in order to

control the dimple geometries more precisely. In addition, injection

molding of functional surfaces would be investigated. By using micro

ball end milling technique, dimples will be machined on the mold

surface, which will be used in the injection molding process to produce

emboss-patterned surfaces. Injection molding is one of the most cost-

effective manufacturing processes for mass production. It is important

to explore the possibility in the mass-production of functional micro-

featured surfaces with minimum manufacturing cost.

ACKNOWLEDGEMENT

The authors would like to acknowledge the Natural Sciences and

Engineering Research Council (NSERC) for funding this work.

REFERENCES

1. Bruzzone, A. A. G., Costa, H. L., Lonardo, P. M., and Lucca, D. A.,

“Advances in engineered surfaces for functional performance,” CIRP

Annals - Manufacturing Technology, Vol. 57, No. 2, pp. 750-769, 2008.

2. K stner, J. and Kayapinar, H., “Efficient Machining of Micro-

dimples for Friction Reduction,” Proc. of the 7th International

Conference on MicroManufacturing (ICOMM), No. MoB1.2, 2012.

3. Etsion, I., “Improving tribological performance of mechanical

components by laser surface texturing,” Tribology Letters, Vol. 17,

No. 4, pp. 733-737, 2004.

4. Dubrujeaud, B., Vardavoulias, M., and Jeandin, M., “The role of

porosity in the dry sliding wear of a sintered ferrous alloy,” Wear,

Vol. 174, No. 1-2, pp. 155-161, 1994.

5. Wakuda, M., Yamauchi, Y., Kanzaki, S., and Yasuda, Y., “Effect of

surface texturing on friction reduction between ceramic and steel

materials under lubricated sliding contact,” Wear, Vol. 254, No. 3-4,

pp. 356-363, 2003.

6. Qiu, Y. and Khonsari, M. M., “Experimental investigation of

tribological performance of laser textured stainless steel rings,”

Tribology International, Vol. 44, No. 5, pp. 635-644, 2011.

7. Zhou, R., Cao, J., Wang, Q. J., Meng, F., Zimowski, K., and Xia, Z.

C., “Effect of EDT surface texturing on tribological behavior of

aluminium sheet,” Journal of Materials Processing Technology, Vol.

211, pp. 1643-1649, No. 10, 2011.

8. Ryk, G. and Etsion, I., “Testing piston rings with partial laser surface

texturing for friction reduction,” Wear, Vol. 261, No. 7-8, pp. 792-

796, 2006.

9. Hua, M., Tan, H. Y., Ma, H. Y., and Mok, C. K., “Patterned PVD TiN

spot coatings on M2 steel: Tribological behaviors under different

sliding speeds,” Wear, Vol. 260, No. 11-12, pp. 1153-1165, 2006.

a··


10. Sano, T., Yanai, M., Ohmura, E., Nomura, Y., Miyamoto, I., Hirose,

A., and Kobayashi, K. F., “Femtosecond laser fabrication of

microspike-arrays on tungston surface,” Applied Surface Science,

Vol. 247, No. 1-4, pp. 340-346, 2005.

11. Greco, A., Raphaelson, S., Ehmann, K., and Wang, Q. J., “Surface

Texturing of Tribological Interfaces Using the Vibromechanical

Texturing Method,” Journal of Manufacturing Science and Engineering,

Vol. 131, No. 6, 2009.

12. Yan, J., Horikoshi, A., Kuriyagawa, T., and Fukushima, Y.,

“Manufacturing structured surface by combining microindentation

and ultraprecision cutting,” CIRP Journal of Manufacturing Science

and Technology, Vol. 5, No. 1, pp. 41-47, 2012.

13. Kogusu, S., Ishimatsu, T., and Ougia, Y., “Rapid generation of

surface dimples using end milling,” International Journal of

Automation Technology, Vol. 1, No. 1, pp. 45-51, 2007.

14. Matsumura, T. and Takahashi, S., “Micro dimple milling on cylinder

surfaces,” Proc. of NAMRI/SME, Vol. 39, 2011.

15. Fleischer, J. and Kotschenreuther, J., “The manufacturing of micro

molds by conventional and energy assisted processes,” Journal of

Advanced Manufacturing, Vol. 33, No. 1-2, pp. 75-85, 2007.

16. Matsumura, T. and Ono, T., “Cutting process of glass with inclined

ball end mill,” Journal of Materials Processing Technology, Vol.

200, No. 1-3, pp. 356-363, 2008.

17. Matsumura, T., Horii, J., and Takahashi, S., “Micro dimple

machining with ball end mill,” Proc. of the 7th International

Conference on MicroManufacturing (ICOMM), No. MoB1.4, 2012.

18. Yan, J., Zhang, Z., Kuriyagawa, T., and Gonda, H., “Fabricating

micro-structured surface by using single-crystalline diamond endmill,”

International Journal of Advanced Manufacturing Technology, Vol.

51, No. 9-12, pp. 957-964, 2010.

19. Nakano, M., Korenaga, A., Korenaga, A., Mikyake, K., Murakami,

T., Ando, Y., Usami, H., and Sasaki, S., “Applying micro-texture to

cast iron surfaces to reduce the friction coefficient under lubricated

conditions” Tribology Letter, Vol. 28, No. 2, pp. 131-137, 2007.

20. Kovalchenko, A., Ajayi, O., Erdemir, A., Fenske, G., and Etsion, I.,

“The effect of laser surface texturing on transitions in lubrication

regimes during unidirectional sliding contact,” Tribology International,

Vol. 38, No. 3, pp. 219-225, 2005.

21. Nuruzzaman, D. M. and Chowdhury, M. A., “Effect of normal load

and sliding velocity on friction coefficient of aluminum sliding

against different pin materials,” American Journal of Materials

Science, Vol. 2, pp. 26-31, 2012.

22. Li, J., Xiong, D., Dai, J., Huang, Z., and Tyagi, R., “Effect of surface

laser texture on friction properties of nickel-based composite,”

Tribology International, Vol. 43, No. 5-6, pp. 1193-1199, 2010.

fabrication of micro-dimpled surfaces through micro ball end milling

Documents