Characterization and prediction of the heat-affected zone in a laser-assisted mechanical micromachining process

ARTICLE IN PRESS

International Journal of Machine Tools & Manufacture 48 (2008) 994–1004 www.elsevier.com/locate/ijmactool

Characterization and prediction of the heat-affected zone in a laser-assisted mechanical micromachining process Ramesh Singh, Matthew J. Alberts, Shreyes N. Melkote The George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA 30332-0405, USA Received 1 August 2007; received in revised form 5 January 2008; accepted 15 January 2008 Available online 26 January 2008

Abstract Laser-assisted mechanical micromachining (LAMM) is a micro-cutting method that employs highly localized thermal softening of the material by continuous wave laser irradiation focused in front of a miniature cutting tool. However, since it is a heat-assisted process, it can induce a detrimental heat-affected zone (HAZ) in the part. This paper focuses on characterization and prediction of the HAZ produced in a LAMM-based micro-grooving process. The heat-affected zone generated by laser heating of H-13 mold steel (42 HRC) at different laser scanning speeds is analyzed for changes in microstructure and microhardness. A 3-D transient finite element model for a moving Gaussian laser heat source is developed to predict the temperature distribution in the workpiece material. The model prediction error is found to be in the 5–15% range with most values falling within 10% of the measured temperatures. The predicted temperature distribution is correlated with the HAZ and a critical temperature range (840–890 1C) corresponding to the maximum depth of the HAZ is identified using a combination of metallography, hardness testing, and thermal modeling. r 2008 Elsevier Ltd. All rights reserved. Keywords: Laser heating; Micromachining; Heat-affected zone; Modeling

1. Introduction There is a growing demand for engineered parts with micro- and meso-scale features for medical, optics, and electronics applications. In light of this demand, mechanical micro-cutting has the potential to be a viable alternative to lithography-based micromachining methods. Lithography-based methods are mostly limited to semiconductor materials and are not suitable for creating free form 3-D structures. They are also expensive [1]. In contrast, mechanical micro-cutting methods such as micro-grooving and micro-milling are capable of generating 3-D free-form features with micron level accuracy [2,3]. Despite its potential, practical use of mechanical microcutting is limited by the low tool/machine stiffness and cutting tool strength especially for hard materials such as heat-treated mold and die steels and ceramics. One solution to this problem is to employ the thermal softening ability of Corresponding author. Tel.: +1 404 894 8499; fax: +1 404 894 9342.

E-mail address: [email protected] (S.N. Melkote). 0890-6955/$ - see front matter r 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijmachtools.2008.01.004

a focused laser beam to heat the material during microcutting, and thus lower its strength. This in turn will decrease the forces acting on the tool thereby minimizing tool deflections and breakage. At the conventional scale, laser-assisted mechanical machining has been studied extensively for machining ceramics [4–8]. Plasma-assisted milling of super alloys at the macro scale has also been reported [9]. Kaldos and Pieper [10] have reported a two-step procedure for die making that consists of roughing by conventional milling followed by finishing with a 100 W Nd:YAG laser. However, laser-assisted mechanical machining of hard materials at the micro scale is a recent development. Singh and Melkote [11–14] investigated laser-assisted mechanical micro-grooving of AISI 1018 steel, hardened H-13 steels and 96 wt% Al2O3. Jeon and Pfefferkorn [15] studied the effect of laser preheating on micro-end milling of soft metals. Their method uses a conventional milling machine and 1-mm-sized 100 W Nd:YAG laser beam. Recent work by Singh and Melkote [13–14] has demonstrated improved dimensional accuracy of the micromachined feature in

ARTICLE IN PRESS R. Singh et al. / International Journal of Machine Tools & Manufacture 48 (2008) 994–1004

laser-assisted mechanical micromachining (LAMM)-based micro-grooving of H-13 hot work die/mold steel (42 HRC). Their approach consists of combining a relatively lowpower (p35 W) continuous wave fiber laser with a mechanical micro-grooving process. A 70–120 mm spot size laser beam is focused in front of a miniature grooving tool to soften the workpiece, hence lowering the forces required to cut the material and in turn, the tool/machine deflections. The process also has the ability to enhance material removal rates in micro-cutting and to minimize catastrophic tool failure. Despite these benefits, an important issue is the presence of a heat-affected zone (HAZ) in the micromachined part. This paper presents the characterization and prediction of the HAZ caused by laser heating in the LAMM process. Specifically, the HAZ caused by pure laser heating of H-13 mold/die steel (42 HRC) is first analyzed using metallographic methods and microhardness measurements. A 3-D transient finite element model for a moving Gaussian laser heat source is developed to analyze the temperatures generated in the workpiece and validated through actual temperature measurements. The computed temperature distribution is correlated with the HAZ via microhardness tests and a critical temperature range corresponding to the formation of the HAZ is identified. Using the model and knowledge of the critical temperature, suitable laser and cutting parameters can be established to achieve the desired amount of thermal softening with the smallest possible HAZ. The paper presents an experimental result illustrating the use of the thermal model to limit the size of the HAZ to the volume of material removed in the LAMM-based micro-grooving process. 2. Experimental work Experiments were conducted to characterize the HAZ produced in pure laser heating, i.e. without mechanical micro cutting, of H-13 steel (42 HRC). The nominal chemical composition of H-13 steel is shown in Table 1. The test setup and procedures used in the experiments are described in the following subsections. 2.1. Experimental setup A schematic of the LAMM setup for 3-D microgrooving is shown in Fig. 1 while a picture of the actual setup is shown in Fig. 2. A 35 W solid-state ytterbium-doped fiber laser (IPG Photonics, Model YLM-30, 1060 nm) is integrated with Table 1 Nominal chemical composition of H-13 steel [16] C

Cr

Mn

Mo

V

Si

0.40%

5.25%

0.40%

1.35%

1.00%

1.00%

995

a precision two-axis motion control stage (Aerotech ATS-125). The positioning resolution of the stage is 0.1 mm with 1 mm accuracy per 25 mm of travel. The only moving component is the workpiece, which is mounted on the stacked X–Y stages. A miniature grooving tool of a given width (100 mm–1 mm) is mounted on a piezoelectric cutting force dynamometer (Kistler Minidynes) and is stationary. The feed velocity is realized by moving the workpiece along the X-axis while the cutting velocity is imparted by moving it in the Y-direction. The 1060 nm laser beam is emitted from a 7 mm diameter single mode fiber through a 5 mm collimator. The collimator and focusing lens are mounted on a small Y–Z stage, which in turn is mounted on a precision slide. The distance of the focusing lens from the workpiece can be adjusted to vary the laser spot size. The focal length of the lens used in the current study is 400 mm, which yields a laser spot diameter of about 110 mm. The components of the setup are mounted on an aluminum base plate and the entire setup is placed on a vibration isolation table. A digital microscope is used to monitor the process and to precisely locate the cutting tool edge relative to the laser beam. The above setup minus the cutting tool is used to run laser scans on the surface of the H-13 steel workpiece of 85 mm  50 mm  50 mm size in order to study the resulting thermal damage. 2.2. Characterization of the HAZ In order to examine the effect of laser heating on the HAZ, experiments were conducted at 10 W laser power and three scanning speeds of 10, 50, and 100 mm/min. Fig. 3 shows the optical images of the laser-scanned H-13 workpiece surface at different speeds. The heat-affected area is clearly visible in each image. The widths of the visible heat-affected areas in Fig. 3 are approximately 100, 85, and 62 mm at 10, 50, and 100 mm/min scanning speeds, respectively. It can be seen that the heat-affected area is confined to smaller widths with increasing laser scanning speed. This is because the faster the scan, the less time there is for the heat to be conducted into the workpiece and consequently, it is confined to a smaller region. Metallographic methods were used to examine the laserscanned surfaces in cross-sectional view. The samples were first cut from the workpiece using wire-EDM, molded in epoxy, polished using 240, 320 and 400 grit SiC paper and followed by 9, 3 and 1 mm diamond slurry. The samples were finally polished with 0.05 mm alumina suspension and etched using 2% Nital solution. Fig. 4(a) shows a cross-sectional view of the laser scan at 10 W laser power, 10 mm/min scan speed and 110 mm spot size. Fig. 4(b) (top) shows the presence of a remelted zone, which is a hard brittle layer formed by solidification of the molten metal pool. The remelted zone has an uneven surface due to fracture of the brittle layer during polishing. This causes the blur observed in the picture. Fig. 4(b) (bottom) shows the tempered martensitic bulk microstructure.

ARTICLE IN PRESS 996

R. Singh et al. / International Journal of Machine Tools & Manufacture 48 (2008) 994–1004

Cutting Direction (Y-axis)

Feed (X-axis)

Workpiece

Tool Dynamometer

X-Y Precision Motion Stages

Focused Laser Beam

Tool Post Fig. 1. Schematic of LAMM-based micro-grooving setup.

Workpiece

Y

Tool Post

X Tool

Y-Z Stage

Lens Dynamometer Laser Collimator

X-Y Stage

Fig. 2. Picture of the LAMM setup for micro-grooving.

Although no sign of microstructure change is evident immediately below and around the laser-scanned region of the workpiece in Fig. 4(a), the presence of re-melted

material on the surface suggests that some microstructure change must have occurred. To determine if any microstructural change did occur and to determine the extent of

ARTICLE IN PRESS R. Singh et al. / International Journal of Machine Tools & Manufacture 48 (2008) 994–1004

997

Laser Scan 100 µm 85 µm

10 mm/min

62 µm

50 mm/min

100 mm/min

Fig. 3. Surface of H-13 steel (42 HRC) exposed to laser scans (CW, 10 W, 110 mm spot size) at 10, 50 and 100 mm/min scanning speeds.

Fig. 4. Micrographs of laser-scanned H-13 samples (CW, 10 W, 110 mm spot size, and 10 mm/min scanning speed): (a) cross section of the bulk and remelted zone at 400  magnification, (b) magnified (1200  ) view of the remelted zone (top) and bulk microstructure (bottom).

Distance from the center of the laser scan (µm) -12.5

12.5

37.5

52

54

62.5

42

44

12.5

46

50

44

42

37.5

44

46

52

50

48

25.0

48

46

50.0

48

46

44

44

42

Depth below the surface (µm)

52

44

46 48

-37.5

50

0

50

-62.5

42

the HAZ, microhardness tests were conducted below the laser scan path on the polished cross section of the sample. A contour map of the microhardness distribution was generated using a square measurement grid of 12.5 mm spacing. Fig. 5 shows a contour plot of the hardness variation for 10 W laser power and 10 mm/min scan speed. The change in hardness is clearly discernible in the microhardness contours (HRC). It can be seen that the maximum hardness observed is 54 HRC near the surface and decreases with increasing depth. The hardness approaches the bulk hardness at about 50 mm below the surface. The increase in hardness can be explained by the mechanism of laser hardening as follows. The use of high-intensity laser radiation rapidly heats the steel surface into the austenitic region described by the critical temperatures Ac1 and Ac3. For H-13 steels, austenitization starts when the temperature reaches 840 1C (Ac1) and is complete when it reaches 890 1C (Ac3). The very steep temperature gradients result in rapid cooling by conduction of heat from the surface to the bulk. This causes phase transformation from austenite to martensite without external quenching. This self-quenching occurs as the low temperature of the bulk provides a sufficiently large heat

62.5

42

42

42

42

75.0 Fig. 5. Hardness contours (HRC) in the cross-sectioned sample (10 W laser power and 10 mm/min scan speed).

sink to quench the hot surface by heat conduction to the interior at a rate high enough to prevent pearlite or bainite formation at the surface, resulting in untempered martensite formation instead [16,17].

ARTICLE IN PRESS 998

R. Singh et al. / International Journal of Machine Tools & Manufacture 48 (2008) 994–1004

In order to correlate the hardness with the workpiece temperature, a 3-D transient finite element model for a moving Gaussian laser heat source is developed and used to calculate the workpiece temperatures produced in the workpiece by laser heating. 3. Thermal modeling 3.1. Physical and mathematical description of the model The thermal model for determining the temperature distribution in the workpiece due to laser heating is based on a 3-D transient heat conduction analysis of a moving Gaussian heat source applied to the workpiece surface. The model is solved using the finite element method and makes use of certain key assumptions, equations, and modeling procedures that are summarized as follows: 1. Heat generated in the workpiece due to micro-cutting is assumed to be small compared to the heat generated by laser irradiation. 2. A scaled model of dimensions 5 mm  2 mm  2 mm is used to reduce the model size, which ensures faster solution times. 3. The Gaussian distribution of laser power intensity Px,y at location (x, y) is given by   2Ptot 2r2 exp  2 , Px;y ¼ (1) pr2b rb pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi where r ¼ x2 þ y2 is the distance measured from the laser beam center and rb is the laser beam radius; Ptot=ZPincident, where Ptot is the total absorbed power, Pincident is the incident laser power, and Z is the average absorptivity of the workpiece material. 4. The average absorptivity, Z, of the workpiece material is determined by calibrating the model against a measured temperature at a known workpiece location for given power and speed combinations. 5. All workpiece physical properties are considered to be temperature dependent. The thermal conductivity is assumed to be isotropic. The temperature-dependent thermophysical properties are given elsewhere [16]. Note that the specific heat data is estimated from alloy steel (Fe+Cr+SXi) data derived from Touloukian and Buyco [18] and the ASM Ready Reference [19]. 6. When temperature at a node exceeds the melting point, the node remains in the mesh and the latent heat of fusion is simulated by an artificial increase in the liquid specific heat as given by Brown and Song [20] and Frewin and Scott [21]. The 3-D transient heat conduction equation for the thermal problem is given by       q qT q qT q qT k k k þ þ þ Q_ qx qx qy qy qz qz qT qT þ rcp V x , (2) ¼ rcp qt qx

where r, cp, k, Q_ and Vx are the density, specific heat, thermal conductivity, rate of volumetric heat generation and laser scan speed, respectively. The initial condition at time t ¼ 0 is given as Tðx; y; z; 0Þ ¼ T 0 .

(3)

The natural boundary condition takes into account the imposed heat flux, radiation and convection at the laser irradiated surface and is given by k

qT  q þ hðT  T 0 Þ þ sðT 4  T 40 Þ ¼ 0, qn

(4)

where h is convective heat transfer coefficient, s is Stefan–Boltzmann constant ¼ 5.67  108 W/m2 K4 and e is emissivity. Note that the temperature-dependent emissivity values for H-13 steel are available in the literature [22] and have been used in the current work. The inclusion of temperature-dependent thermophysical properties and a radiation term in Eq. (4) makes the analysis highly nonlinear. To avoid this, one can use the following lumped convection coefficient that accounts for radiation and convection as suggested by Frewin and Scott [21]: h ¼ 2:4  103 T 1:61 .

(5)

Previous study [21,22] has shown only a 5% difference in the calculated temperatures using lumped convection coefficient and the explicit radiation term in Eq. (4). Consequently, the current work uses the lumped convection coefficient approach to model the combined convection and radiation heat losses. The lumped convection coefficient model given in Eq. (5) has been established over a temperature range of 25–2800 1C using temperaturedependent emissivity data reported by Gogol and Staniszewski [23] (and reproduced in the Appendix here). Similar to Frewin and Scott [21], the emissivity above 1600 1C is assumed to be constant. The boundary conditions at the remaining surfaces in the model mimic the actual block. Since only a small portion of the actual block is modeled, the heat flux on the model surfaces is assumed to be approximately equal to the heat conducted from the smaller portion of the block into the bulk. Experiments were conducted on a small block having the model dimensions stated earlier and temperatures were measured at the side and top surfaces using thermocouples (excluding the front and bottom faces). The average temperature measurements on the boundary surfaces of the scaled workpiece vary from 150 1C for 10 W laser power and 10 mm/min scanning speed to 55 1C for 5 W laser power and 100 mm/min scanning speed. The difference between the predicted maximum temperatures using ambient and measured boundary conditions (as explained above) is 10%. Consequently, in order to improve model prediction accuracy, the measured temperature values are used as boundary conditions in the model. It should be noted that one can use ambient temperature boundary

ARTICLE IN PRESS R. Singh et al. / International Journal of Machine Tools & Manufacture 48 (2008) 994–1004

value of h varies from 5 W/m2 K at 25 1C to 300 W/m2 K at 1900 1C. A contour plot of the temperature distribution on the front face of the block is shown in Fig. 7 for the case of 10 W laser power and 10 mm/min laser scan speed. The maximum temperature for the case shown is 1876 1C and the temperature at a distance of 200 mm from the center of the laser beam along the Y-axis is about 363 1C. This indicates that the temperature distribution in the semicircular region decays very rapidly. The temperature distribution for this low-speed case appears to be symmetric about the Y-axis.

conditions for greater generality, albeit with a slight loss in prediction accuracy. Since the finite element model employs half symmetry, the heat flux at the bottom face of the scaled block is set to zero. The initial workpiece temperature is set to 25 1C. 3.2. Finite element model The finite element model is created in ANSYSs (version 10) based on the physical and mathematical description of the problem given above and is shown in Fig. 6. Note that a very finely mapped mesh is used in the area where the Gaussian laser heat flux is incident. This is to capture the steep temperature gradients as accurately as possible. The size of the mapped element is 25 mm along the X-axis (length), 12.5 mm along the Y-axis (height) and 20 mm along the Z-axis (depth). The mapped mesh portion is 5 mm  200 mm  200 mm. An eight-noded 3-D thermal element (Solid70) is used. The model contains 48,610 nodes and 107,288 elements. The moving laser heat source is symmetric and hence the semi-circular Gaussian distribution of heat flux is defined by a matrix of 5  5 elements with the total area of heat flux application measuring 100 mm  50 mm. The heat flux table is computed using Eq. (1) and is made to sweep the mesh on the front face of the block to simulate the moving heat source. The heat flux matrix is swept through a distance of 1 mm, which comprises 41 load steps. The laser scan speed is realized by specifying the time interval for the total number of load steps. The time increment during one load step is an order of magnitude lower than the load step time interval. The heat transfer coefficient on the front face of the block is implemented by approximating a polynomial fit to the function given in Eq. (5). The

3.3. Model validation The model was validated by temperature measurements made using a K-type thermocouple of 75 mm diameter. Although the thermocouple bead size is larger than the grid size used in the model, the actual contact area of the thermocouple with the workpiece is much smaller, thereby enabling it to measure temperatures over a smaller area. The thermocouple was glued to the front face of the block using conducting cement (Omega Bond 400) and was calibrated using a temperature-controlled furnace. The thermocouple bead was always in contact with the metal surface as seen in Fig. 8. The laser was scanned close to the thermocouple at a constant speed and the temperature recorded. The maximum temperature is observed when the laser beam center is closest to the thermocouple. The distance between the thermocouple and laser beam center was measured using a microscope camera (see Fig. 8). The thermal model was calibrated by varying the absorptivity, Z, till the model output matched the measured temperature at a point 250 mm from the laser beam center along the Y-axis. Absorptivity obtained in this manner was

1 ELEMENTS MAT NUM

Natural B.C. on front face

JUL 19 2006 14:52:01

e t Fac

Fron

Temperature B.C. on sides esh

ed m

Y

ZY

pp Ma

X

X

Z

e

tom Bot

999

Fac

Symmetry B.C. on bottom face

Fig. 6. Finite element model of the block.

ARTICLE IN PRESS R. Singh et al. / International Journal of Machine Tools & Manufacture 48 (2008) 994–1004

1000 11

NODAL SOLUTION JUL 20 2006 10:37:08

STEP=41 SUB =10 TIME=6 TEMP (AVG) RSYS=0 SMN =150 SMX =1876

Y

X

MX

(Laser scan direction)

150

341.827

533.653

725.48

917.307

1109

1301

1685

1493

1876

Fig. 7. Simulated temperature distribution (10 W laser power, 10 mm/min scan speed, and 110 mm spot size).

Thermocouple bead

Laser scan

Fig. 8. Location of thermocouple with respect to laser scan.

between 0.4 and 0.6 (see Table 2) for the different cases examined. This is in agreement with values reported in the literature [24]. Once the absorptivity was determined, simulations were performed for three laser scanning speeds (10, 50, and 100 mm/min) and two laser powers (5 and 10 W). Figs. 9 and 10 show the predicted and measured temperatures as a function of the distance from the laser beam center at 5 and 10 W laser powers, respectively. Six temperature readings were taken at each location of the thermocouple. The error bars in the figures represent one standard deviation of the measurements. The prediction error is 4–6% at a distance of 20 mm from the laser beam center for a laser power of 5 W and between 5% and 15%, when the thermocouple is between 100 and 200 mm from the laser beam center. At 10 W laser power, it

Table 2 Absorptivity values used in simulation Laser power (W)

Scanning speed (mm/min)

Absorptivity

10 10 10 5 5 5

10 50 100 10 50 100

0.6 0.52 0.48 0.44 0.42 0.4

can be seen that the model slightly under-predicts the temperature close to the laser beam center. At 60 mm from the laser beam center, the errors are between 10% and 12% at all three speeds. It is evident from the validation results that the predicted and measured temperatures are in fairly

ARTICLE IN PRESS R. Singh et al. / International Journal of Machine Tools & Manufacture 48 (2008) 994–1004

good agreement. The prediction errors may be attributed to the uncertainty associated with the exact location of the thermocouple, approximations in the temperature1000 10 mm/min (Sim.) 50 mm/min (Sim.) 100 mm/min (Sim.) 10 mm/min (Exp.) 50 mm/min (Exp.) 100 mm/min (Exp.)

900 700 600

4. Results and discussion

500

4.1. Correlation of temperature distribution with the HAZ hardness

400 300 200 100 0 100

150

200

Distance from laser beam center (µm) Fig. 9. Simulation and experimental results for 5 W laser power.

2000 10 mm/min (Sim.) 50 mm/min (Sim.) 100 mm/min (Sim.) 10 mm/min (Exp.) 50 mm/min (Exp.) 100 mm/min (Exp.)

1800 1600 1400 1200 1000 800 600 400 200 0

100 50 150 Distance from laser beam center (µm)

0

200

Fig. 10. Simulation and experimental results for 10 W laser power.

As described earlier, the mechanism of laser hardening depends on the temperature. Experiments show that the maximum depth of the HAZ depends on the laser scanning speed, which influences the resulting temperature distribution in the material. Figs. 11 and 12 show the temperature distribution and the hardness contours in the cross section containing the HAZ for 10 W laser power and scan speeds of 10 and 50 mm/min, respectively. The horizontal axis in these plots represents the distance measured from the center of the laser scan and the vertical axis represents the depth below the surface. The units of the temperature contours in Fig. 11(a) are given in 1C and the hardness contours in Fig. 11(b) are in HRC. Fig. 11 shows that hardness decreases from 54 HRC near the center of the laser scan to 43–44 HRC (transition hardness close to the bulk hardness) at a depth of approximately 50 mm from the surface (see dashed line in Fig. 11). The width of the HAZ is about 125 mm at the surface. The temperature corresponding to the transition hardness is close to Ac1 for H-13 (840 1C). Fig. 12 shows that if the scan speed is increased to 50 mm/min, the hardness decreases from 52 HRC near the center of the laser scan to the transition hardness at a depth of about Distance from the center of the laser scan (µm)

Distance from the center of the laser scan (µm)

46 48

42

52

44

48

50

52

46

25.0

44

37.5

24

1000

800

12.5

44

Depth below the surface ((µm)

120 0 14 00

800

890

840

54

62.5 44

600

52

46

62.5

37.5 50

800

Ac1

12.5

50

0

-12.5

48

600

0 084

84

800

50.0

1000 0

-37.5

42

00

89

-62.5 0

25.0 37.5

112.5 600

600

87.5

12

89

Depth below the surface (µm)

62.5

1000

0

00

1800 16 0

37.5

14

12.5

12.5

840 890

-112.5 -87.5 -62.5 -37.5 -12.5 0

46

50.0

48

46

44

44 Ac1

62.5

42

50

0

Temperature (°C)

dependent thermophysical properties and the assumption of average temperature boundary conditions at the sides (excluding the front and bottom faces of the block shown in Fig. 7). Next, the thermal model is used to analyze the temperature distribution below the surface and its correlation with the HAZ, which is characterized by the microhardness contours obtained earlier.

50

Temperature (°C)

800

1001

42

42

42

42

00

0

60

75.0

75.0

Fig. 11. Temperature and hardness contours at 10 W laser power and 10 mm/min scan speed. (a) temperature contour (1C) and (b) hardness contour (HRC).

ARTICLE IN PRESS R. Singh et al. / International Journal of Machine Tools & Manufacture 48 (2008) 994–1004

Distance from the center of the laser scan (µm)

70

0

0 60

70

Ac1

700

50 600

0

12.5

25

37.5

50

62.5

52

46

50

48

44

20 42

0

Depth below the surface (µm)

13 1 0040 00

0

00 1 10 10

890 700

0

0

890 840 800

-25 -12.5

46

8

84

89

0 70

00

-50 -37.5

10 00

8

-62.5 0

44

800

62.5

10

0

120

1000

60 0

Distance from the center of the laser scan (µm) 50

0 80

Depth below the surface (µm)

37.5

1100

30

60

25

48

1500

40

40

12.5

800 840 890

20

0

00 11 00 12 00 13

10

-25 -12.5

00

840

-50 -37.5

14

-62.5 0

50

1002

44

48

46

30

44 Ac1

40

42

42

50 600

60

Fig. 12. Temperature and hardness contours at 10 W laser power and 50 mm/min scan speed. (a) temperature contour (1C) and (b) hardness contour (HRC).

00 11 10

840 0

0

800

0 60

28 32

700

36

0

50

0

60

0 60

0

50

40

0

0

70

0

60

70

0

24

0

0 890

84

Ac1

0 80

16

60

1000

0

80

0 70

60

0

20

84

1200

89

70

12

00

0

62.5 0 80

00

50.0

84

Depth below the surface (µm)

37.5 10

4 8

-37.5

0 89

In order to illustrate the utility of the thermal model in establishing the laser parameters and cutting conditions that minimize the HAZ in the final micromachined part in a process such as laser-assisted micro-grooving, an example case of creating a 25 mm deep groove in H-13 steel is selected. A 300 mm grooving tool is used in the LAMM setup described earlier with the laser spot located 100 mm from the edge of the grooving tool. The laser beam diameter is 110 mm. In order to select an ideal speed for the operation which results in no residual HAZ, thermal model simulations were carried out at different laser scan speeds and laser power of 10 W. Fig. 13 shows the temperature distribution at 100 mm/min scan

-50.0

00

4.2. Application of thermal model in laser-assisted microgrooving

Distance from the center of the laser scan (µm) -62.5 0

11

35 mm (dashed line in Fig. 12) from the surface. The width of the HAZ in this case is about 110 mm at the surface. It is evident from Fig. 12 that the temperature corresponding to the transition hardness is close to Ac1 for this case as well. This correspondence of the critical temperature to the transition hardness (and consequently depth of the HAZ) can be explained by the mechanism of laser hardening discussed earlier. The initial microstructure of H-13 used in the current study is tempered martensite. As predicted by the thermal model, the temperature rise in the material can even exceed the melting point, which is corroborated by the presence of the small remelted zone near the surface as shown in Fig. 4. Austenitization starts in the portion of the workpiece volume where the temperatures reach Ac1 (840 1C) and the regions where the temperatures exceed Ac3 (890 1C) are totally converted to austenite. Once the austenite is quenched by the cold interior, untempered martensite is formed, which accounts for the hardness increase.

Fig. 13. Temperature contours (1C) in the cross section of the HAZ at 10 W laser power and 100 mm/min scan speed.

speed and a laser power of 10 W. The depth corresponding to Ac1 (840 1C) is 19 mm, which should correspond to the maximum HAZ depth beyond which no phase transformation (and consequently, hardness change) should occur. Therefore, a speed of 100 mm/min was selected for the test case to investigate the presence of residual HAZ, if any. The cross section of the machined surface was analyzed for increase in hardness and/or microstructural changes related to any residual HAZ. Fig. 14(a) shows the microstructure of the machined groove cross section and the area close to the base of the groove where the hardness was measured. Fig. 14(b) shows a magnified view of the center of the groove base. Any change in microhardness near the base of the groove indicates that the microstructure in that area has changed.

ARTICLE IN PRESS R. Singh et al. / International Journal of Machine Tools & Manufacture 48 (2008) 994–1004

200X

150 µm

1003

50 µm

600X

Epoxy Groove Bulk Hardness measurement

Fig. 14. Microstructure of cross section of test cut with 300 mm grooving tool at 10 W laser power and 100 mm/min cutting speed: (a) cross section of groove, (b) magnified view of bulk near center of groove base.



45

Hardness (HRC)

40

 35 30 25

 20 0

50

100 150 200 250 Distance from edge of groove (µm)

300

Fig. 15. Hardness at the base of the groove.

Microhardness tests were conducted near the base of the groove in the area shown in Fig. 14(a). Fig. 15 shows the hardness as a function of distance measured from the left edge of the groove. The hardness measurement varies between 40.4 and 42.5 HRC, which is within the range of the base hardness (4272 HRC) of the H-13 steel workpiece used in the current work. Consequently, it can be concluded that the HAZ and any associated phase transformation was confined to the volume of material removed by the micro-grooving tool. This demonstrates that the thermal model can be used to establish and/or minimize the size of the HAZ in LAMM-based microgrooving. 5. Conclusions The paper focused on characterizing and predicting the HAZ produced by laser heating in the LAMM-based micro-grooving process. The main conclusions of this work are as follows:



The size of the HAZ depends on laser scanning speed and laser power. The maximum width and depth of the HAZ decrease by 32% and 62%, respectively, when the scan speed is increased from 10 mm/min to 100 mm/min at a laser power of 10 W.

The transient 3-D finite element model developed to analyze the temperatures in the HAZ predicts the temperature to within 15% accuracy. A critical temperature range (840–890 1C) corresponding to the depth of HAZ produced in H-13 steel (42 HRC) at different scan speeds was identified. If the temperature in the material removal region is below the critical range then there will be no residual HAZ in the material after micro cutting. The thermal model can be used to determine the laser parameters for a given cut geometry that will yield no residual HAZ in the material after micro-grooving.

Appendix. Emissivity of H-13 steel [23] Temperature (1C)

Emissivity

25 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600

0.20 0.40 0.45 0.48 0.51 0.54 0.56 0.57 0.58 0.59 0.61 0.62 0.62 0.62 0.62 0.62 0.62

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.

ARTICLE IN PRESS 1004

R. Singh et al. / International Journal of Machine Tools & Manufacture 48 (2008) 994–1004

[3] D. Cox, G. Newby, H.W. Park, S.Y. Liang, Performance evaluation of a miniaturized machining center for precision manufacturing, in: Proceedings of the 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, in: Proceedings of the International Conference 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. Lo´pez de Lacalle, J.A. Sa´nchez, A. Lamikiz, A. Celaya, Plasma assisted milling of heat-resistant superalloys, Journal of Manufacturing 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] R. Singh, S.N. Melkote, Preliminary investigation of laser assisted mechanical micromachining, in: Proceedings of the 2nd JSME/ ASME International Conference on Materials and Processing, Seattle, WA, USA, 2005, pp. 1–6. [12] R. Singh, S.N. Melkote, Experimental characterization of laserassisted mechanical micromachining (LAMM) process, in: Proceedings of the ASME International Mechanical Engineering Congress and Exposition, Orlando, FL, November 2005, pp. 1–8. [13] R. Singh, S.N. Melkote, Characterization of a hybrid laser-assisted mechanical micromachining (LAMM) process for a difficult-to-

[14]

[15]

[16]

[17]

[18] [19] [20]

[21] [22]

[23]

[24]

machine material, International Journal of Machine Tools and Manufacture 47 (2007) 1139–1150. R. Singh, S.N. Melkote, Hybrid laser-assisted mechanical micromachining (LAMM) process for hard-to-machine materials, in: Proceedings of the 4th International Congress on Laser Advanced Materials Processing (LAMP’06), Kyoto, Japan, 2006, pp. 1–6. Y. Jeon, F. Pfefferkorn, Effect of laser preheating the workpiece on micro-end milling of metals, in: Proceedings of the ASME International Mechanical Engineering Congress and Exposition, Orlando, FL, USA, 2005, pp. 1–10. R. Singh, M.J. Alberts, S.N. Melkote, Characterization of heat affected zone in a laser-assisted mechanical micromachining (LAMM) process for difficult-to-machine materials, in: Proceedings of the 1st International Conference on Micromanufacturing, Urbana, IL, 2006, pp. 1–6. E. Kennedy, G. Byrne, D.N. Collins, A review of the use of high power diode lasers in surface hardening, Journal of Materials Processing Technology 155–156 (2004) 1855–1860. Y.S. Touloukian, E.H. Buyco, Specific Heat—Metallic Elements and Alloys, IFI/Plenum, 1970. F. Cverna, ASM Ready Reference Thermal Properties of Metals, ASM, Materials Park, OH, 2005. S. Brown, H. Song, Finite element simulation of welding of large structures, Journal of Engineering for Industry 114 (11) (1992) 441–451. M.R. Frewin, D.A. Scott, Finite element model of pulsed laser welding, Welding Journal 78 (1) (1999) 15S–22S. S. Kou, D.K. Sun, Y.P. Le, A fundamental study of laser transformation hardening, Metallurgical Transactions A 14A (1983) 643–653. W. Gogol, B. Staniszewski, Thermophysical properties research in Poland, in: Proceedings of the Symposium on Thermophysical Properties, Gaithersburg, MD, USA, 1977, pp. 55–66. C. Sainte-Catherine, M. Jeandin, D. Kechemair, J.P. Ricaud, L. Sabatier, Study of dynamic absorptivity at 10.6 mM (CO2) and 1.06 mM (Nd-YAG) wavelengths as a function of temperature, Journal de Physique IV (Colloque) 1 (C7) (1991) 151–157.