A

coDdoua©

K

1

th�Tt8bA[oesrclG

0d

Materials Science and Engineering A 472 (2008) 115–124

Metallurgical analysis and nanoindentation characterization ofTi–6Al–4V workpiece and chips in high-throughput drilling

Rui Li a, Laura Riester b, Thomas R. Watkins b, Peter J. Blau b, Albert J. Shih a,∗a Department of Mechanical Engineering, University of Michigan, Ann Arbor, MI 48109, United States

b Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, United States

Received 15 October 2006; received in revised form 2 March 2007; accepted 10 March 2007

bstract

The metallurgical analyses, including scanning electron microscopy (SEM), X-ray diffraction (XRD), electron microprobe, and nanoindentationharacterization are conducted to study the Ti–6Al–4V hole surface and subsurface and the chips in high-throughput drilling tests. The influencef high temperature, large strain, and high strain rate deformation on the � → � phase transformation and mechanical properties is investigated.iffusionless � → � phase transformation in the subsurface layer adjacent to the hole surface can be observed in dry drilling, but not in otherrilling conditions with the supply of cutting fluid. Nanoindentation tests identify a 15–20 �m high hardness subsurface layer with peak hardness

ver 9 GPa, relative to the 4–5 GPa bulk material hardness, adjacent to the hole surface in dry drilling. For drilling chips, the � phase is retainednder all conditions tested due to rapid cooling. On the chips, the saw-tooth feature and narrow shear bands are only formed at the outmost edgend no significant change of hardness across the shear bands can be found in nanoindentation. 2007 Elsevier B.V. All rights reserved.

ah

ecirDomphpm0p

eywords: Ti drilling; Chip; Phase transformation; Nanoindentation

. Introduction

Ti–6Al–4V is a widely used titanium (Ti) alloy in indus-ry, sharing 56% of the Ti market in the US in 1998 [1]. Tias two crystal structures: the hexagonal closed-pack (HCP)phase, and body centered cubic (BCC) � phase [1,2]. Pure

i is 100% � at room temperature. The allotropic transforma-ion from � to � phase takes place at the � transus temperature83 ◦C. For Ti–6Al–4V, vanadium (V) is added to pure Ti to sta-ilize the � phase by lowering the � transus temperature [1,2].luminum (Al) is added to increase the � transus temperature

1,2]. With 6 wt% of Al and 4 wt% V, the � transus temperaturef Ti–6Al–4V is 980 ◦C, beyond which Ti is 100% � [2]. Anxample of the microstructure of the Ti–6Al–4V used in thistudy is shown in Fig. 1. The dark region is � phase and the lightegion is � phase. Because Ti–6Al–4V is a two-phase alloy, it

an be heat treated and aged to provide exceptional propertiesike the high strength–density ratio at elevated temperature [2].enerally, higher � content results in higher creep resistance

∗ Corresponding author. Tel.: +1 734 647 1766; fax: +1 734 936 0363.E-mail address: shiha@umich.edu (A.J. Shih).

hc

caof

921-5093/$ – see front matter © 2007 Elsevier B.V. All rights reserved.oi:10.1016/j.msea.2007.03.054

nd high-temperature strength, while higher � content leads toigher density and room temperature strength [2].

The machinability of Ti–6Al–4V is poor due to the inher-nt properties of work-material, particularly the low thermalonductivity and high strength [3–6]. To increase productiv-ty and reduce cost, machining Ti at high material removalate is important and has been studied extensively [7–12].rilling is an important machining process and is usually onef the final steps in component fabrication. With the advance-ent of tool material, tool geometry, and knowledge in process

arameter selection, the high-throughput drilling of Ti alloysas been demonstrated to be technically feasible. For exam-le, high-throughput drilling of Ti–6Al–4V at 156 mm3/minaterial removal rate with up to 183 m/mm cutting speed and

.75 mm/min feed rate using a 4 mm diameter WC-Co spiraloint drill has been demonstrated feasible [13]. Effects of suchigh material removal rate drilling on the machined surface andhip are investigated in this study.

In high-throughput drilling of Ti–6Al–4V, the workpiece and

hips undergo large deformation at high strain rate and temper-ture, which can alter the microstructure and material propertiesf the Ti work-material and chips. The high temperature and theollowing cooling process may cause the � → � phase trans-

116 R. Li et al. / Materials Science and Eng

Fr

fodiwdpdrmh

rsoIhui

bicnt[tln

if[aSgss

TbhemotrtActh

eCms

2

2

isglfddsTpi

ttihitsynthetic CIMTECH 500 metalworking fluid.

In Exps. III and IV, the cutting speed was reduced and feed perrevolution was increased while maintaining the same 0.75 m/minfeed rate and internal cutting fluid supply. In Exp. III (91 m/min

Table 1Experimental conditions for high-throughput drilling tests with 0.75 m/min feedrate

Exp. I Exp. II Exp. III Exp. IV

Cutting speed (m/min) 183 183 91 61

ig. 1. SEM micrograph of an etched microstructure of Ti–6Al–4V with darkegion for HCP � phase and light region for BCC � phase.

ormation. Bayoumi and Xie [14] reported the disappearancef � phase in scanning electron microscopic (SEM) and X-rayiffraction (XRD) study of Ti–6Al–4V chips generated in turn-ng. Cantero et al. [15] reported the formation of an “� case”here the � phase is absent adjacent to the hole surface in thery drilling of Ti–6Al–4V. In high-throughput drilling, the tem-erature is higher and plastic deformation is more severe thanrilling at lower speed and feed rate. One of the goals of thisesearch is to gain better understanding of the phase transfor-ation on surface and subsurface layer of the hole and chip in

igh-throughput drilling of Ti–6Al–4V.The change of V concentration due to diffusion has been

eported as another cause of the phase transformation in theubsurface layer adjacent to the hole surface in deep-hole drillingf Ti–6Al–4V at a relatively low feed rate, 0.04 m/min [16].n this study, the feed rate of high-throughput drilling is muchigher, 0.75 m/min. Since the drilling time is short, diffusion isnlikely to occur. However, the change in V concentration isnvestigated in this study.

Mechanical properties in the hole subsurface can be affectedy the combination of phase transformation, thermal soften-ng, and strain hardening during drilling. This layer is usuallyalled the subsurface layer [3,17] or deformed layer [18] under-eath the machined surface. The microhardness evaluation ofhe cross-section of drilled Ti–6Al–4V holes by Cantero et al.15] found a 125 �m thick hardened subsurface layer. High-hroughput drilling is expected to create a narrower subsurfaceayer due to the shorter drilling time. Microhardness testing isot adequate and nanoindentation evaluation is utilized.

Serrated or saw-tooth type chips are generated in Ti machin-ng. Most of researchers attributed the adiabatic shear bandormation for the observed serrated chips [19–25]. Nakayama26] and Vyas and Shaw [27] proposed that the crack initiatedt the chip free face was the cause of serrated chip formation.

heikh-Ahmad et al. [28] combined these two theories and sug-ested that flow instability and shear band formation caused theaw-tooth formation at high cutting speed, while at low cuttingpeed, the cracking followed by shear localization dominated.

FCTP

ineering A 472 (2008) 115–124

he high temperature and shear strain concentrated in the shearands are expected to influence the mechanical properties. Highardness across the shear band was reported by Sheikh-Ahmadt al. in turning of commercially pure (CP) Ti [28]. Until now,ost studies of Ti machining chips were constrained to orthog-

nal cutting in which the cutting conditions were constant alonghe tool cutting edge. Drilling has a more complicated materialemoval process. During the chip formation in drilling, the cut-ing speed and rake angle vary along the cutting edge of the drill.s a result, complicated chip morphology at different stages of

hip formation process is created. The chip morphology andhe associated changes in material properties are investigated inigh-throughput drilling of Ti–6Al–4V.

In this study, metallurgical studies, including SEM, XRD,lectron microprobe, and nanoindentation tests were conducted.hanges in microstructure, phase, chemical composition, andechanical properties are discussed on the drilled hole surface,

ubsurface layer, and drilling chips.

. Experimental setup

.1. High-throughput drilling of Ti–6Al–4V

High-throughput drilling of a 6.35 mm thick Ti–6Al–4V platen the as-rolled condition was conducted using a Kennametalpiral point drill with 3.97 mm diameter (K285A01563) andrade K715 WC tool material with 9.5% Co. This drill has aow negative rake angle chisel edge which can reduce the thrustorce and make the drill self-centering [29]. Four drilling con-itions, marked as Exps. I–IV in Table 1, were performed. Therill feed rate was the same, 0.75 m/min, for all tests in thistudy. The drilling time was only 0.57 s to penetrate the 6.35 mmi–6Al–4V plate. The associated peripheral cutting speed, feeder revolution, and cutting fluid supply in Exps. I–IV are listedn Table 1.

The application of cutting fluid internally from the through-he-drill holes was critical to increase the tool life [13]. Theool life, quantified by the number of holes drilled, is also listedn Table 1. In Exp. I, without using any cutting fluid, only 10oles could be drilled. Under the same drilling conditions withnternal cutting fluid supply, the tool life improved 10 timeso 101 holes in Exp. II. The cutting fluid was water-based 5%

eed (mm/rev) 0.051 0.051 0.102 0.152utting fluid supply No Yes Yes Yesool life (number of holes) 10 101 205 164ercentage of � phase 0 13.6 7.8 11.4

R. Li et al. / Materials Science and Engineering A 472 (2008) 115–124 117

4V sp

cbIEn1

2

estttH�tbpsmt

2

e

2

Cdmlooedo

XSi

S

wit

y

wctpiob

2m

p�mtf

2p

ct

Fig. 2. Macroscopic pictures of the Ti–6Al–

utting speed and 0.102 mm/rev), the tool life once again dou-led to 205 holes by reducing the cutting speed by half in Exp.I. Further reduction of cutting speed and increase of feed fromxp. III to IV (61 m/min cutting speed and 0.152 mm/rev) didot help to further improve the tool life, which was reduced to64 holes.

.2. Sample preparation

After drilling, each specimen was cut in half axially alongach drilled hole, as shown in Fig. 2(a). One of the sectionedpecimens was metallographically mounted in epoxy potting andhen polished with fine diamond compound for nanoindentationesting and chemical composition analysis. After nanoindenta-ion, the specimens were etched with a solution containing 6 mlNO3 and 3 ml HF in 100 ml H2O to distinguish the � andphase structure of Ti–6Al–4V in the SEM. The other sec-

ioned specimen was used for XRD. An original Ti–6Al–4Vulk material was also polished and etched to extract the �–�hase information in SEM and for XRD evaluation. The chips, ashown in Fig. 2(b), were cleaned by acetone, metallographicallyounted in epoxy, and sectioned and polished for nanoindenta-

ion testing and SEM study.

.3. SEM microstructure observations

The polished and etched hole and chip cross-sections werexamined by a Hitachi S-4700 SEM to identify the �–� phase.

.4. Quantitative analysis of phase content by XRD

A PANalytical XPERT-PRO MPD X-ray diffractometer withu tube (generator setting: 40 mA, 45 kV) and proportionaletector was used with parallel beam optics, incident parabolicultilayer mirror and diffraction-side 0.09◦ radial divergence

imiting slits, to examine the phase transformation of �–� phasen the cylindrical surfaces of the machined hole. Parallel beam

ptics were employed to eliminate sample surface displacementffects [30]. The beam width was confined to be less than the holeiameter in order to obtain signal from the cylindrical surfacesnly. The scan rate was 15 s/step.

icdb

ecimen: (a) hole cross-section and (b) chip.

The Rietveld method [31] was utilized for the quantitativeRD analysis of the �–� phase fraction using X’Pert High-core PlusTM software made by PANalytical. The scan profile

s refined with least squares fitting [31]:

y =∑

i

wi(yi − yci)2 (2)

here wi(= 1/yi), and yi and yci are observed and calculatedntensities of the spectrum pattern lines at the ith step, respec-ively. The calculated intensities yci are represented as:

ci = s∑

k

Lk|Fk|2φ(2θi − 2θk)PkA + ybi (3)

here s is the scale factor, k represents the Miller indices, Lkontains the Lorentz, polarization, and multiplicity factors, Fk ishe structure factor for the kth Bragg reflection, φ is the reflectionrofile function, 2θk is the measured position at the ith step, 2θi

s the calculated positions of the Bragg peak, Pk is the preferredrientation function, A is an absorption factor, and ybi is theackground intensity at the ith step.

.5. Analysis of chemical composition by electronicroprobe

Wavelength dispersive spectroscopy was used to detect theossible changes in chemical composition, especially the Al asstabilizer and V as � stabilizer, using a JEOL 8200 electronicroprobe. Spectra taken from a region less than 10 �m from

he edge of the hole surface were compared with those takenrom a region far away from the hole surface.

.6. Nanoindentation characterization of mechanicalroperties

The subsurface regions adjacent to the hole surface and thehips were narrow, only a few �m wide. Nano-indentation usinghe Hysitron triboindenterTM (by Hysitron Inc.) was applied to

nvestigate mechanical properties of the subsurface layer adja-ent to the hole surface. Array of indents were made along theirection parallel and perpendicular to the hole axis. The spacingetween each indent was 5–12 �m. For each indent, the applied

118 R. Li et al. / Materials Science and Engineering A 472 (2008) 115–124

with

lTesiiT

H

wao

3l

3

othagthslag

med

ttwograin structure within the subsurface layer is elongated parallelwith the hole edge, indicating the severe plastic deformation.Although not shown, the micrographs of Exps. III and IV showedsimilar features and dimensions as those in Exp. II.

Fig. 3. SEM micrographs of the polished and etched cross-sections

oad was 2 mN. A MTS Nano-indenterTM II was used for chips.o investigate the influence of shear bands on mechanical prop-rties, nanoindentation was conducted on the chips with distinctaw-tooth formations. For each indent, the displacement of thendenter was 150 nm. For both machines, Berkovich diamondndenters were used. The area of indents was less than 1 �m2.he hardness, H is calculated as follows [32,33]:

= Pmax

A(4)

here Pmax is the peak indentation load and A is the projectedrea of the contact at peak load evaluated from the shape functionf the indenter and the maximum indent displacement.

. Metallurgical analysis of hole surface and subsurfaceayer

.1. Microstructural observations

Fig. 3 shows the SEM micrographs of the hole cross-sectionsf Exps. I and II, which have the same drilling peripheral cut-ing speed and feed per revolution. In Exp. I (dry drilling), theole surface is rough with debris welded onto the hole surfacend 10–15 �m thick subsurface layer. In this layer, there is aradation of distinct grain boundaries between � and � phaseso indiscernible ones from the bulk toward the hole surface. Theigh magnification micrograph of the subsurface layer shows

ome tiny acicular or elongated grains. This phenomenon isikely due to two factors: one is the � → � phase transformations a result of high temperature and the other is the decrease ofrain size due to large plastic deformation. Because the ther-

Ffw

drilled holes in Exps. I (dry) and II (internal cutting fluid supply).

al conductivity of Ti–6Al–4V is low, the heat affected zone isxpected to be significantly narrower than 125 �m observed inrilling at lower peripheral cutting speed (0.04 m/min) [15].

In Exp. II, the � structure is observable at regions very closeo the hole surface, suggesting that the extent of the � → � phaseransformation is significantly reduced due to the better coolingith internal cutting fluid supply. The subsurface layer is aboutnly 3 �m. As shown in the high magnification micrograph, the

ig. 4. X-ray diffraction patterns of the as-polished bulk material and hole sur-ace of Exp. I (intensity plotted as square root of counts to help distinguish theeak peaks from the background).

R. Li et al. / Materials Science and Engineering A 472 (2008) 115–124 119

dge o

3t

artTp

pdtt

Fig. 5. X-ray diffraction spectra of as-polished bulk material and hole e

.2. Quantitative XRD analysis of the β → α phaseransformation

X-ray diffraction analysis results of as-polished bulk materialnd drilled hole surface of Exp. I are shown in Fig. 4. The X-

ay diffraction pattern of the as-polished bulk material showshe BCC � phase has a (1 1 0) preferred orientation/texture.he HCP or phase is also textured on the (1 0 0) and (0 0 2)lanes.

nr(p

Fig. 6. Nanoindentation on the workpiece cross-section in Exp. I

f Exps. I–IV and calculated profiles from 30◦ to 45◦ diffraction angle.

In Exp. I (dry drilling), compared to the pattern of as-olished bulk material, the � peaks are clearly broader due to theecreased size of crystallite and likely rms strains (i.e., nanos-rains) [34], resulting from the severe plastic deformation inhe subsurface layer. Drilling process also changes the origi-

al texture and makes the � crystallites in the material moreandomly distributed. The peaks corresponding to � (1 0 0) and0 0 2) planes are weakened relative to the (1 0 1) planes. Alleaks corresponding to � phase disappear. Because the most

: (a) array of indents and (b) indents close to the hole edge.

1 d Engineering A 472 (2008) 115–124

p4tdc

tt1i�o�b

wtmpteXtbm

3

ds(VsaVftdcftlcea

TCsb

AV

20 R. Li et al. / Materials Science an

rominent changes are concentrated in the range of 30◦ and5◦ 2θ diffraction angle, the close-up view of the XRD pat-erns along with the fitted profiles of the bulk material and fourrilled hole surfaces of Exps. I–IV are shown in Fig. 5 for mutualomparison.

Quantitative analysis by Rietveld method shows 7.3% � inhe bulk material, as labeled in Fig. 5. In Exp. I, the � is reducedo 0%. The X-ray penetration depth is estimated to vary from 4 to2 �m at 30–90◦ 2θ, respectively, for 95% of the total diffractionntensity. There are two possible causes for the disappearance of

peaks. The main cause is the transformation of a large portionf the � to � phase. The other cause is that the residual broadenedpeaks are submerged by broadened neighboring or peaks and

ecome indiscernible.The amount of � phase increases to 13.6, 7.8, and 11.4%

ith the introduction of cutting fluid for Exps. II–IV, respec-ively. These values are higher than the 7.3% � phase in the bulk

aterial and indicates that only a small amount, if any, of the �hase transformed to � on the hole surface during drilling dueo the better cooling with cutting fluid. Because of the broad-ned � peaks, the peaks corresponding to � are not obvious inRD patterns of Exps. II–IV as in Fig. 5. Since only one or two

ests were conducted for each hole, quantitative results may note determinative, but the difference between Exps. I and II isanifested.

.3. Chemical composition analysis

The Al and V contents were measured at seven points ran-omly chosen within a region less than 10 �m from the holeurface and seven points in the region far from the hole surfacei.e., the bulk) in Exps. I and II. For comparison, the Al and

contents were also measured in the bulk material. Table 2ummarizes results of average and standard deviation of the Alnd V contents. No statistically significant change of Al and

composition can be observed between regions close to andar away from the hole surface, which supports the hypothesishat the � phase decomposition under dry drilling condition is aiffusionless transformation (the transformation without chemi-al composition change). The same diffusionless transformationor the decomposition of � phase is also observed in the heatreatment of Ti–6Al–4V [2]. Any � → � transformation is more

ikely due to the high temperature. This matches with the sameonclusion observed in turning of Ti–6Al–4V [14] but is differ-nt from the result reported in Ref. [16] for drilling of Ti–6Al–4Vt low feed rate.

able 2omparison of Al and V content in the chemical analysis of cross-sectional

ubsurface layer (within 10 �m from the hole surface) in Exps. I and II and theulk material

Bulk Exp. I Exp. II

Avg. Std. Avg. Std. Avg. Std.

l (%) 6.7 0.3 6.7 0.6 6.5 0.2(%) 4.0 0.8 4.1 0.1 4.1 0.5

Fi

3

of

ig. 7. Nanoindentation hardness profile of subsurface adjacent to the hole edgen Exps. I–IV.

.4. Nanoindentation hardness

Fig. 6 shows the nanoindents and associated hardness valuesf the hole cross-section in Exp. I, which has a 15–20 �m subsur-ace layer of almost 0% � phase (Fig. 3). Fig. 6(a) shows a matrix

nd En

ohAtahu

thoutwhsI

r[em�ct

atshtt

R. Li et al. / Materials Science a

f indents in directions roughly parallel and perpendicular to theole axis. The spacing between each indent is about 8–12 �m.nother set of nanoindentations with a line of nanoindents close

o the hole surface is shown in Fig. 6(b). Higher hardness valuesre found at indents closer to the hole surface. Adjacent to theole surface, the hardness can reach 10.0 GPa. All indents arender the same 2 mN load.

Fig. 7 compares the nanoindentation hardness, H, versus dis-ance from the hole surface in Exps. I–IV. The layer of highardness exists in all conditions, but the extent and the depthf high hardness layer are different. The highest hardness existsnder the dry drilling condition in Exp. I. At the region lesshan 5 �m from the hole surface, the hardness is over 8 GPa,hich is about twice the hardness measured in the bulk. The

igh hardness layer is about 15–20 �m wide, which is in rea-onable agreement with the microstructural observations of Exp.as in Fig. 3. The thickness of hardened layer is significantly nar-

dIo

Fig. 8. SEM micrographs of chip morphology in Exp. I: (a) spiral con

gineering A 472 (2008) 115–124 121

ower than prior results observed in drilling at moderate speed15] and demonstrates a benefit of high-throughput drilling. Thexistence of this hardened layer is the result that plastic defor-ation outweighs the thermal softening. The high-temperature→ � phase transformation as stated in Section 3.2 may also

ontribute to the formation of this layer [3,15]. Beyond this layer,he hardness is stabilized around 4.1–5 GPa.

When the cutting fluid is supplied at the same cutting speednd feed (Exp. II), the peak hardness reduces to 5–5.5 GPa inhe subsurface layer close to the hole edge. This hardness is onlylightly higher than that of the bulk material. The thickness ofardened layer is also reduced to between 5 and 15 �m. Relativeo the results of Exp. I, the benefit of cutting fluid to lubricatehe tool–chip interface and reduce the temperature and plastic

eformation on the hole surface is apparent. For Exps. III andV, no significant change in the hardness profile from Exp. II isbserved. The peak hardness remains in the range of 5–5.5 GPa

e chip, (b) outer edge of the chip, and (c) inner side of the chip.

1 d Eng

a5tas

4

sdcflic

4

pc

snwsd

cSfAtTee

bt

F

22 R. Li et al. / Materials Science an

s in Exp. II. The thickness of machining affected layer are about�m, which is smaller than that of Exp. II. The decrease of layer

hickness is possibly the result of smaller plastic deformation andlower strain hardening effect at high feed and lower cutting

peed.

. Drilling chips

In Section 3, significant differences in hardness and depth ofubsurface layer are observed between dry drilling (Exp. I) andrilling with supply of cutting fluid (Exp. II) at 183 m/min highutting speed and 0.051 mm/rev feed. With the supply of cuttinguid, different cutting speeds and feeds in Exps. II–IV do not

nfluence the product quality metallurgically. As a result, onlyhips in Exps. I and II are investigated.

.1. Chip morphology

In all Ti–6Al–4V drilling experiments in this study, the spiraloint drill generates the same chip morphology: a continuouship with spiral cone followed by folded long ribbon [13]. The

aeoa

ig. 9. SEM micrographs of chips in: (a) Exp. I (dry drilling), (b) Exp. II (internal cu

ineering A 472 (2008) 115–124

piral cone is generated at the start of drilling from the begin-ing of contact to the full engagement of drill cutting edge withorkpiece. Due to the increased resistance to eject the chip, the

piral cone is changed to folded ribbon chip with the increase ofrilling depth.

Serrated chip formation with the saw-tooth shape surface isommonly observed in orthogonal turning of Ti alloys [17–23].urprisingly, as shown in Fig. 8, the saw-tooth is not a commoneature on the chip surface in high speed drilling of Ti–6Al–4V.s labeled as the point F in Fig. 8(a) and magnified in Fig. 8(b),

he saw teeth can be seen only at the outmost edge of the chip.his region is generated by the outmost point on the drill cuttingdge. This saw teeth region is less than 50 �m from the chipdge. It is only a very small part of the whole chip.

Other than this narrow region, the saw-tooth formationecomes indiscernible. The free surface on the chip degradeso lamellae [24], as shown by an example point G in Fig. 8(a)

nd its close-up view in Fig. 8(c). This observation is differ-nt from Ti chips formed in turning [17–23]. Contrary to therthogonal cutting, chips in drilling are not generated uniformlylong the cutting edge. The rake and inclination angles as well

tting fluid supply), and (c) close-up view of regions H (Exp. I) and J (Exp. II).

nd Engineering A 472 (2008) 115–124 123

acwidctotgr

4

BlbEpeisc

Erogc

4

bTfFTtdetsi

otnTctBmes

F(

5

psocatHlTotthUca

R. Li et al. / Materials Science a

s the cutting speed vary along the drill cutting edge. Near theenter of the drill, the strain rate is low, where plowing of theork-material occurs. The strain rate and cutting parameters

n this region are not likely reaching the critical cutting con-ition to initiate the chip saw teeth formation. This and theontinuously changing cutting conditions along the drill cut-ing edge likely inhibit the serrated chip formation in drillingf Ti–6Al–4V. At the outmost region of the drill cutting edge,he chip is less affected by the changing cutting speed and tooleometry. This can explain the saw teeth formation in this narrowegion.

.2. Microstructural observations

SEM micrographs of chip cross-sections are shown in Fig. 9.oth the saw teeth (outmost of edge of the chip) and small lamel-

ae regions of the chip are examined. As shown in Fig. 9(a and), the light � phase can be observed in both Exps. I and II.ven in dry drilling (Exp. I), which expects to have high tem-erature in the contact surface with the tool, the � phase stillxists. Under high cooling rate, the � → � phase transformations retarded [2]. The high cooling rate in chip is likely the rea-on for the � phase in retained and distributed across the wholehip.

Fig. 9(c) shows the close-up view of the chip free edge inxps. I and II, as marked by H and J in Figs. 9(a and b),

espectively. The narrow shear bands initiate from the valleyf saw-tooth chips are observed in both drilling conditions. Therain structures are elongated along the both sides of shear bands,learly identifying the severe plastic shear deformation.

.3. Nanoindentation hardness

Nanoindention on the chip cross-sections was conductedefore etching to minimize the influence of chemical etching.he thermal softening and strain hardening are two competing

actors which determine the hardness of an indent in the chip.ig. 10(a) shows an example of indents on a chip before etching.he spacing between each indent was about 5 �m. After inden-

ation, the sample was etched to expose the crystal structure andetermine if the indent is close to the shear band. An example oftched chip sample is shown in Fig. 10(b). Two indents close tohe shear band are marked by circles in Fig. 10(b). There is noignificant difference of hardness between these two and otherndents.

Over 140 and 60 indents were made in a similar configurationn the saw-tooth chip cross-sections in Exps. I and II, respec-ively. These indentation results show that the shear band doesot change the nanoindentation hardness in the drilled chip ofi–6Al–4V. This result is different form that observed of CP Tihip at low cutting speed turning [28]. At low cutting speed, the

hermal effect is weak, so the strain hardening effect dominates.esides, the CP Ti is a single phase material, no phase transfor-ation occurs. In high speed drilling of Ti–6Al–4V, both thermal

ffect and phase transformation can occur and counteract thetrain hardening effect.

tswct

ig. 10. Nanoindentation on the chip cross-section in Exp. I: (a) before etch andb) after etch (circled indents: indents close to the shear band).

. Concluding remarks

Metallurgical studies, including SEM, XRD, electron micro-robe, and nanoindentation tests were conducted on the holeurface and subsurface and the chips in high-throughput drillingf Ti–6Al–4V. High-throughput drilling decreased the size ofrystallite and likely rms strains, changed the original texturend made the � crystallites in the material more randomly dis-ributed. In dry drilling, the transformation of BCC � phase intoCP � phase was identified in a 10–15 �m wide subsurface

ayer adjacent to the hole surface by SEM and XRD analysis.his transformation was proved to be diffusionless, i.e., with-ut chemical composition change. High hardness was found inhis layer by nanoindentation testing. No obvious � → � phaseransformation occurred with the supply of cutting fluid. Theigh hardness layer was also narrower than that of dry drilling.nlike chips formed in orthogonal cutting, drilling chips had

omplicated morphology. The saw-tooth feature only formedt the outmost region of the drill cutting edge, mainly dueo the variable cutting speed along the cutting edge. Narrowhear bands could be observed in the cross-sections of chips

ith distinct saw teeth, but no significant change of mechani-

al properties along the chips was found using nanoindentationests.

1 d Eng

A

AWaVMpwaTrbc

R

[

[[

[

[[

[

[

[[

[[[

[

[

[[

[[

[[

[[32] W.C. Oliver, G.M. Pharr, J. Mater. Res. 7 (6) (1992) 1564–1580.

24 R. Li et al. / Materials Science an

cknowledgments

Assistances from Mr. Parag Hegde of Kennametal and Drs.ndrew Payzant, Paul Becker, and Ray Johnson and Mr. Larryalker of Oak Ridge National Laboratory are greatly appreci-

ted. A portion of this research was sponsored by the Heavyehicle Propulsion Systems Materials Program and Advancedaterials for High BMEP Engines Program, Office of Trans-

ortation Technologies, US Department of Energy. Experimentsere sponsored by the Assistant Secretary for Energy Efficiency

nd Renewable Energy, Office of FreedomCAR and Vehicleechnologies, as part of the High Temperature Materials Labo-atory User Program, Oak Ridge National Laboratory, managedy UT-Battelle, LLC, for the US Department of Energy underontract number DE-AC05-00OR22725.

eferences

[1] G. Lutjering, J.C. Williams, Titanium, Springer, Berlin, 2003.[2] M.J. Donachie Jr., Titanium: A Technical Guide, ASM International, Mate-

rial Park, OH, 1988.[3] A.R. Machado, J. Wallbank, Proc. Inst. Mech. Eng. Part B: Manage. Eng.

Manuf. 204 (1) (1990) 53–60.[4] E.O. Ezugwu, Z.M. Wang, J. Mater. Process. Technol. 68 (3) (1997)

262–274.[5] X. Yang, C.R. Liu, Mach. Sci. Technol. 3 (1) (1999) 107–139.[6] M. Rahman, Y.S. Wong, A.R. Zareena, JSME Int. J., Ser. C: Mech. Syst.,

Mach. Elements Manuf. 46 (1) (2003) 107–115.[7] T. Kitagawa, A. Kubo, K. Maekawa, Wear 202 (2) (1997) 142–148.[8] A.R. Zareena, M. Rahman, Y.S. Wong, 33rd International SAMPE Tech-

nical Conference—Advancing Affordable Materials Technology, vol. 33,

November 5–8, 2001, Seattle, WA, 2001, pp. 739–750.

[9] S. Lei, W. Liu, Int. J. Mach. Tools Manuf. 42 (6) (2002) 653–661.10] R. Balkrishna, Y.C. Shin, Proceedings of 2002 ASME International

Mechanical Engineering Congress and Exposition, MED 13, New Orleans,LA, 2002, pp. 277–286.

[

[

ineering A 472 (2008) 115–124

11] N. Norihiko, Chin. J. Mech. Eng. 15 (2002) 109–114.12] K. Hirosaki, K. Shintani, H. Kato, F. Asakura, K. Matsuo, JSME Int. J.,

Ser. C: Mech. Syst., Mach. Elements Manuf. 47 (1) (2004) 14–20.13] R. Li, P. Hegde, A.J. Shih, Int. J. Mach. Tools Manuf. 47 (1) (2007) 63–

74.14] A.E. Bayoumi, J.Q. Xie, Mater. Sci. Eng. Al90 (1995) 173–180.15] J. Cantero, M. Tardio, J. Canteli, M. Marcos, M. Miguelez, Int. J. Mach.

Tools Manuf. 45 (11) (2005) 1246–1255.16] L. Reissig, R. Volkl, M.J. Mills, U. Glatzel, Scr. Mater. 50 (2004) 121–

126.17] A.L. Mantle, D.K. Aspinwall, J. Mater. Process. Technol. 118 (1–3) (2001)

143–150.18] C.H. Che-Haron, J. Mater. Process. Technol. 118 (1–3) (2001) 231–237.19] M.C. Shaw, S.O. Dirke, PA. Smith, N.H. Cook, E.G. Loewen, C.T. Yang,

Machining Titanium, MIT Report, 1954.20] R. Komanduri, B.F. Von Turkovich, Wear 69 (2) (1981) 179–188.21] R. Komanduri, Wear 76 (1) (1982) 15–34.22] J.Q. Xie, A.E. Bayoumi, H.M. Zbib, Int. J. Mach. Tools Manuf. 36 (7)

(1996) 835–847.23] J. Sheikh-Ahmad, J.A. Bailey, J. Manuf. Sci. Eng. 119 (3) (1997) 307–

312.24] J. Barry, G. Byrne, D. Lennon, Int. J. Mach. Tools Manuf. 41 (7) (2001)

1055–1070.25] R. Komanduri, Z.B. Hou, Metall. Mater. Trans. 33A (9) (2002) 2995–3010.26] K. Nakayama, International Conference on Production Engineering,

August 26–29, Tokyo, Japan, 1974.27] A. Vyas, M.C. Shaw, J. Manuf. Sci. Eng. 121 (2) (1999) 163–172.28] J.Y. Sheikh-Ahmad, V. Quarless, J.A. Bailey, Mach. Sci. Technol. 8 (3)

(2004) 415–430.29] H. Ernst, W.A. Haggerty, ASME Trans. 80 (1958) 1059–1072.30] T.R. Watkins, O.B. Cavin, J. Bai, J.A. Chediak, in: T.C. Huang, et al. (Eds.),

Advances in X-Ray Analysis, V. 46 CD ROM, ICDD, Newtown Square,PA, 2003, pp. 119–129.

31] R.A. Young, Rietveld Method, Oxford University Press, New York, 1995.

33] J. Qu, L. Riester, A.J. Shih, R.O. Scattergood, E. Lara-Curzio, T.R. Watkins,Mater. Sci. Eng. A 344 (1–2) (2003) 125–131.

34] B.D. Cullity, Elements of X-Ray Diffraction, Addison-Wesley, Reading,MA, 1978.