ARTICLE IN PRESS

0890-6955/$ - se

doi:10.1016/j.ijm

�CorrespondE-mail addr

International Journal of Machine Tools & Manufacture 47 (2007) 1636–1645

www.elsevier.com/locate/ijmactool

Tool wear in friction drilling

Scott F. Millera, Peter J. Blaub, Albert J. Shiha,�

aMechanical Engineering, University of Michigan, Ann Arbor, MI 48109-2125, USAbMetals and Ceramics Division, Oak Ridge National Laboratory, P.O. Box 2008 – M/S 6063, Oak Ridge, TN 37831-6063, USA

Received 23 May 2006; accepted 31 October 2006

Available online 12 December 2006

Abstract

This study investigates the tool wear in friction drilling, a nontraditional hole-making process. In friction drilling, a rotating conical

tool uses the heat generated by friction to soften and penetrate a thin workpiece and create a bushing without generating chips. The wear

of a conical tungsten carbide tool used for friction drilling a low carbon steel workpiece is studied. Tool wear characteristics are

quantified by measuring its weight change, detecting changes in its shape with a coordinate measuring machine, and making observations

of wear damage using scanning electron microscopy. Energy dispersive spectrometry is applied to analyze the change in chemical

composition of the tool surface due to drilling. In addition, the thrust force and torque during drilling and the hole size are measured

periodically to monitor the effects of tool wear. Results indicate that the carbide tool is durable, showing minimal tool wear after drilling

11,000 holes, but observations also indicate the progressively severe abrasive grooving on the tool tip.

r 2006 Elsevier Ltd. All rights reserved.

Keywords: Tool wear; Friction drilling; Wear measurement; Chipless hole making

1. Introduction

Friction drilling is a novel machining process that utilizesfrictional heat between a rotating conical tool and the workmaterial to soften and penetrate thin-walled workpiecesand to form a hole [1–5]. The frictional heat increases theductility of the workpiece material, which is extruded ontoboth the front and back sides of the material being drilled.The process forms a bushing in situ and is clean andchipless. The added height of the formed bushing canlengthen the threaded portion of the hole, and conse-quently increase the fastener clamp load for joining thinsheet metal. The technology and principles of the frictiondrilling process have been reviewed by Miller et al. [1].

Fig. 1 shows the stages in friction drilling a ductile metal,square tube workpiece. First, the tool comes into contactwith the workpiece. Then, at the main thrust stage, the toolpenetrates the workpiece and high axial force is encoun-tered. The friction force on the contact surface producesheat and softens the work material. Then, in the material

e front matter r 2006 Elsevier Ltd. All rights reserved.

achtools.2006.10.009

ing author. Tel.: +1734 647 1766; fax: +1 734 936 0363.

ess: shiha@umich.edu (A.J. Shih).

separation stage, the tool penetrates the workpiece andmakes a hole. The ductile work material encompasses thetool. Finally, the tool retracts and leaves a hole with abushing. Cross-sections of several holes friction drilled witha 5.3mm diameter carbide tool in AISI 1015 carbon steeltube are shown in Fig. 2.Tool wear in friction drilling is a concern because it affects

the characteristics and tolerances that are achievable. It ispromoted by the high temperature and forces generated inthe process. Miller et al. [1] have observed high peaks in theexperimentally measured thrust force, torque, and tempera-ture. Tool wear in friction stir welding, a process similar tofriction drilling, was characterized by Fernandez and Murr[6] and Liu et al. [7] for threaded steel and WC–Co tools,respectively. A preliminary study of performance of toolwear for coated and uncoated friction drilling tools has beenreported [8]; however, there is a lack of additional publishedresearch on the wear of friction drilling tools.The goals of the current research were to quantify the

wear and surface degradation of tungsten carbide toolsused in friction drilling of steel using changes in tool shapeand mass, to characterize worn tool surface features, and toanalyze the surface chemistry of the worn tool tip. Thrust

ARTICLE IN PRESS

t

Fig. 1. Stages of friction drilling in a square steel tube.

Fig. 2. Cross-section of friction drill holes in AISI 1015 steel tube.

Fig. 3. Setup for friction drilling of carbon steel square tube.

S.F. Miller et al. / International Journal of Machine Tools & Manufacture 47 (2007) 1636–1645 1637

force, torque, and hole inside diameters were also measuredas the tool wear progressed.

Fig. 4. Close up view of square tube workpiece after friction drilling tool

wear test.

2. Experimental setup and wear measurements

2.1. Machine tool and workpiece

A Milacron Sabre 500 vertical machining center wasused for the friction drilling tool wear study. Fig. 3 showsthe drilling setup. The workpiece was fixed with blocksacross two vices that provided support in the axialdirection to minimize workpiece bowing. Holes were drilledfrom left to right relative to the view in Fig. 3. After eachhole was drilled, the tool was indexed 7.6mm to the rightautomatically. Depending on the length of the square tubeworkpiece, 70–80 holes could be drilled on each side of thesquare tube. The spindle speed was 2800 rpm and tool feedrate was 254mm/min. These are common process para-meters for friction drilling of low-carbon steel [1,2].

Each hole took about 2.3 s to drill. The time betweeneach hole was set to at least 10 s to allow heat dissipationfrom the tool because of the high temperature in theprocess. The maximum temperature generated in frictiondrilling was measured to be about 1/2 to 2/3 of the meltingtemperature of the carbon steel workpiece [1] by an

infrared camera system. Even higher temperatures areexpected at the tool–workpiece interface.The work-material was AISI 1015 carbon steel, which is

a low-carbon steel similar to the AISI 1020 steel used infriction drilling experiments in [1]. The workpiece wassquare tubing with wall thickness of 1.5mm, width of19mm, and average length of 0.6m. As shown in Fig. 4,square tubing was ideal for drilling many holes because ithas four usable, well-supported sides.A Kistler model 9272A piezoelectric drilling dynam-

ometer was used to measure the axial thrust force andtorque during the drilling process.

ARTICLE IN PRESSS.F. Miller et al. / International Journal of Machine Tools & Manufacture 47 (2007) 1636–16451638

2.2. Tool geometry and tool wear characterization

A picture and an illustration of the key dimensions of thefriction drilling tool are shown in Figs. 5(a) and (b),respectively. The tool used in this study has d ¼ 5.3mm,a ¼ 901, b ¼ 361, hc ¼ 0.940mm, hn ¼ 5.518mm, andhl ¼ 7.043mm. The transition from the cylindrical to theconical region is rounded and not as well defined asillustrated in Fig. 5(b). Furthermore, the transitionbecomes even less distinct as tool wear progresses.

The cylindrical and conical regions have the trilobularcross-sectional shape. The lobes are defined as regionsof the tool that have a larger radius or raised profile.The tool used has three lobes 120o apart from each otherthat provide primary contact with the workpiece. Profileson the tool are measured using the coordinate measuring

�

r

z

CL

notch

da

tum

An

gu

lar

da

tum

Scan in the valley

of trilobular tool

(Fig. 9)

a b

c

Fig. 5. Friction drill tool: (a) picture of the tool, (b) illustration of regions in

CMM measurement.

machine (CMM). Fig. 6(a) shows the CMM measured toolprofiles at two angular positions (601 apart from eachother), one on the ridge and another on the valley of thelobe, showing the difference in radius of the trilobulargeometry. The difference in radius between two scans isabout 0.087mm. Fig. 6(b) shows three cross-sectionalCMM measurements of the trilobular geometry at 8.98,13.0, and 16.0mm axial locations. The dashed circlesrepresent the original circular cross-section of the toolbefore the grinding to make the trilobular shape. Measure-ment results show a cam-like lobular shape of the toolcross-section with about 0.082mm maximum distance tothe dashed circle. This distance matches the measurementin Fig. 6(a).A notch, as shown in Fig. 5(a) was ground along the

axial direction in the shoulder region of the tool. This

d

�

Shank region

Shoulder region

hl, Cylindrical region

hc , Center region

hn, Conical region

CLR

ad

ial

Axial datum

Scan on the apex

of trilobular tool

(Fig. 10)

the tool and key dimensional parameters and (c) three reference data for

ARTICLE IN PRESS

0.0

0.5

1.0

1.5

2.0

2.5

3.0

10 11 12 13 14 15 16 17

Axial distance (mm)

Ra

dia

l d

ista

nc

e (

mm

)

-3

-2

-1

0

1

2

3

-3 -2 -1 0 1

Trilobular tool

Circular geometry

Angular datum (notch in tool)

Axial position: 8.98 mm

13.0 mm

16.0 mm

2 3

a

b

Fig. 6. CMM profiles of tool (a) at different angular positions showing difference in radius of the peak and valley for trilobular geometry and (b) cross-

sectional scans of the tool showing deviation from circular cross-section.

S.F. Miller et al. / International Journal of Machine Tools & Manufacture 47 (2007) 1636–1645 1639

notch aligns with the apex of one of the lobes and marksthe angular datum for CMM measurement of tool wear, asdescribed later in this section. The trilobular geometrydecreases the overall contact area and torque in frictiondrilling. The tool material, provided by FormdrillTM, was acobalt-bonded combination of tungsten carbide andtitanium carbide and was designed for friction drilling.

Throughout the study, measurements of drilling thrustforce, torque, tool weight, and tool profile were conductedafter drilling hole numbers 1, 2, 3, 4, 5, 10, 50, 100, 200,500, 1000, and every thousand holes after that. In total,11,000 holes were drilled and the tool was still performingwell at the end of the experiment. After drilling the 9000thhole, scanning electron microscopy and light elementenergy dispersive X-ray analysis were performed on thetool.

One of the key wear measurements is the change of thetool shape, which was measured using a CMM (LegendTM

by Electronic Measuring Devices). A specially designedtool fixture was built to hold the tool in the shoulder region(Fig. 7). The tool was positioned in the fixture with thefront of the tool shoulder flush with the front face of the

fixture, and the slot was in line with a gap machined intothe fixture. The CMM incorporates a scanning measure-ment head with two probes, marked as Probes 1 and 2 inFig. 7(a). Probes 1 and 2 are made of tungsten carbide with1.0 and 0.7mm diameter ball tips, respectively. The firststep in the measurement was to establish the datum usingProbe 1. Probe 1 was programmed to measure the back endof the tool and establish three datum surfaces, marked asthe axial, radial, and angular datum in Fig. 5(c). As shownin Fig. 7(b), Probe 1 first came into contact with the backof the tool shoulder region to establish the axial datum.Probe 1 then traced the shank region of the tool to find theradial datum. Finally, Probe 1 touched around theshoulder region to find the position of the notch forangular datum. These three datum planes do not changeduring the tool wear test and were the references forprecision measurement of small change in tool geometry orthe tool wear.After the datum surfaces were established and the

relative location between Probes 1 and 2 remainedunchanged, Probe 2 was programmed to perform 12 scans,301 apart from each other in the angular direction, of the

ARTICLE IN PRESS

Fig. 7. CMM for tool wear measurement: (a) tool, tool holder, and CMM scanning head with Probes 1 and 2, (b) Probe 1 measuring the axial datum, and

(c) Probe 2 performing axial scanning of the tool conical region.

S.F. Miller et al. / International Journal of Machine Tools & Manufacture 47 (2007) 1636–16451640

conical and center regions of the tool. The path of thesescans was along the axial direction. Fig. 7(c) shows theProbe 2 of CMM scanning the tool for profile measure-ment. The tool profile was scanned eight consecutive timeson the worn tool to determine repeatability of the scans.After each scan the tool was repositioned in the fixture. Thestandard deviation was calculated at four points on thetool (Fig. 10) and for tool radius at each axial position ofthe eight scans.

The CMM was also used to measure hole sizes. Thediameters of the 1st, 5000th, and 11,000th hole weremeasured at three depths (0.46, 1.99, and 4.46mm from thetop of the hole).

3. Results and observations of tool wear

3.1. Observations of tool wear by optical microscopy

The trilobular geometry of the tool influenced the toolwear and material adhesion from the workpiece to the tool.

Fig. 8 shows optical micrographs of the conical region ofthe tool after 2, 1000, 5000, and 11,000 drilled holes. Thelobe apex is in the center of the tool in each micrograph. Inthe very early stage of tool wear, as shown in Fig. 8(a), theonly observation that can be made is the adhesion of workmaterial to the tool. A patch of wear including circulargrooves in the conical region occurs on the lobe apex afterdrilling 1000 holes, as shown in Fig. 8(b). At 5000 holes, asshown in Fig. 8(c), this patch area has grown significantly.Fig. 8(d) shows the patch of wear extending around thetool conical region. Circular grooves around the toolperipheral develop in the area intersecting tool conical andcylindrical regions. These grooves are likely due to theabrasive wear.

3.2. CMM measurements of tool profile and tool wear

Fig. 9 shows the tool profile of the new tool and the toolused after drilling the first hole. The profile was measuredat 301 in the angular direction from the angular datum, as

ARTICLE IN PRESS

Fig. 8. Optical micrographs of friction drilling tool after: (a) 2, (b) 1000, (c) 5000, and (d) 11,000 holes.

0.9

1.4

1.9

2.4

11 12 13 14 15 16

Axial distance (mm)

Rad

ial d

ista

nce (

mm

)

1

New

Fig. 9. CMM profiles of unused tool and after drilling one hole showing

buildup of work material on the tool.

0.0

0.5

1.0

1.5

2.0

2.5

3.0

10 11 12 13 14 15 16 17

Axial distance (mm)

Rad

ial d

ista

nce (

mm

) New

11000

5000

110005000

New

A

D

C

B

Fig. 10. Comparison of CMM measured profiles on the lobe apex of the

new tool and tool after drilling 5000 and 11,000 holes.

S.F. Miller et al. / International Journal of Machine Tools & Manufacture 47 (2007) 1636–1645 1641

shown in Fig. 5(c). The evidence of material adhesion onthe tool is apparent. The horizontal axis represents axialdistance from the axial datum and the vertical axisrepresents radial position of the tool surface. A 0.045mmthick raised area along the tool profile can be noticed inFig. 9 because a layer of work material adhered to the toolsurface. As the lobe plows through the workpiece, the workmaterial builds up on its front side. The tool profiles are ameasure of the balance between tool wear and work-material adhesion to the tool. The continual transfer ofwork material between the tool and workpiece will explainthe fluctuations in thrust force and torque, to be discussedin Section 3.6. Along the apex of the trilobular lobe, therewas no measurable wear or material deposition afterdrilling the first hole.

The CMM measured tool profiles of the new tool andtool after drilling 5000 and 11,000 holes are shown inFig. 10. The measurement trace was made in line with theangular datum, as illustrated by the trace of dashed line inFig. 5(c). Wear of the tool after drilling 5000 and 11,000

holes can be quantified in Fig. 10. Large wear occurs at thetool center region.Wear was quantified at four locations, marked as A–D,

in Fig. 10. Point A is located at the transition of thecylindrical region to the conical region and defined at10mm from the axial datum. The wear at Point Arepresents the radial wear of the tool. This wear will affectthe hole inner diameters. Point B is the area with significantwear at the intersection of the conical and cylindricalregions. The wear at point B is measured as the maximumdistance from a point on the profile of the worn tool to theprofile of the new tool. Point C indicates the wear in themiddle of the tool center region. The wear is measured asthe distance from the profile of the worn tool to the straightline of the new tool profile in the center region. Point D isthe wear in the axial direction at the tool tip. The averagewear at A–D for 12 measured tool profiles are shown inFig. 11. Six data points for every 2000 holes after hole

ARTICLE IN PRESS

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

10 11

Hole number (X 1000)

Wear

(mm

)

A

D

C

B

0 1 2 3 4 5 6 7 8 9

Fig. 11. Tool wear at A–D in Fig. 10.

17.26

17.28

17.30

17.32

17.34

17.36

17.38

17.40

10 11

Hole number (X1000)

Weig

ht

(g)

00 1 2 3 4 5 6 7 8 9

2

4

6

8

10

12

14

Weig

ht

loss p

er

1000 h

ole

s (

mg

)Tool weight Tool weight loss per 1000 holes

Fig. 12. Tool weight and weight loss at different stages of tool wear.

S.F. Miller et al. / International Journal of Machine Tools & Manufacture 47 (2007) 1636–16451642

number 1000 are used to illustrate the relative change intool wear at four representative locations on the tool.

Repeatability of eight CMM measurements of the toolwear were conducted at A–D. Standard deviations of theseeight measurements at A–D were 1.88, 9.67, 7.17, and6.19 mm, respectively. This demonstrates that the repeat-ability of CMM measurement was adequate compared tothe amount of wear, 16.9, 129, 231, and 297 mm at A–D,respectively.

Wear was the smallest at Point A. The wear wasnegligible before hole number 3000. It steadily increasesto 0.02mm after 11,000 holes. After drilling 1000 and 3000holes, wear was the largest at Point B, an area with largeinitial tool wear. The optical micrograph of tool wear inFig. 8(b) confirms such observation of high tool wear atPoint B. Starting from hole number 5000, a clear trendemerged showing increasing wear at the tool tip. The wearat Point C is steady throughout the drilling of 11,000 holes,reaching 0.22mm at the end of tool wear test. The wear atPoint D, the drill tip, is slow before hole number 5000. Thewear rate at Point D increases rapidly after drilling 7000holes. After drilling 11,000 holes, the wear at A–D are0.0169, 0.129, 0.231, and 0.297mm, respectively. It shouldbe noted that the center region supports the drill andgenerates the peak force in the process [1]. As shown in theprofile in Fig. 10 and optical micrograph in Fig. 8, shape ofthe center region was worn from a cone to a funnel shapewith concave cross-sectional profile after drilling 5000 and11,000 holes. The drill tip becomes sharper and, as a result,will generate lower peak thrust force in friction drilling.This will be discussed in Section 3.6.

3.3. Tool weight

Fig. 12 shows results of the weight measurements duringthe tool wear study. The weight ultimately changed from17.388 g for new tool to 17.299 g after 11,000 holes, whichindicates a decrease of 0.089 g. After drilling the first hole,the tool weight actually increased to 17.399 g. The increaseof 0.011 g was due to material adhesion from the workpiece

to the tool. It should be noted that every weightmeasurement was a balance of the loss due to tool wearand the gain due to work material transferred to the tool.After drilling 100 holes, the weight of tool started to dropdue to tool wear.The change of tool weight for every 1000 holes is also

shown in Fig. 12, labeled tool weight loss, scaled with theright Y-axis. The general trend was increasing tool weightloss which peaked at 12mg/1000 holes from hole number5000 to 6000. Inexplicably, after this peak the trendchanged to decreasing tool weight loss. Change of toolweight was 7mg/1000 holes from hole number 10,000 to11,000. The decreasing wear rate is likely due to theincrease of contact area, as shown in the optical micro-graphs in Fig. 8(d). The high tool wear rate at the start isconsistent with the observation of high wear rate of a new,sharp cutting tool at the start of a machining process. Afterthe initial run-in of the tool, the wear rate decreases. This isobserved in the wear of friction drilling tool. The work-material deposition to the tool because of changinggeometry of the worn tool is also a possible reason.Further study is needed to investigate this phenomenon.

3.4. Scanning electron microscopy

An SEM image of the tool tip after drilling 9000 holes isshown in Fig. 13(a). The tool outer diameter near thebottom of the figure is about 5.3mm. In the tool tip (centerregion), the funnel shape and circular grooves can be seen.The conical region has formed a rough, serrated appear-ance due to the different forms of wear. Lower inFig. 13(a), the cylindrical region can be seen with littleevidence of wear or use.The result of element detection by light element EDS

X-ray for new tool and tool after drilling 9000 holes isshown in Figs. 13(b) and (c), respectively. It should benoted the relative height or electron count of each elementis not directly quantitative. W, C, Ti, and Co, as shown inFig. 13(b), are elements that comprise the WC, TiC, andCo binder of the tool and are expected. Tool elements and

ARTICLE IN PRESS

0

100

200

300

400

500

600

700

Counts

C CoTi

W

WW

0

100

200

300

400

500

600

Counts

Fe

C

O

W

WW W

Fe

MnTiFe

a

b

c

Fig. 13. SEM characterization of friction drilling tool: (a) SEM micrograph of the tool center and conical regions and EDS analysis of elemental

composition of the tool surface for (b) new tool and (c) tool after 9000 drilled holes.

S.F. Miller et al. / International Journal of Machine Tools & Manufacture 47 (2007) 1636–1645 1643

deposited elements from the workpiece of C, Fe, and Mnare also present in Fig. 13(c), confirming material transferfrom the workpiece to the tool. Co cannot be seen in thegraph because of the larger Fe peak. O is present due toeither high-temperature oxidation in the friction drillingprocess, which is an evidence of oxidation wear, oroxidization on the drill surface after drilling. As discussedin Section 4, SEM microscopy and the chemical analysisinformation suggested that a combination of adhesive,oxidative, and abrasive wear contributes to friction drillingtool wear.

3.5. Hole inner diameters

An important effect of tool wear is the dimensionalalteration of friction drilled holes. The hole becomes toosmall when it cannot be tapped or utilized with theintended M6 fastener for the 5.3mm diameter hole drilledin this study. The CMM was used to measure innerdiameter of hole numbers 1, 5000, and 11,000 at threedepths, 0.46, 1.99, and 4.46mm from the top of the hole,for comparison. Fig. 14 shows the measured holediameters. A general trend of reducing hole diameter isobserved. Near the top of the hole, the diameter wasreduced by about 0.09mm. Almost no change is tooldiameter occurred in the middle of the hole (1.99mm fromthe top of the hole). Small change in diameter near the top

of the hole is because of the lack of wear in the cylindricalregion. An important observation in this study is that, evenwith the tool worn in the center and conical regions, thehole diameter is largely determined by the tool cylindricalregion, which does not exhibit a significant wear. There-fore, the hole diameter in the top of the hole, does notchange drastically.The most notable change in diameter between hole

numbers 1 and 11,000 is 0.29mm, which is near the bottomof the bushing at a depth of 4.46mm. The wear in the toolconical region causes the decrease in hole diameter. Twiceof the 0.13mm wear at Point B, as shown in Figs. 10 and11, matches well with the 0.29mm diametrical reduction ofthe hole.In this study, no particular limit was determined for the

reduction of hole size. Such criterion is usually applicationdependent. Even with 0.29mm diametrical reduction of thehole size at the bottom of the bushing after 11,000 holes,the hole drilled could still be tapped and coupled with theM6 fastener due to the thin wall thickness near the bottomof the bushing.

3.6. Thrust forces and torque

The measured thrust force and torque for friction drilledhole numbers 1, 2, 4000, and 11,000 are shown in Fig. 15.These hole numbers were chosen to demonstrate the effect

ARTICLE IN PRESS

0

500

1000

1500

2000

2500

Fo

rce

(N

) 1

11000

2

4000

00

0 2 4 6 8

1

2

3

4

5

6

7

Time from contact (s)

To

rqu

e (

N-m

)

1

11000

2

4000

10

Distance from contact (mm)

0.5 1 1.5 2 2.5

Fig. 15. Thrust force and torque of hole numbers 1, 2, 4000, and 11,000.

0

1

2

3

4

54.4 4.6 4.8 5.0 5.2 5.4 5.6

Ho

le d

ep

th f

rom

to

p (

mm

)

1 5000 11000

Hole diameter (mm)

Fig. 14. Hole diameter at hole depths of 0.46, 1.99, and 4.46mm for hole

numbers 1, 5000, and 11,000.

1400

1500

1600

1700

1800

Th

rus

t F

orc

e (

N)

Hole number (X 1000)

0 1 2 3 4 5 6 7 8 9 10 11

Fig. 16. Peak thrust force at different stages of tool wear.

S.F. Miller et al. / International Journal of Machine Tools & Manufacture 47 (2007) 1636–16451644

of tool wear on drilling process. The horizontal axisrepresents the time and distance of tool travel from theinitial contact between tool and workpiece. General shapesof force and torque are similar to those described in Ref.[1]. The secondary peaks in thrust force and torque afterhole drilling can be identified. After penetration, theshoulder of the tool, which is marked in Fig. 5(b),contacted the back-extruded work-material and createdthe secondary peak [1]. Maximum peak forces were 1600,2100, 1700, and 1400N for holes 1, 2, 4000, and 11,000,respectively. After drilling the first hole, the force increased

significantly in the second and following holes. This islikely due to the adhesion of work material on the tool, asshown in Figs. 8(a) and 9, which increased the size of thetool and changed the frictional contact interaction betweenthe tool and workpiece. The first hole was drilled withcarbide tool sliding on steel workpiece. Every hole afterthat was a combination of carbide tool, transferred steelwork material, and possibly an oxide layer sliding on steelworkpiece. The changing condition of work-materialexchange to and from the tool caused variability of thethrust force and torque due to varying coefficient offriction, size of the tool, and amount of work materialtransferred between the tool and workpiece. No clear trendon the effect of tool wear in torque can be seen. The reasonfor this is thought to be the varying frictional contactcondition from hole to hole.The peak thrust force decreased at later stages of tool

wear, as shown in Fig. 16. This surprising observation wascredited to the change of tool tip geometry, as shown inFig. 10 after drilling 11,000 holes. The worn tool tip issharp with the concave cross-sectional shape in the centerregion. This self-sharpening tool tip in friction drilling andthe reduction in peak thrust force was an interestingphenomenon caused by the tool wear in friction drilling.The peak thrust force was the lowest at the end of the11,000 hole test. At about 5000 holes, the peak thrust forcebecame smaller than that of the first hole.

4. Friction drilling tool wear mechanisms

Microscopy of the tool surface, augmented by energydispersive X-ray analysis, suggest that adhesive, oxidative,and abrasive wear all occur to some extent during frictiondrilling; however, it is difficult to determine their propor-tional contributions [9]. They are described in theirproposed order of importance; namely,

�

Adhesive wear. In friction drilling, most materialtransfer is observed to be from the workpiece to thetool. Work-material adhesion to the tool duringmachining is a familiar phenomenon [10]. Due to thehigh temperature of the process, the Co matrix in thetool material will soften, be removed from the tool, and

ARTICLE IN PRESSS.F. Miller et al. / International Journal of Machine Tools & Manufacture 47 (2007) 1636–1645 1645

lose hard WC particles embedded in the matrix.Adhesive wear arises from the strong adhesive forcescreated whenever atoms come into intimate contact.When these adhesive forces are greater than the shearstrength of either bulk material, a break is likely withinone of the materials [11].

� Oxidative wear. Oxygen was detected on the tool by

EDS X-ray. High temperature increases potential foroxidation in normal atmospheric air. Oxidative wearoccurs when a reaction with oxygen produces an oxidelayer on the surface of the tool. This is likely to happenafter every hole drilling when the tool has fresh surfaceunder high temperature. The removal rate and wear ofthe oxide layer are important to determine the contribu-tion of oxidative wear.

� Abrasive wear. The circular grooves in the tool conical

and center regions suggested abrasive wear. This is theform of wear which occurs when the hard particles ofWC, that were dislodged due to adhesive wear, slide onthe surface of the tool and plough grooves into it. This isthe three-body wear with the three bodies being the tool,workpiece, and dislodged WC particles.

5. Conclusions

The wear of a friction drilling tool was minimal afterproducing 11,000 holes in a low carbon steel workpiece.The hard carbide tool proved to be durable. Precisemeasurements of tool dimensions indicated that the wearwas concentrated at the tool center region and at theintersection between the conical and cylindrical regions.The tool tip self-sharpened during friction drilling, whichreduced the thrust force as tool wear progressed. Accord-ing to the self-sharpening observation, the tool couldpossibly be initially shaped as a steady-state worn toolgeometry to reduce the thrust force and tool wear rateresulting in a more durable tool. The torque did not displayany obvious changes at different stages of tool wear.

Adhesive, oxidative, and abrasive wear all occur to someextent during friction drilling; however, it is difficult todetermine their proportional contributions. The relativeinfluence of these wear modes, especially those associatedwith diffusion wear, may change as the tool continues towear out.

The current results suggest further investigations. Addi-tional study of tool wear rates and mechanisms is neededfor larger numbers of tools, and with larger numbers ofholes drilled in order to establish ultimate tool failure

modes. The tribological aspects of changing frictionalcontact conditions due to transfer of work material to thetool need clarification. Diffusion wear is thought to occurfrom chemical reactions in the contact zone between thetool and the workpiece, especially at elevated temperatures,but no direct evidence for this phenomenon was found inthis study.

Acknowledgements

This research is sponsored by the US Department ofEnergy, Office of FreedomCAR and Vehicle Technologies,and was performed at Oak Ridge National Laboratory,which is managed under Contract DE-AC05-00OR22725with UT-Battelle LLC. Program management and techni-cal guidance by Dr. Phil Sklad and assistance by RandyParten, are greatly appreciated, as was the assistance ofBrian Jolly and Chris Cofer in machining.

References

[1] S.F. Miller, H. Wang, R. Li, A.J. Shih, Experimental and numerical

analysis of the friction drilling process, Journal of Manufacturing

Science and Engineering 128 (3) (2006) 802–810.

[2] S.F. Miller, P. Blau, A.J. Shih, Micostructural alterations associated

with friction drilling of steel, aluminum, and titanium, Journal of

Materials Engineering and Performance 14 (5) (2005) 647–653.

[3] S.F. Miller, J. Tao, A.J. Shih, Friction drilling of cast metals,

International Journal of Machine Tool and Manufacture 46 (2006)

1526–1535.

[4] J.A. van Geffen, Piercing tools, US Patent 3,939,683, 1976.

[5] J.A. van Geffen, Method and apparatuses for forming by frictional

feat and pressure holes surrounded each by a boss in a metal plate or

the wall of a metal tube, US Patent 4,175,413, 1979.

[6] G.J. Fernandez, L.E. Murr, Characterization of tool wear and weld

optimization in the friction-stir welding of cast aluminum 359+20%

SiC metal–matrix composite, Materials Characterization 52 (2004)

65–75.

[7] H.J. Liu, J.C. Feng, H. Fujii, K. Nogi, Wear characteristics of a

WC–Co tool in friction stir welding of AC4A+30 vol% SiCp

composite, International Journal of Machine Tools and Manufacture

45 (2005) 1635–1639.

[8] M. Kerkhofs, M.V. Stappen, M. D’Olieslaeger, C. Quaeyhaegens,

L.M. Stals, The performance of (Ti, Al)N-coated flowdrills, Surface

and Coatings Technology 69/69 (1994) 741–746.

[9] B. Bhushan, Modern Tribology Handbook: Principles of Tribology,

vol. 1, CRC Press, Boca Raton, Florida, 2001.

[10] N. Sato, O. Terada, H. Suzuki, Adhesion of aluminum to WC–Co

cemented carbide tools, Journal of the Japan Society of Powder and

Powder Metallurgy 44 (4) (1997) 365–368.

[11] E. Rabinowicz, Friction and Wear of Materials, Wiley, New York,

1965.