ARTICLE IN PRESS

0890-6955/$ - se

doi:10.1016/j.ijm

�CorrespondE-mail addr

International Journal of Machine Tools & Manufacture 47 (2007) 1139–1150

www.elsevier.com/locate/ijmactool

Characterization of a hybrid laser-assisted mechanical micromachining(LAMM) process for a difficult-to-machine material

Ramesh Singh, Shreyes N. Melkote�

George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA 30332-0405, USA

Received 26 April 2006; received in revised form 18 August 2006; accepted 4 September 2006

Available online 3 November 2006

Abstract

Mechanical micro-cutting is emerging as a viable alternative to lithography based micromachining techniques for applications in

optics, semiconductors and micro-mold/dies. However, certain factors limit the types of workpiece materials that can be processed using

mechanical micromachining methods. For difficult-to-machine materials such as mold and die steels or ceramics, limited cutting tool/

machine stiffness and strength are major impediments to the efficient use of mechanical micromachining methods. In addition, at micron

length scales of cutting, the effect of tool/machine deflection on the dimensional accuracy of the machined feature can be significant. This

paper presents experimental characterization of a novel hybrid laser assisted mechanical micromachining (LAMM) process designed for

3D micro-grooving that involves highly localized thermal softening of the hard material by focusing a solid-state continuous wave laser

beam in front of a miniature cutting tool. Micro-scale grooving experiments are conducted on H-13 mold steel (42 HRc) in order to

understand the influence of laser variables and cutting parameters on the cutting forces, groove depth and surface finish. The results show

that the laser variables significantly influence the process response. Specifically, the mean thrust force is found to decrease by 17% and

the 3D average surface roughness increases by 36% when the laser power is increased from 0 to 10W. The groove depths are found to

be influenced by the machine (stage) deflection and tool thermal expansion, which affect the actual depth of cut, in the presence of

laser heating. In particular, it is found that the accuracy of groove depth improves with laser heating. Explanations for the observed

trends are given.

r 2006 Elsevier Ltd. All rights reserved.

Keywords: Micromachining; Laser assisted; Micro-grooving; Hard material

1. Introduction

There is growing need for effective ways to manufactureparts with micro- and meso-scale features that haveapplications in the fields of optics, semiconductors andmicro-molding of plastics. In response to this demand,mechanical micro-cutting (e.g. micro-grooving, micro-milling) is emerging as a viable alternative to lithographybased micromachining techniques. Lithography basedmethods are primarily limited to semiconductor materialslike silicon and are generally not suitable for creating freeform 3D shapes. They are also cost-prohibitive in manycases [1]. In contrast, mechanical micromachining methods

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

achtools.2006.09.004

ing author.

ess: shreyes.melkote@me.gatech.edu (S.N. Melkote).

such as micro-grooving/milling are capable of generating3D free-form features with micron level accuracy [2,3].Despite potential advantages, practical use of mechanicalmicromachining is limited by the properties of the work-piece and tool material. In particular, very low toolstiffness and bending strength limit the utility of mechan-ical micromachining, especially for hard-to-machine mate-rials such as mold/die steels. At micro/meso-length scalesof cutting, the effect of tool and machine (e.g. motionstage) deflections, arising from the increased cutting forces,on the dimensional accuracy can be significant. Oneapproach to address this situation is to employ the thermalsoftening ability of a laser source to heat the materialduring micro-cutting.At the conventional scale, laser assisted mecha-

nical machining has been investigated for processing

ARTICLE IN PRESSR. Singh, S.N. Melkote / International Journal of Machine Tools & Manufacture 47 (2007) 1139–11501140

hard-to-machine ceramics. Laser assisted hot machining ofceramics was initially suggested by Konig and Zaboklicki[4]. Laser assisted mechanical machining of silicon nitridehas been studied extensively [5,6]. Laser assisted mecha-nical machining of other ceramics such as sintered mulliteand magnesia-partially-stabilized-Zirconia have also beenreported [7,8]. These studies focus on material removalmechanisms, temperature prediction, wear modeling andevaluation of surface integrity. Plasma assisted milling ofsuper alloys has been reported at the macro scale [9].Kaldos and Pieper [10] have reported a two-step procedurefor die making that consists of roughing by conventionalmilling followed by finishing with a 100W Nd:YAG laser.

Pure laser micromachining is primarily limited toablation processes. The use of ultra-short pulsed (e.g.femtosecond) UV lasers for pure laser micromachining hasdrawn the attention of researchers in the past few years.These laser systems have been successfully applied tomicromachining of masks for lithography, MEMS, photo-nics, coronary stents and dental surgery [11–15]. Typicaletch rates for nanosecond and femtosecond laser vary from1 to 0.032 mm/pulse. The typical feature size is between 3and 20 mm [16]. The typical material removal rates are ofthe order of 0.1mm3/min for deep drilling of hardened steeland 1mm3/min for grooving in PMMA [17,18]. Incontrast, for pure mechanical micromachining a materialremoval rate of 25mm3/min has been reported [19].Although ultra-short laser micromachining can give finedetail, it is a slower process and much better suited forprocessing thin films. In addition, generating a sculptedsurface with complex 3D features can be difficult with thesesystems. A logical solution to this limitation is to develop ahybrid process that combines the beneficial aspects of laserheating and 3D mechanical micro-cutting.

However, no detailed studies have been reported on laserassisted mechanical machining of hard materials at themicro-scale. Initial investigations by Singh and Melkote[20,21] revealed evidence of laser softening in AISI 1018and hardened H-13 steels. Recently, Jeon and Pfefferkorn[22] have studied the effect of laser preheating on micro-end milling of metals (Al 6061 T6 and 1018 steel). Theyhowever used a conventional milling machine and milli-meter sized 100W Nd:YAG laser beam. Hard-to-cutmaterials were not studied in their work.

The work described in this paper builds on previouswork by the authors [20,21] to further understand theprocess. The present paper describes the experimental setupand characterization of a laser assisted mechanical micro-machining (LAMM) process for 3D micro-grooving of ahard mold steel with a particular focus on the cuttingforces, groove depth and surface finish.

2. Basic approach

The basic approach consists of combining a low-power(0–10W) continuous wave fiber laser with a mechanicalmicro-grooving process. The laser beam is focused in front

of a miniature cutting tool to soften the workpiece materialjust ahead of the cutting tool thereby lowering the forcesrequired to cut the material. Once modeled and under-stood, local thermal softening can be controlled andconfined to the material volume being removed by thetool thereby minimizing the thermal damage. Preliminarywork [20,21] has shown that thermal softening andconsequently lower cutting forces are observed undercertain, but not all, conditions. Initial investigation of theLAMM process applied to 1018 steel [20] revealed that theeffect of thermal softening can be offset by thermalexpansion of the tool, which results in an increased depthof cut. Other factors such as tool/machine deflections canalso alter the nominal depth of cut. Hence, the expecteddecrease in cutting force is not observed under allconditions. To further understand the LAMM process, adesign of experiments method is employed in the presentpaper to investigate the effects of laser variables andcutting parameters on the cutting forces and surface finishin micro-cutting of heat treated mold steel. The effects oftool/machine deflections and tool thermal expansion on thedepth of cut are also evaluated.

3. LAMM setup for 3D micro-grooving

A schematic of the first generation LAMM setup for 3Dmicro-grooving is shown in Fig. 1. A 10W solid-stateytterbium fiber laser (Model YLM-10) is integrated with aprecision 2-axis motion control stage (Aerotech ATS-125).The positioning resolution of the stage is 0.1 mm with 1 mmaccuracy per inch of axial travel. The only movingcomponent is the workpiece, which is mounted on thestage. All other components including the tool holder andlaser are stationary. As such, the machine is capable ofgenerating features on the order of a few microns. The laserbeam is emitted from a 7 mm diameter single mode fiberthrough a collimator. The beam has a near infraredwavelength of 1064 nm. A red aiming beam that is collinearwith the laser beam allows the laser beam to beapproximately spotted. The collimator and focusing lensare mounted on a small Y–Z stage mounted on the carriageof a precision slide. The distance of the focusing lens fromthe workpiece can be adjusted to vary the spot size of thelaser. The focal length of the lens used in the current studyis 250mm, which yields a laser spot diameter of about70 mm. The laser controller is used to modulate the laserpower. The components of the setup are mounted on analuminum base plate and the entire setup is placed on avibration isolation table.The setup is instrumented to measure the cutting forces

using a piezoelectric force dynamometer (Kistler Mini-dynes 9256C2). The tool holder is mounted on thedynamometer and holds a micro-grooving tool of300–500 mm cutting width. The cutting tool material usedin this study is tungsten carbide (WC) coated with TiAlN.The rake angle is 01, the back clearance angle is 2.51 andthe side clearance angle is 51. The workpiece is fixed to the

ARTICLE IN PRESS

Fig. 1. Schematic (top) and actual view (bottom) of first generation

LAMM setup for micro-grooving.

Fig. 2. Sinusoidal profile cut in H-13 steel (42 HRC) using 1W laser beam

with 70mm spot size located 200mm from the tool edge.

R. Singh, S.N. Melkote / International Journal of Machine Tools & Manufacture 47 (2007) 1139–1150 1141

X–Y stages that move it in the X and Y directions. Thestage has a maximum speed of 30m/min, 100mmmaximum travel, and can withstand up to 180N axialforce. The feed velocity is obtained by moving theworkpiece along the X-axis while the cutting velocity isimparted by moving it along the Y-axis.

A digital microscope is used to monitor the process andto precisely spot the laser beam at a known distance fromthe tool cutting edge. The geometric capability of themachine is illustrated in Fig. 2, which shows a whitelight interferometer image of a sinusoidal profile cut inH-13mold steel (42 HRC).

4. Experimental work

4.1. Design of experiment

The workpiece material used in this study (H-13 steel, 42HRC) finds application in the fabrication of molds and diesfor micro-injection molding of polymeric biomedicalimplants. The nominal chemical composition of H-13 steelis given in Table 1.A full factorial design of experiment was used to

investigate the effects of laser and cutting parameters onthe cutting force and surface roughness. The maximumlaser power was limited by the first generation LAMMsetup that had a 10W fiber laser system. Consequently, thethree levels of 0, 5 and 10W of laser power were selected inorder to explore the entire range available. It is knownfrom previous experiments [11] that variation in spot size inthe feasible range of the setup does not have a significanteffect on the cutting force. Hence, the laser spot size wasfixed at 70 mm. The combination of laser location, speedand depths of cut were selected so that there isapproximately 50% reduction in the flow stress of H-13tool steel due to thermal softening. The five factors andtheir respective levels are given in Table 2.The full factorial design consisted of 96 test runs. Three

replications were conducted for each test condition. Themeasured responses of the experiments included the cutting(Y) force, thrust (X) force, and the three dimensionalarithmetic surface roughness parameter Sa.

4.2. Experimental procedure

The distance of the laser from the tool was determinedby burning a spot on the workpiece with the laser andmeasuring its location relative to the tool cutting edgeusing the digital microscope as seen in Fig. 3. If thespot was not at the desired location, the laser collimatorwas translated (on the precision carriage) and a newspot burnt again. This procedure was repeated untilthe desired location of the laser beam spot from thetool edge was achieved. The normal distance betweenthe laser spot center and the tool cutting edge is indicatedin Fig. 3.The depth of cut was controlled using the X–Y stages.

G-code programming was used in the NView controllersoftware for controlling the depth of cut and imparting aconstant cutting speed. The laser controller was used to setthe laser power. The lengths of cut were 5mm for 10mm/min cutting speed and 10mm for 50mm/min cutting speed.

ARTICLE IN PRESS

Table 1

Nominal composition of H-13mold steel (wt%)

Carbon (%) Chromium (%) Manganese (%) Molybdenum (%) Vanadium (%) Silicon (%)

0.40 5.25 0.40 1.35 1.00 1.00

Table 2

Factors and their levels

Level Depth of cut (mm) Laser location (mm) Width of cut (mm) Cutting speed (mm/min) Laser power (W)

0 10 100 300 10 0

1 15 200 500 50 5

2 20 10

3 25

Fig. 3. Magnified image of laser beam location with respect to the tool

edge.

R. Singh, S.N. Melkote / International Journal of Machine Tools & Manufacture 47 (2007) 1139–11501142

5. Results and discussion

5.1. Effect of laser variables and cutting parameters on

forces

Analysis of variance (ANOVA) was carried out on theexperimental force data to identify the main effects andinteractions. The ANOVA performed on the cutting forcedata showed that the main effects of depth of cut, width ofcut and laser power were statistically significant at a risklevel (a) of 5%. Further, two-way interaction effects ofdepth of cut and laser power, depth of cut and tool width,tool width and laser power, cutting speed and laser powerand laser location and cutting speed on the cutting forcewere also found to be statistically significant at a risk (a)level of 5%.

ANOVA was also performed to identify the main effectsand interactions on the thrust force. The results showedthat the main effects of depth of cut, width of cut, cuttingspeed, laser location and laser power are statistically

significant at a 5% a level. Further, all two-way interactioneffects except the depth of cut and cutting speed, and laserlocation and laser power, were found to be statisticallysignificant at 5% a level. The main effect plots of thecutting and thrust forces are shown in Fig. 4.Analysis of the results in Fig. 4 shows a 112% increase in

the mean cutting force as the depth of cut is increased from10 to 25 mm. There is a 45% increase in the mean cuttingforce when the tool width is increased from 300 to 500 mm.The effect of laser power on the cutting force, even thoughstatistically significant, is not very pronounced. Anexplanation for this interesting observation is given laterin the paper.On the other hand, all main effects on the thrust force

are statistically significant. The mean thrust force increasesby 97% when the depth of cut is increased from 10 to25 mm. The increase in tool width from 300 to 500 mmresults in a 46% increase in the mean thrust force. Theseresults can be easily explained by noting that an increase inthe depth of cut and tool width result in a larger uncut chiparea thereby increasing the cutting and thrust forces.An 8% increase in the mean thrust force due to change

in laser location is seen. The mean thrust force drops by4% with increase in the cutting speed. Although the effectsof laser location and cutting speed are statisticallysignificant, the observed change is no more than 8%,which indicates that these factors are not as important asthe others. A 17% drop is observed in the mean thrustforce when the laser power is increased from 0 to 10W.This clearly indicates that laser heating does inducethermal softening, which in turn produces a lower thrustforce.Fig. 5 shows the mean and range of the cutting and

thrust forces for all conditions evaluated for the 300 mmtool located 200 mm from the laser beam center. Pairedstudent t-tests were performed on the difference in themean forces and it was observed that the differences incutting force with and without laser heating are statisticallyinsignificant (at a risk level of 5%) for all but the two casesof 0 and 10W laser power and depths of cut of 20 and

ARTICLE IN PRESS

15

20

25

30

35

16

14

12

10

8

0 1 2 3 0 1 0Levels

1 0 1 0 1 2

0 1 2 3 0 1 0

Levels

1 0 1 0 1 2

Thr

ust f

orce

(N

)

Depth of cut Laser location Tool width Cutting speed Laser power

Cut

ting

forc

e (N

)

Depth of cut Laser location Tool width Cutting speed Laser power

Fig. 4. Main effect plots for cutting and thrust forces.

R. Singh, S.N. Melkote / International Journal of Machine Tools & Manufacture 47 (2007) 1139–1150 1143

25 mm at 10mm/min cutting speed. The thrust force datawith and without laser heating were found to bestatistically different in all but three cases (0 and 5W laserpowers for 10 mm depth of cut at 10 and 50mm/min cuttingspeeds, and 0 and 10W laser powers for 15 mm depth of cutat 50mm/min cutting speed). When testing for the alternatehypothesis, the thrust forces at 5 and 10W laser powerswere found to be statistically lower than the thrust forcedata without laser heating, clearly indicating thermalsoftening due to laser heating.

For the cases shown in Fig. 5, the cutting force does notexhibit a large change with increase in laser power for theconditions investigated. In contrast, the thrust force is seento generally decrease with increase in laser power. Thisresult can be explained as follows. The LAMM setup has afinite stiffness and can deflect under the machining forces.Since the depth of cut is of the order of a few microns, asmall deflection of the machine and/or thermal expansionof the tool in the thrust (X) direction can easily alter thedepth of cut. The reduced thrust force due to laser heatingcauses a smaller deflection of the machine stage and this inturn yields a higher effective depth of cut than would beobtained without laser heating. In addition, thermalexpansion of the tool due to laser heating yields a higher

actual depth of cut. The cumulative increase in depth of cutdue to smaller machine deflection and tool thermalexpansion can offset the effect of reduced material strengthdue to thermal softening. Consequently, the measuredcutting forces do not always exhibit the expected decreasewith laser heating. These factors affecting the depth of cutare analyzed in detail in the next section of the paper.Note also that the decrease is more prominent in the

thrust force data than in the cutting force data due to theinherently greater sensitivity of the thrust force to changesin material strength with laser heating. This can beseen in the plot of specific cutting and thrust pressuresshown in Fig. 6. These quantities were computed bydividing the respective measured force component by theuncut chip area. It can be seen in Fig. 6 that, for a givendepth of cut, the laser power has a more noticeable effecton the specific thrust pressure than on the specific cuttingpressure.

5.2. Factors affecting the depth of cut

The main factors that can contribute to the change indepth of cut are the stage deflection, tool deflection andtool thermal expansion. These factors are analyzed below.

ARTICLE IN PRESS

02468

10121416

Without Laser

0

5

10

15

20

25

0

5

10

15

20

25

30

0

5

10

15

20

25

30

35

For

ce (

N)

For

ce (

N)

For

ce (

N)

For

ce (

N)

CuttingForce

10 mm/min

ThrustForce

CuttingForce

50 mm/min

ThrustForce

CuttingForce

10 mm/min

ThrustForce

CuttingForce

50 mm/min

ThrustForce

CuttingForce

10 mm/min

ThrustForce

CuttingForce

50 mm/min

ThrustForce

CuttingForce

10 mm/min

ThrustForce

CuttingForce

50 mm/min

ThrustForce

5 Watt Laser

10 Watt Laser

10 μm depth of cut 15 μm depth of cut

Without Laser

5 Watt Laser

10 Watt Laser

Without Laser

5 Watt Laser

10 Watt Laser

25 μm depth of cut

Without Laser

5 Watt Laser

10 Watt Laser

20 μm depth of cut

Fig. 5. Cutting and thrust forces for 300mm tool width and tool located 200mm from the laser spot.

0

0.001

0.002

0.003

0.004

0.005

0 5 10 15 20 25 30

Depth of cut (μm)

Kc,

Kt (

N/μ

m2 ) Kc (0 W)

Kt (0 W)Kc (10 W)Kt (10 W)

Fig. 6. Plot of specific cutting (Kc) and thrust (Kt) pressures calculated

from the measured force data and cutting conditions.

R. Singh, S.N. Melkote / International Journal of Machine Tools & Manufacture 47 (2007) 1139–11501144

5.2.1. Stage deflection

The setup shown schematically in Fig. 7 was designed todetermine the static stiffness of the stage in the depth of cut(X) direction. Two precision stages (Aerotech ATS-125)stacked perpendicular to each other and configured toperform coordinated X–Y motion were used in theexperiment.

A piezoelectric actuator (Physik Instruments, PI-129077)with a maximum displacement of 30 mm was used to applya known displacement to the stage. The piezoelectricactuator was fastened to an L-bracket as shown in Fig. 7and a 3-axis piezoelectric load cell (Kistler 9251A) wassandwiched between the L-bracket and the actuator. Thetip of the actuator was in contact with the workpiece,

which was mounted on the top stage. The stage wastranslated in the X direction to make initial contact withthe actuator. Once contact was made, the stage was set atthat location. The actuator displacement was controlled bythe current supplied by a variable power supply. Themeasured load vs. deflection curve in the depth of cut or X

direction is shown in Fig. 8.Fig. 8 yields a maximum deflection of 22 mm for a normal

load of 60N. The highest thrust force observed in thecurrent set of experiments is 20N, which yields a deflectionof 7 mm. Hence, in the current experiments, the actualdepth of cut can vary by up to 7 mm from the nominaldepth of cut due to stage/machine deflection. It can be seenfrom Fig. 8 that the stage deflection is strongly dependenton the thrust force. As discussed earlier, the mean thrustforce decreases by up to 17% in the presence of laserheating. This reduction in the thrust force produces asmaller stage deflection, which results in a higher actualdepth of cut compared to the case without laser heating.This increase in the actual depth of cut tends to partiallyoffset the effect of laser softening of the material and thuspartly explains the observed behavior of the cutting force.

5.2.2. Tool deflection

The load-deflection curve for the cutting tool wassimulated using the ANSYSs finite element software.The tool holder, dynamometer and the tool post are

ARTICLE IN PRESS

0

5

10

15

20

25

0 20 40 60 80Load (N)

Def

lect

ion

(m)

Loading

Fig. 8. Load vs. deflection curve for the stage in the X direction.

δx

δy

60402000

Load (N)

Def

lect

ion

(μm

)

3.5

3

2.5

2

1.5

1

0.5

Fig. 10. Load deflection curve for 300mm wide grooving tool.

Support

Load cellPiezoelectric actuator

Workpiece

X-Y stage

Fig. 7. Schematic of setup for determining static stiffness of the stage.

Fig. 9. Finite element model of the tool.

R. Singh, S.N. Melkote / International Journal of Machine Tools & Manufacture 47 (2007) 1139–1150 1145

considered to be rigid and consequently, only the tool ismodeled.

The actual shape of the 300 mm wide grooving tool wasmodeled using solid 3D 8-node brick elements (Solid 45).The modulus of elasticity of the tungsten carbide tool wastaken to be 600GPa. The model is shown in Fig. 9. All sixdegrees of freedom at nodes on the tool base and the faceof the tool in contact with tool holder were constrained.Note that the tool is mounted in the tool holder (secured byset screws) and these faces are in complete contact with thetool holder (not modeled). The applied cutting forces areuniformly distributed on the tool edge.

It is evident from Fig. 10 that the static tool deflection inthe X (depth of cut) direction is negligible. The maximumdeflection in the X direction is 0.1 mm for the extreme caserepresenting the resultant of the maximum cutting andthrust forces. This indicates that, in the current experi-

mental setup, tool deflection has a negligible effect on thedepth of cut.

5.2.3. Tool thermal expansion

The tool thermal expansion was analyzed using acombination of an analytical thermal model and finiteelement simulation. Specifically, the temperature distribu-tion derived from an analytical thermal model was used inthe ANSYSs finite element software to estimate the toolthermal expansion. The analytically determined 3D tran-sient temperature rise in the workpiece is based on thefollowing solution of a moving disk shaped Gaussian heatsource acting on a semi-infinite body with coordinatesaligned with the laser beam center [23,24]:

TðX ; y; zÞ ¼9qv

4lap3=2r2oe�XV

Z ri¼ro

ri¼0

e�ð3ri=roÞ2

ri dri

�

Z $¼v2t=4a

$¼0

d$

$3=2exp �$�

u2

4$

� �

� I0riV

2

2$

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiX þ

2$

V

� �2

þ y2

s24

35, ð1Þ

ARTICLE IN PRESS

Fig. 11. Distribution of vector sum of heat flux in the workpiece and the tool.

-150 -100 -50 0 50 100 150810

820

830

840

850

860

870

Tem

pera

ture

(°C

)

Distance along tool -workpiece interface (μm)

Fig. 12. Temperature rise at the tool-workpiece interface at 10 mm depth

of cut, 10mm/min cutting speed, 10W laser beam with 70 mm spot size

located at 100mm from the laser beam center.

R. Singh, S.N. Melkote / International Journal of Machine Tools & Manufacture 47 (2007) 1139–11501146

where q is the laser power, a the thermal diffusivity of themedium, l the thermal conductivity, ro the beam radius ofthe laser, v the scan velocity, V ¼ v=2a, ri the instantaneous

radius (varies from 0 to ro), u ¼ V

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffir2i þ X 2 þ y2 þ z2

q,

$ ¼ v2t=4a (for transient effects); for quasi-steady state

v2t=4a ¼ 5 and I0 the modified Bessel function of the firstkind of order zero and time step used in o is 0.005 s. Notethat the heat generated due to cutting induced deformationis ignored because it is very small compared to the heatgenerated by laser heating.

The tool thermal expansion analysis also assumes thatthe heat conducted to the tool is negligible. In order toverify this assumption, a finite element analysis of the toolin perfect contact with the workpiece under static conditionwas conducted to determine the fraction of heat transferredfrom the workpiece to the tool. Employing half symmetry,the tool was modeled as a 150 mm wide solid, 1mm inlength and 3mm in height using solid brick 8-nodedelements (Solid 70). The workpiece dimensions were scaleddown to 5mm� 2mm� 2mm. The spot size of the laserbeam was 70 mm (semi-circular region of laser spot ismodeled due to half symmetry) and the laser power was10W. The vector sum of the heat flux in the workpiece andtool is shown in Fig. 11. The distribution reveals that lessthan 10% of the heat is conducted into the tool while morethan 90% remains in the workpiece. Consequently, theearlier assumption ignoring the heat transfer to the tool isconsidered to be valid for the conditions investigated in thiswork and the solution of Eq. (1) can be used to analyze thetool thermal expansion.

Fig. 12 shows the distribution of temperature rise in theworkpiece after 1min at 10 mm below the surface for

10mm/min laser speed, 10W laser beam and 70 mm spotsize located 100 mm from the laser beam center calculatedusing Eq. (1). It can be seen from Fig. 12 that thetemperature rise varies from 868 1C at the center of theirradiated spot to 810 1C at the tool edge. The workpiecetemperature profile at the tool edge location obtained fromthis distribution is superimposed onto the tool edge forthermal expansion analysis. Based on thermal equilibriumconsiderations, it is reasonable to assume that the tool andworkpiece temperatures at the interface are identical. Notehowever that, in reality the tool–workpiece contact occursover a finite area, and therefore superposing the workpiece

ARTICLE IN PRESSR. Singh, S.N. Melkote / International Journal of Machine Tools & Manufacture 47 (2007) 1139–1150 1147

temperatures only on the tool edge may underestimate toolthermal expansion, particularly at large depths of cut.

Fig. 13 shows the portion of the actual tool modeled inANSYS. The circled portion is the tool overhang, which isin contact with the workpiece during cutting. The remain-ing portion fits into the slot of the tool holder. Since thethermal mass of the remainder of the tool and the toolholder are significantly higher than that of the tooloverhang portion, the temperature observed at the end ofthe tool overhang is close to ambient. The tool wasmodeled as a 300 mm wide solid, 1mm in length and 3mmin height, using solid brick 8-noded elements (Solid 70).The thermal conductivity of the tool material was takento be 42W/mK (to account for TiAlN coating that acts asa thermal barrier), the specific heat was assumed to be200 J/kgK and the coefficient of thermal expansion was

Fig. 13. Tool geometry used in thermal modeling.

Fig. 14. Thermal expansion of

6.5� 10�6/1C. Convection boundary conditions (h ¼5W=m2 K, estimated for natural convection on a heatedplate) were applied on all the tool surfaces, while at the endof the overhang (1mm from the edge of the tool, as shownin Fig. 13) the ambient temperature boundary conditionwas applied. All displacement degrees of freedom at theend face of the tool (shown in Fig. 14) were restricted in thesubsequent mechanical analysis.Fig. 14 shows the top view of the tool overhang after

model solution. It reveals that the maximum tool thermalexpansion can be as high as 2.2 mm in the extreme case (atthe center of the tool), which represents 22% of the 10 mmdepth of cut. The workpiece thermal expansion based on alinear coefficient of thermal expansion of 13.1� 10�6/1C isabout 0.05 mm, which is negligible.

5.3. Measurement of groove depth

A white light interferometer image of the cut groovegeometry is shown in Fig. 15. The difference between theperpendicular trace of the groove (shown by the double-headed arrow in Fig. 15) for the machining pass (with/without laser heating) and the preceding clean-up passgives the actual depth of cut. Note that clean-up cuts arerequired for removing the laser marks made during thelaser spot alignment process.Fig. 16 shows the profiles of the cut grooves at 10W laser

power, 10mm/min cutting speed and laser beam centerlocated at 100 mm from the tool. The upper trace in each ofthe graphs in Fig. 16 shows the groove profile created bythe clean up cut used to remove the laser burn marks. Thisprofile represents the initial condition and is marked‘‘before machining’’ in Fig. 16. The subsequent lower trace

tool solved in ANSYSs.

ARTICLE IN PRESSR. Singh, S.N. Melkote / International Journal of Machine Tools & Manufacture 47 (2007) 1139–11501148

represents the machined groove depth (the actual depth ofcut is the difference between the two traces) and is labeledas ‘‘after machining’’ in the graphs. It can be seen that, fordifferent nominal depths of cut with and without laser

Fig. 15. 3D plot of a typical micro-groove (10W laser power, laser spot

location: 100mm from the tool, 10mm nominal depth of cut, 10mm/min

cutting speed).

040

50

60

70

80

90

100

-10

-5

0

5

10

15

Hei

ght

(μm

)

Before machining

After machining ≈ 7 μm

≈ 18 μm

0.10 0.2

Distance (mm)

0.3 0.4 0.5 0.6 0.7

Hei

ght

(μm

)

0.1 0.2

Without laser

Distance (mm)0.3 0.4 0.5 0.6 0.7

Before machining

After machining

(a)

(c)

Fig. 16. Measured depth of groove machined at 10 mm nominal depth of cut (a

cut, (c) without laser heating, (d) with laser heating.

heating, the dimensional accuracy of the groove depthincreases appreciably with laser heating. Table 3 lists thenominal and measured values of the mean depth of cutwith and without laser heating and their standarddeviation. A 30% difference between the nominal andmeasured depths of cut is observed for a 10 mm nominaldepth of cut. A 28% difference is observed between thenominal and measured depths of cut for 25 mm nominaldepth of cut. In the presence of laser heating, negligibledifference is observed between the nominal and measureddepths of cut for a nominal depth of cut of 10 mm while the

-15

-10

-5

0

5

10

15

20

≈ 22 μm

≈ 10 μm

40

50

60

70

80

90

110

100

Hei

ght

(μm

)

0 0.1 0.2

10 W laser

Distance (mm)0.3 0.4 0.5 0.6 0.7

0 0.1 0.2

Distance (mm)

0.3 0.4 0.5 0.6 0.7

Before machining

After machining

Hei

ght

(μm

)

Before machining

After machining

(b)

(d)

) without laser heating, (b) with the laser heating; 25 mm nominal depth of

Table 3

Nominal and measured depths of cut

Nominal depth

of cut (mm)

Measured depth of cut

without laser (mm)

Measured depth of cut with

laser with 10W laser (mm)

10 770.76 1071.01

15 1070.81 1470.87

20 1271.15 1971.86

25 1870.99 2271.23

ARTICLE IN PRESSR. Singh, S.N. Melkote / International Journal of Machine Tools & Manufacture 47 (2007) 1139–1150 1149

difference is reduced to 12% for a nominal depth of cutof 25 mm.

It is evident from Table 3 and Fig. 16 that higher actualdepths of cut are observed with laser heating compared towithout laser heating, which, as noted earlier, can beattributed to reduced stage deflection (due to lower thrustforces arising from thermal softening) and tool thermalexpansion. As a result, the measured depths of cut with10W laser power are closer to the nominal depths of cutthan those obtained without laser heating.

5.4. Effect of laser variables and cutting parameters on

surface roughness

Three dimensional surface roughness data were acquiredby a white light interferometer based instrument (ZygoNewView 200). Fig. 17 shows a sample surface plot of themeasured surface.

To investigate the effects of the main factors and theirinteraction on the surface roughness, a full factorial designof experiments was conducted. The ANOVA performed onthe surface roughness data showed that the main effect oflaser power and depth of cut were statistically significant ata a level of 5%. The two-way interaction effects of width ofcut and laser location, depth of cut and laser location anddepth of cut and laser power were found to be statisticallysignificant at an a level of 5%.

Fig. 17. 3D surface plot of LAMM surface.

3D S

urfa

ce r

ough

ness

, Sa

(μm

)

0.95

0.90

0.85

0.80

0.75

Depth of cut Laser location To

0 1 2 3 0 1 0

Fig. 18. Main effect plot for 3D a

Fig. 18 shows the main effects plot for the 3D averagesurface roughness parameter, Sa. The surface roughness islowest for 10 mm depth of cut while an increase of 7.3% isobserved when the depth of cut is increased to 15 mm. Thesurface roughness exhibits no distinct pattern as a functionof the depth of cut. The surface roughness decreases if thedepth of cut is increased to 20 mm and increases again by4% when the depth of cut is increased to 25 mm. Anincrease of 36% is observed when the laser power isincreased from 0 to 10W. It has been observed that burrformation increases with laser heating in laser assistedmicro-milling [22], and in laser cutting of mild steel [25] thesurface roughness deteriorates with increase in laser power.The surface roughness results in the present work appear tobe consistent with the trends reported in these prior studies.

6. Conclusion

This paper experimentally characterized a new laser-assisted mechanical machining process for micromachiningapplications. The process attempts to overcome thelimitations of low tool stiffness and bending strength inpure mechanical micro-cutting and the geometry limita-tions of pure laser micromachining. The study revealedthat machine stiffness and thermal expansion of the toolcan affect the actual depth of cut in LAMM-based micro-grooving process. The following specific conclusions can bedrawn from this study:

�

Lev

ol w

ver

The main effects of depth of cut, width of cut and laserpower on the cutting force are statistically significant ata risk level (a) of 5%. The two-way interaction effects ofdepth of cut and laser power, depth of cut and toolwidth, tool width and laser power, cutting speed andlaser power, and laser location and cutting speed on thecutting force are also statistically significant.

� The main effects of depth of cut, width of cut, cutting

speed, laser location and laser power are statisticallysignificant on thrust force at a risk level (a) of 5%. Alltwo-way interaction effects except the depth of cut and

els

idth Cutting speed Laser power

1 0 1 0 1 2

age surface roughness, Sa.

ARTICLE IN PRESSR. Singh, S.N. Melkote / International Journal of Machine Tools & Manufacture 47 (2007) 1139–11501150

cutting speed, and laser location and laser power, arestatistically significant on thrust force.

� Thermal softening can be induced by laser heating and

up to 17% reduction in the thrust force is observed withlaser heating for the conditions investigated.

� The depth and width of cut influence the cutting and

thrust forces considerably (up to 112% and 45%,respectively).

� Decrease in the cutting force with laser heating is not

observed because of an increase in the actual depth ofcut due to smaller stage deflection (as a result of lowerthrust forces) and thermal expansion of the tool.

� Laser induced thermal softening improves the accuracy

of the groove depth by reducing the stage deflection(due to lower thrust force).

� The main effect of laser power and depth of cut are

statistically significant on surface roughness at a risklevel (a) of 5%. The two-way interaction effects of widthof cut and laser location, depth of cut and laser locationand depth of cut and laser power are statisticallysignificant on surface roughness.

� Laser power appears to significantly affect the surface

roughness in the LAMM process; an increase of up to36% is observed in the current study.

Acknowledgments

The authors gratefully acknowledge the support of TheTimken Company.

References

[1] T.-R. Hsu, MEMS and Microsystems Design and Manufacture, first

ed., McGraw-Hill Company, New York, 2002.

[2] T. Masuzawa, H.K. Tonshoff, Three-dimensional micromachining by

machine Tools, Annals of the CIRP 46 (2) (1997) 621–628.

[3] D. Cox, G. Newby, H.W. Park, S.Y. Liang, Performance evaluation

of a miniaturized machining center for precision manufacturing,

Proceedings of the ASME International Mechanical Engineering

Congress and Exposition, Anaheim, CA, Paper No. IMECE2004-

62186, 2004, pp. 1–6.

[4] W. Konig, A.K. Zaboklicki, Laser-assisted hot machining of ceramics

and composite materials, Proceedings of the International Con-

ference on Machining of Advanced Materials, NIST Special

Publication 847, 1993, pp. 455–463.

[5] S. Lei, Y.C. Shin, Experimental investigation of thermo-mechanical

characteristics in laser-assisted machining of silicon nitride ceramics,

Journal of Manufacturing Science and Engineering 123 (2001)

639–646.

[6] J.C. Rozzi, F.E. Pfefferkorn, Y.C. Shin, Experimental evaluation of

the laser assisted machining of silicon nitride ceramics, Journal of

Manufacturing Science and Engineering 122 (2000) 666–670.

[7] P.A. Rebro, Y.C. Shin, F.P. Incropera, Laser-assisted machining of

reaction sintered mullite ceramics, Journal of Manufacturing Science

and Engineering 124 (2002) 875–885.

[8] F.E. Pfefferkorn, Y.C. Shin, F.P. Incropera, Laser assisted machining

of magnesia-partially stabilized Zirconia, Journal of Manufacturing

Science and Engineering 126 (2004) 42–51.

[9] L.N. Lopez de Lacalle, J.A. Sanchez, A. Lamikiz, A. Celaya, Plasma

assisted milling of heat-resistant superalloys, Journal of Manufactur-

ing Science and Engineering 126 (2004) 274–285.

[10] A. Kaldos, H.J. Pieper, Laser machining in die making—a modern

rapid tooling process, Journal of Materials Processing Technology

155–56 (2004) 1815–1820.

[11] C. Momma, U. Knop, S. Nolte, Laser cutting of slotted tube

coronary stents—state-of-the-art and future developments, Progress

in Biomedical Research 4 (1) (1999) 39–44.

[12] J. Servin, T. Bauer, C. Fallnich, Femtosecond lasers as novel tool in

dental surgery, Applied Surface Science 197 (2002) 737–740.

[13] I.V. Hertel, Surface and bulk ultrashort pulsed laser processing of

transparent materials, Proceedings of the SPIE 4088 (2000) 17–24.

[14] M.R. Kasaai, V. Kacham, F. Theberge, S.L. Chin, The interaction of

femtosecond and nanosecond laser pulses with the surface of glass,

Journal of Non-Crystalline Solids 319 (2003) 129–135.

[15] F. Theberge, S.L. Chin, Enhanced ablation of silica by the

superposition of femtosecond and nanosecond laser pulses, Applied

Physics A 80 (2005) 1505–1510.

[16] M. Henry, P.M. Harrison, I. Henderson, M.F. Brownell, Laser

milling: a practical industrial solution for machining a wide variety of

materials, Proceedings of the SPIE 5662 (1) (2004) 627–632.

[17] D. Gomez, I. Goenaga, I. Lizuain, M. Ozaita, Femtosecond laser

ablation for microfluidics, Optical Engineering 44 (5) (2005) 1105–1113.

[18] G. Dumitru, V. Romano, H.P. Weber, M. Sentis, J. Hermann, S.

Bruneau, W. Marine, H. Haefke, Y. Gerbig, Metallographical analysis

of steel and hard metal substrates after deep-drilling with feratosecond

laser pulses, Applied Surface Science 208 (2003) 181–188.

[19] X. Liu, M.B.G. Jun, R.E. Devor, S.G. Kapoor, Cutting mechanisms

and their influence on dynamic forces, vibrations and stability in

micro-endmilling, Proceedings of IMECE, Anaheim, CA, Paper No.

IMECE2004-62416, 2004, pp. 1–10.

[20] R. Singh, S.N. Melkote, Preliminary investigation of laser assisted

mechanical micromachining, Proceedings of the Second JSME/

ASME International Conference on Materials and Processing,

Seattle, WA, USA, 2005, pp. 1–6.

[21] R. Singh, S.N. Melkote, Experimental characterization of laser-

assisted mechanical micromachining (LAMM) process, Proceedings

of IMECE, IMECE2005-81350, Orlando, FL, USA, MED, 2005,

pp. 1–8.

[22] Y. Jeon, F. Pfefferkorn, Effect of laser preheating the workpiece on

micro-end milling of metals, Proceedings of IMECE, Orlando,

Florida, USA, 2005, pp. 1–10.

[23] Z.B. Hou, R. Komanduri, General solutions for stationary/moving

plane heat source problems in manufacturing and tribology,

International Journal of Heat and Mass Transfer 43 (2000)

1679–1698.

[24] R. Komanduri, Z.B. Hou, Thermal analysis of laser surface

transformation hardening—optimization of process parameters,

International Journal of Machine Tools & Manufacture 44 (2004)

991–1008.

[25] A. Hiroshi, J. Suzuki, H. Kawakami, E. Hiroshi, Selection of

parameters on laser cutting of mild steel plates taking account of

some manufacturing purposes, Proceedings of the SPIE 5603 (2005)

418–425.