Académique Documents
Professionnel Documents
Culture Documents
1INTRODUCTION
Laser processing of metals and alloys are widely used in industry, due to its
distinct advantages of high productivity, non-contact, reduced processing
cost, improved quality, greater material utilization and minimum heat affected
zone (HAZ). Macro laser material processing can be generally divided into
two major categories, those requiring phase change such as cutting, drilling,
*Corresponding author: Tel: +91 3324146372; E-mail: dipten@gmail.com
P. Kalvettukaran et al.
parameter effects such as material, laser power, beam diameter, scan velocity,
plate thickness, pass number and pulse duration on bending angle by numerical simulation and experimental investigations with Taguchi experimental
design method. Venkadeshwaran et al [15] used FEM simulation and response
surface methodology (RSM) to predict bending angle and optimize the process parameters for laser bending of stainless steel. They have identified the
optimum values of parameters that would increase the productivity and
reduce the total operating cost and the HAZ. Acherjee et al [16] presented
experimental investigation with DOE technique on laser transmission welding of dissimilar thermoplastics between PMMA and ABS. They implemented graphical optimization technique which allows identifying optimal
region in a graphical window. Olabi et al [17] developed mathematical models to determine the relationship between laser welding parameters and
the magnitude of the residual stress at different locations by using response
surface methodology. Reisgen et al [18] presented statistical analysis of the
CO2 laser beam welding of dual phase steel sheets using RSM. Their results
indicate that the proposed models predict the responses adequately within the
limits of welding parameters being used and the welding speed is the most
significant parameter during the welding process. Caydas and Hascalik [19]
presented optimization of laser cutting process of St-37 steel with multiple
performance characteristics such as workpiece surface roughness, top kerf
width and the HAZ based on the grey relational analysis. They have observed
that laser power has stronger influence on responses rather than cutting speed.
Yang et al. [20] performed experimental investigation and numerical simulation to predict depth and width of the HAZ during laser assisted machining of
Ti6Al4V alloy. They have found that the depth and width of the HAZ are
strongly dependent on laser parameters (laser power, scan speed, angle of
incidence and the beam diameter) and material properties (thermal conductivity, specific heat and density). Their parametric studies showed that the
depth and width of HAZ increased with laser power and decreased with an
increase of the laser spot size and the laser scan speed.
As noted from the literature, selection of appropriate process parameters
is important for laser processing techniques. This manuscript aims to develop
a mathematical model that can be used to predict the material temperature
for given values of process parameters. It further develops regression models
to predict the required process parameters, when the targeted maximum
temperature is known. In the present work a 3-D FEM model is developed
to simulate laser processing of commercially pure Ti metal sheet with a moving laser beam. Statistical techniques are used to construct the design layout
and to develop regression models based on results of numerical simulation.
The developed model would be of significant assistance to any laser material
processing application that involves phase change, such as, cutting, drilling,
welding, etc., as well as no phase change, such as laser bending and straightening, surface hardening, semiconductor annealing, laser shock peening, etc.
P. Kalvettukaran et al.
FIGURE 1
Schematic diagram showing the simulation work piece and irradiation path.
FIGURE 2
FEM mesh used for the modelling.
(iv) Melting is not involved in the work piece, so phase change and heat
generation are neglected.
2.2 Governing equations and boundary conditions
Due to laser heating a transient temperature field is generated based on the
mechanism of the heat conduction. The governing equation for heat conduction within the specimen can be written as
rc
T (r , t )
= kr (rT )
t
(1)
where r is the material density, c is the specific heat capacity, k is the thermal
conductivity, T(r,t) is the temperature r is the coordinate (m) in the reference
configuration, t is time and r is the gradient operator.
Material cooling occurs due to heat loss through natural convection and
radiation from its surfaces exposed to ambient air.
The convection and radiation boundary conditions can be expressed as
and
where, h is the heat transfer coefficient, Ts is the sheet metal surface temperature, T0 is the ambient temperature, e is the emissivity and is the Stefan
Boltzmann constant.
The moving heat flux, Q, produced by the laser beam is applied on the top
surface of the sheet metal. In this work the laser beam is assumed to have a
Gaussian distribution and expressed as
P. Kalvettukaran et al.
Q=
2r 2
(4)
exp
2
2
pR
R
2 AP
FIGURE 3
Temperature distribution on the top surface at plate centre where (a) and (b) present for the
model, and (c) and (d) are from Jung [23].
TABLE 1
Validation of temperature field.
Temperature at
Plate Centre (C)
Parameter
Laser
power
(W)
Scanning
speed
(mm/s)
Spot
diameter
(mm)
Plate
thickness
(mm)
Present
(ft)
Percentage
Error
Jung [23].
(fs)
(ft fs )
100
(fs )
1500
5.0
16.00
6.00
723
728
-0.69156
3000
10.0
16.00
4.00
878
844
3.76283
FIGURE 4
Temperature distribution at three instances where (a) the laser beam is at the plate centre, (b) the
laser beam is at plate end and (c) after 30 seconds when P=650W, V=5.0 mm/s, D=2.50 mm and
S=1.00 mm.
P. Kalvettukaran et al.
The maximum temperature histories along the scan line of the plate is
shown in Figure 5. It can be seen from Figure 5 that the temperature starts
rising gradually on the top surface and reaches a steady maximum along a
major part of its journey along the width of the plate. Afterwards, the temperature increases rapidly when the beam reaches near the top edge of the
plate. From Figure 5 it can be seen that a much higher temperature (1688 K)
is obtained at the end of the scan line compared to that at the beginning
(1125K). The reason for this is that heat from the incident laser beam and the
heat retained by the material behind the beam is flowing into the cold region
ahead of the beam, and as the beam reaches the edge the heat flowing ahead
of the beam cannot travel any further and so a heat build up occurs in the end
of the plate.
3.2 Response surface methodology (RSM)
In this section, again, the temperature distribution of commercially pure Ti
sheet is presented at different time intervals for P= 650W, V= 5.0mm/s,
D= 2.50mm and S=1.00mm. The plate size is taken as 100.0mm 50.0mm
(L W).
RSM is a set of mathematical and statistical technique that is useful for
empirical model and optimization. A model predicting the response for
some independent input variables can be obtained by conducting experiments
and applying regression analysis [11, 24]. If all variables are assumed to be
measurable, the response surface can be expressed as:
FIGURE 5
Graph showing the variation of local maximum temperature along the laser scan line.
y = f ( x1 , x2 , x3 ,..... xn ) z (5)
TABLE 2
Process parameters and their units and limits.
Parameter
Symbol
Unit
Low Actual
High Actual
Laser power
450
650
Scanning speed
mm/s
5.0
15.0
Spot diameter
mm
1.50
2.50
Plate thickness
mm
1.00
2.00
10
P. Kalvettukaran et al.
TABLE 3
Design layout and numerically calculated response.
Process Parameters
Sample
Response
Laser
power,
P (W)
Scanning
velocity,
V (mm/s)
Spot
diameter,
D (mm)
Plate
thickness,
S (mm)
Maximum
temperature
(K)
450
5.0
2.50
2.00
1082
650
5.0
1.50
2.00
1958
550
10.0
2.00
1.50
1289
650
15.0
2.50
2.00
1096
650
15.0
1.50
1.00
1770
368
10.0
2.00
1.50
997
550
19.1
2.00
1.50
1116
650
5.0
1.50
1.00
2260
450
15.0
1.50
2.00
1305
550
10.0
1.09
1.50
1984
550
10.0
2.91
1.50
980
450
5.0
2.50
1.00
1311
550
10.0
2.00
2.41
1243
650
5.0
2.50
2.00
1371
450
5.0
1.50
2.00
1534
550
10.0
2.00
1.50
1289
550
10.0
2.00
1.50
1289
450
5.0
1.50
1.00
1761
650
15.0
2.50
1.00
1181
650
15.0
1.50
2.00
1669
550
10.0
2.00
1.50
1289
550
10.0
2.00
1.50
1289
450
15.0
2.50
2.00
868
550
10.0
2.00
0.59
1667
550
0.9
2.00
1.50
1953
450
15.0
1.50
1.00
1386
550
10.0
2.00
1.50
1289
450
15.0
2.50
1.00
938
650
5.0
2.50
1.00
1688
732
10.0
2.00
1.50
1554
11
TABLE 4
ANOVA for response surface quadratic model.
Source
Sum of
Squares
Degree of
Freedom
Mean
Square
F-Value
p-value
Prob>F
Model
3475680.5
14
248262.8941
142.4999256
<0.0001
645561.36
645561.3587
370.5444824
<0.0001
807898.06
807898.0639
463.7238054
<0.0001
1557032.5
1557032.49
893.7179871
<0.0001
210762.45
210762.4528
120.9751218
<0.0001
PV
8556.25
8556.25
4.911184947
0.0426
PD
17822.25
17822.25
10.22975789
0.0060
PS
2450.25
2450.25
1.406414132
0.2541
VD
12.25
12.25
0.007031353
0.9343
VS
34040.25
34040.25
19.5386955
0.0005
DS
6.25
6.25
0.003587425
0.9530
P2
1410.3397
1410.339708
0.809518088
0.3825
V2
101250.37
101250.368
58.11649765
<0.0001
D2
60558.571
60558.57084
34.75989382
<0.0001
43681.116
43681.11635
25.07243723
0.0002
significant
Standard deviation=41.739629
R2=0.99253731
Mean=1413.5333
Adjusted R2=0.985572132
Coefficient of variation=2.9528578
Predicted R2=0.959443764
Adequate Precision=48.56869825
FIGURE 6
Graph showing the comparison of actual and predicted results for maximum temperature.
12
P. Kalvettukaran et al.
The mathematical model for maximum temperature in terms of actual factors as determined by Design-Expert is given:
Tmax = 2486.67930 + 4.76588P - 96.68044 V - 1030.19362D - 797.64368S
-0.04625PV - 0.66750PD - 0.24750 P S + 0.35000 VD
D
+ 18.45000 VS + 2.50000DS - 8.2541110-4 P 2 + 2.79748V 2
+ 216.34959D 2 + 183.74483S 2
(6)
PTmax =
bp + bp 2 4 a p c p
2a p
(7)
where
ap=-8.2541110-4 (7a)
bp=4.76588-0.046250V-0.66750D-0.24750S (7b)
and
cp=-(Tmax-2486.67930+96.68044V+1030.19362D+797.64368S
-0.35000VD-18.45000VS-2.500DS-2.79748V2-216.34959D2
-183.74483S2(7c)
The specified maximum temperature of the commercially pure Ti sheet is
assumed as 1941 K. The predictions of maximum laser power as obtained
from Equation (7) are compared with the FEM simulation for a few number
of process parameters to assess its accuracy. Results show that the FEM
simulation and prediction from Equation (7) are in good agreement. Comparisons of results are furnished in Table 5 and in Figure 7.
13
TABLE 5
Comparisons of maximum temperature.
Percentage Deviation
Sample
Laser
Power,
P (W)
Scanning
Speed,
S (mm/s)
Spot
Diameter,
D (mm)
Plate
Thickness,
S (mm)
Equaqtion
(7)
(K) (fa)
FEM
(K)
(fb)
753.7168
15.0
1.50
1.00
1941
1954
-0.67500
546.8736
10.0
1.09
1.50
1941
1977
-1.83100
632.7047
5.0
1.50
2.00
1941
1924
0.87800
732.3225
10.0
2.00
0.59
1941
1976
-1.74600
753.7168
15.0
1.50
1.00
1941
1954
-0.67500
1063.1131
10.0
2.00
1.50
1941
1966
-1.25700
(fa fb )
100
(fb )
FIGURE 7
Graph showing the comparison of developed model from Equation (7) and the FEM model for
maximum temperature.
Figures 8(a) to (c) show the 3-D surface plots demarcating the attainment
of maximum temperature with varying combination of laser process parameters for 1.00, 1.50 and 2.00mm thick plates. It is observed Figures 8(a) to (c)
that more laser power is required for the material to reach the specified
maximum temperature when the scanning speed increases and the laser
spot diameter is kept constant. Similarly, when the laser spot diameter is
increased at a fixed scanning speed, more power is required to attain the same
maximum temperature. The 3-D surfaces in Figures 8(a) to (c) correspond to
14
P. Kalvettukaran et al.
FIGURE 8
3-D surface plots demarcating the attainment of maximum temperature with combination of
laser process parameters for CP Ti when S is (a) 1.00 mm, (b) 1.50 mm and (c) 2.00 mm.
15
the melting point of CP Ti; hence, if one wishes to perform laser cutting or
drilling, one should choose a parameter combination which lies above this
surface. On the other hand, for jobs, such as laser bending, heat treatment,
shock peening, etc., one should choose a laser process parameter combination that would lie below the surface, such that the maximum temperature
does not exceed the melting point of the working substance.
4.2Estimation of scanning speed needed to achieve a specific
temperature
In this section, we attempt to develop a mathematical relation to find the scanning speed such that the maximum temperature reaches the specified temperature, VTmax (P,D,S) corresponding to known values of other process
parameters (P,D,S). The mathematical relation for finding VTmax (P,D,S) is
derived from the regression Equation (6) which is found out from statistical
analysis:
VTmax =
-bv - bv 2 - 4av cv
2av
(8)
where
av=2.79748 (8a)
bv=-96.68044-0.046250P+0.35000D+18.45000S (8b)
and
cv=-(Tmax-2486.67930-4.76588P+1030.19362D+797.64368S
+0.66750PD+0.24750PS-2.5000DS+8.25411 10-4P2
-216.34959D2-183.74483S2(8c)
The predictions of minimum scanning speed as obtained from Equation (8)
are compared with FEM simulation for a few number of process parameter
combinations. Results show that the FEM simulation and prediction
results from Equation (8) are in good agreement. Comparisons are furnished
in Table 6 and in Figure 9.
Figures 10(a) to (c) show the 3-D surface plots demarcating the attainment
of maximum temperature with varying combinations of laser process parameters for 1.00, 1.50 and 2.00mm thick plates. It is observed from Figures
10(a) to (c) that higher scanning speed is required for the material to reach the
specified maximum temperature when the laser power is increased and the
laser spot diameter is kept constant. Similarly, when the laser spot diameter is
16
P. Kalvettukaran et al.
TABLE 6
Comparison of maximum temperature.
Percentage Deviation
Sample
Laser
Power,
P (W)
Scanning
Speed,
S (mm/s)
Spot
Diameter,
D (mm)
Plate
Thickness,
S (mm)
Equaqtion
(7)
(K) (fa)
FEM
(K)
(fb)
(fa fb )
100
(fb )
650
10.3240
1.50
1.00
1941
1930
0.57000
550
10.1913
1.09
1.50
1941
1979
-1.92000
650
5.5731
1.50
2.00
1941
1930
0.59600
550
5.2597
2.00
0.59
1941
1985
-2.21700
450
3.2743
1.50
1.00
1941
1917
1.26300
732
4.1087
2.00
1.50
1941
1937
0.23200
FIGURE 9
Graph showing the comparison of the developed model from Equation (8) and the FEM model
for maximum temperature.
17
FIGURE 10
3-D surface plots demarcating the attainment of maximum temperature with combination of
laser process parameters for CP Ti when S is (a) 1.00mm, (b) 1.50mm and (c) 2.00mm.
18
P. Kalvettukaran et al.
DTmax =
-bd - bd 2 - 4 ad cd
2 ad
(9)
where
ad=216.34959 (9a)
bd=-1030.19362-0.66750P+0.35000V+2.50000S (9b)
and
cd=-(Tmax-2486.67930-4.76588P+96.68044V+797.64368S
+0.046250PD+0.24750PS-18.45000VS+8.25411 10-4P2
-2.79748V2-183.74483S2(9c)
TABLE 7
Comparison of maximum temperature.
Percentage Deviation
Sample
Laser
Power,
P (W)
Scanning
Speed,
S (mm/s)
Spot
Diameter,
D (mm)
Plate
Thickness,
S (mm)
Equaqtion
(9)
(K) (fa)
FEM
(K)
(fb)
(fa fb )
100
(fb )
650
15.0
1.2971
1.00
1941
1981
-2.01918
550
10.0
1.0978
1.50
1941
1972
-1.57201
650
5.0
1.5429
2.00
1941
1925
0.831169
550
10.0
1.4946
0.59
1941
1942
-0.06693
450
15.0
0.8225
1.00
1941
1958
-0.86823
732
10.0
1.5091
1.50
1941
1971
-1.52207
19
FIGURE 11
Graph showing the comparison of the developed model from Equation (9) and the FEM model
for maximum temperature.
The predictions of minimum spot diameter as obtained from Equation (9) are
compared with FEM simulation for a few number of process parameters.
Comparison results are furnished in Table 7 and in Figure 11.
Figures 12(a) to (c) show the 3-D surface plots demarcating the attainment
of maximum temperature with varying combination of laser process parameters for 1.00, 1.50 and 2.00mm thick plates. We can realize from the 3-D
surface plots that a larger spot diameter is required for the material to reach
the specified maximum temperature when the laser power is increased and
the scanning speed is kept constant. Likewise, when laser scanning speed is
increased at a fixed laser power, spot diameter is required to be reduced to
attain the same maximum temperature.
5CONCLUSIONS
In the present study, laser processing on commercially pure Ti sheet metal is
simulated and analysed, using finite element method (FEM) and response
surface methodology (RSM). In addition, mathematical models are devel-
20
P. Kalvettukaran et al.
FIGURE 12
3-D surface plots demarcating the attainment of maximum temperature with combination of
laser process parameters for CP Ti when S is (a) 1.00mm, (b) 1.50mm and (c) 2.00mm.
21
oped for estimating the input process parameters such that the material temperature attains a specified maximum temperature and predicting the
maximum temperature attained when process parameters are known. It is
found that the percentage error between predicted results of developed mathematical models and numerical simulation are less than 2.5%, which gives an
assurance to use the developed model for laser processing of Ti with satisfactory precision for diverse operations, such as cutting, drilling, bending, shock
peening etc.
Acknowledgments
K. Paramasivan expresses gratitude for the grants received for his Senior
Research Fellowship under the State Government Fellowship of Jadavpur
University.
NOMENCLATURE
A
c
f
h
k
P
PTmax (V,D,S)
Q
r
R
r
t
T0
Ts
T(r,t)
y
(x1,x2,x3,....xn)
Absorption coefficient
Specific heat (J/kgC)
Function of response
Heat transfer coefficient 10 W/m2K
Thermal conductivity (W/mC),
Laser power (W),
Specified maximum temperature
Moving heat flux (W/m2)
Distance (m) of a point from the centre of the laser beam
Laser beam radius (m)
Coordinate (m) in the reference configuration
Time (seconds)
Ambient temperature (K)
Sheet metal surface temperature (K)
Temperature (K)
Response
Independent parameters
Greek symbols
e
r
z
Emissivity
Material density (kg/m3)
Stefan Boltzmann constant (5.6703 108 W/m2K4).
Experimental error
22
P. Kalvettukaran et al.
Mathematical operators
r
Gradient operator
References
[1] Sparks M. Theory of laser heating of solids: Metals. Journal of Applied Physics 47(3)
(1976), 837849.
[2] Majumdar D. and Manna I. Laser processing of materials. Sadhana 28(34) (2003),
495562.
[3] Shen H. and Vollertsen F. Modelling of laser forming - A review. Computational Materials
Science 46(4) (2009), 834840.
[4] Ji Z. and Wu S. FEM simulation of the temperature field during the laser forming of sheet
metal. Journal of Materials Processing Technology 74(13) (1998), 8995.
[5] Venkadeshwaran K., Das S. and Misra D. Finite element simulation of 3-D laser forming
by discrete section circle line heating. International Journal of Engineering, Science and
Technology 2(4) (2010), 163175.
[6] Jamil M.S., Sheikh M.A. and Li L. A study of the effect of laser beam geometries on laser
bending of sheet metal by buckling mechanism. Optics & Laser Technology 43(1) (2011),
183193.
[7] Li L.Q., Chen Y.B., Wang X.Y. and Lin S.Y. FEM simulation for laser forming processing.
Acta Metallurgica Sinica (English Letters) 17(3) (2004), 317322.
[8] Shukla P. and Lawrence J. Examination of temperature distribution and the thermal effects
on Si3N4 engineering ceramics during fibre laser surface treatment. Optics and Lasers in
Engineering 49(7) (2011), 9981011.
[9] Scintilla LD. and Tricarico L. Estimating cutting front temperature difference in disk and
CO2 laser beam fusion cutting. Optics & Laser Technology 44(5) (2012), 14681479.
[10] Acherjee B., Kuar A., Mitra S. and Misra D. Finite element simulation of laser transmission welding of dissimilar materials between polyvinylidene fluoride and titanium. International Journal of Engineering, Science and Technology 2(4) (2010), 176186.
[11] Acherjee B., Kuar A., Mitra S. and Misra D. Modeling of laser transmission contour welding process using FEA and DoE. Optics & Laser Technology 44(5) (2012), 12811289.
[12] Zahrani G. and Marasi A. Experimental investigation of edge effect and longitudinal
distortion in laser bending process. Optics & Laser Technology 45 (2013), 301307.
[13] Gollo H., Naeini M., Liaghat G.H., Jelvani S. and Torkamany M.J. A numerical and experimental study of sheet metal bending by pulsed Nd:YAG laser with DOE method.
Advanced Materials Research 8386 (2010), 10761083.
[14] Gollo H., Mahdavian S.M. and Naeini M. Statistical analysis of parameter effects on bending angle in laser forming process by pulsed Nd:YAG laser. Optics & Laser Technology 43
(2011), 475482.
[15] Venkadeshwaran K., Das S. and Misra D. Bend angle prediction and parameter optimisation for laser bending of stainless steel using FEM and RSM. International Journal of
Mechatronics and Manufacturing Systems 5(34) (2012), 308321.
[16] Acherjee B., Kuar A., Mitra S., Misra D. and Acharyya S. Experimental investigation on
laser transmission welding of PMMA to ABS via response surface modeling. Optics &
Laser Technology 44(5) (2012), 13721383.
[17] Olabi A.G., Benyounis K.Y. and Hashmi M.S.J. Application of response surface methodology in describing the residual stress distribution in CO2 laser welding of AISI304. Strain
43(1) (2007), 3746.
[18] Reisgen U., Schleser M., Mokrov O. and Ahmedn E. Statistical modeling of laser welding
of DP/TRIP steel sheets. Optics & Laser Technology 44(1) (2012), 92101.
23
[19] Caydas U. and Hascalk A. Use of the grey relational analysis to determine optimum laser
cutting parameters with multi-performance characteristics. Optics & Laser Technology
40(7) (2008), 987994.
[20] Yang J., Sun S., Brandt M. and Yan W. Experimental investigation and 3D finite element
prediction of the heat affected zone during laser assisted machining of Ti6Al4V alloy.
Journal of Materials Processing Technology 210(15) (2010), 22152222.
[21] Adamus K., Kucharczyk Z., Wojsyk K. and Kudla K. Numerical analysis of electron beam
welding of different grade titanium sheets. Computational Materials Science 77 (2013),
286294.
[22] Ellis L.D. Effects of long-term thermal exposure on commercially pure titanium grade 2
elevated-temperature tensile properties. NASA/TM-2012215484, E-16688, National
Aeronautics and Space Administration (NASA). 2012.
[23] Jung H. A Study on Laser Forming Processes with Finite Element Analysis. PhD Thesis,
University of Canterbury Christchurch. 2006.
[24] Montgomery D.C. Design and Analysis of Experiments. New York: Wiley. 2001.