Power scaling of kW-diode lasers optimized for material processing

applications
Stephan Schneider*, Ihab Kardosh, Michael Voß, Sebastian Liebl, Jens Meinschien
LIMO Lissotschenko Mikrooptik GmbH, Bookenburgweg 4-8, 44319 Dortmund, Germany

ABSTRACT

Cutting of metal sheets is a key application in material processing and high power diode lasers gain further importance in
this field due to their exceptionally good current to light conversion efficiency. Both, power scaling and process
optimization are under investigation to improve the performance of this application. We report first results of a laser
system combining these approaches. The presented diode laser power scaling is realized by means of an asymmetric,
noncircular beam shape. The beam parameter product of the laser light is manipulated accordingly. In addition, the
power scaling of this approach allows the generation of spot geometries which inherently support the interaction
processes of the laser light with materials as metals. The setup is based on conventional, highly reliable and well tested
principles and components, as they are passively cooled laser diode bars, coarse wavelengths coupling and fiber delivery.
First results of material processing with stainless steel of 6 mm thickness are presented.
*[email protected]; phone +49 231 222410; fax +49 231 22241140; www.limo.de
Keywords: high power diode laser, beam shaping, laser processing, scalability

1. INTRODUCTION
High power diode lasers are usually called direct lasers. This term means that the electrical current is directly converted
into laser light without any optical conversion like fiber pumping. The beam parameter product (BPP) of diode lasers for
kW power applications, like cutting of metal sheets, is higher compared to fiber or CO2 lasers. However, for the
customers power and BPP are not the key parameters but reliability, process quality, wall-plug efficiency and velocity.
Efficiency is the main advantage of direct diode lasers in both aspects electrical-to-optical conversion and wavelength
dependent absorption rates in metals. Fiber lasers match with 1 µm wavelength the local absorption maximum of
aluminum, figure 1. Direct diode lasers emit a multi wavelength range typically in the region of 0.8 - 1 µm which is well
placed at the aluminum absorption peak and in the rising absorption slope of copper. The CO2 laser with 10.6 µm
wavelength addresses the lower absorption values. The overall energy conversion efficiency of CO2 lasers is in the range
of 5-10%, of fiber lasers 20-40% and of direct diode lasers 30-50% [1-4].
Many efforts were done to understand and improve the physical process of laser material processing like laser cutting,
e.g. the strategy of oscillating the laser beam in cutting direction over the region of the cutting front [5]. On the same
bases, by manipulating the laser beam shape in cutting direction, we designed a cutting head with a double fiber
connection forming an asymmetric laser beam. The asymmetry originates from the crossing of two laser beams in a focal
plane. In this paper the opto-mechanical design of this cutting head is presented, together with first results of stainless
steel cutting. A theoretical model is described and its reliability is tested on the basis of the experimental results. The
predictions of the model for an enlarged set of parameters are promising for further development of high power direct
diode lasers for metal cutting applications.

High-Power Diode Laser Technology and Applications XIII, edited by Mark S. Zediker,
Proc. of SPIE Vol. 9348, 934808 · © 2015 SPIE · CCC code: 0277-786X/15/$18
doi: 10.1117/12.2076265

Proc. of SPIE Vol. 9348 934808-1

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/14/2015 Terms of Use: http://spiedl.org/terms

 0.03

Cu
0.25

0.20

__ Steel
4 0.15
Ag
ó

0.10 Al
/ 1

..--- ....

0.05 \ \
.
o
0.1 0.2 0.3 0.5 0.8 1 2 4 5
rsr...
8 10 20
Wavelength [µm)

Figure 1: Spectral dependent absorption rates of metals [6].

2. CUTTING HEAD WITH ASYMMETRIC BEAM

For job shops the cutting velocity per power, i.e. per cost, is an important parameter. In the last three years LIMO lasers
increased both, power and velocity and furthermore also the velocity to power ratio [7,8]. Achieved values for samples
of 6 mm stainless steel are summarized in table 1, the data of the year 2014 are realized by the asymmetric laser beam.

Table 1: Evolution of process parameters, the data of 2014 are achieved with the asymmetric laser beam.

Year Optical power P [kW] Cutting velocity V [m/min] V/P [m/(min kW)]
2012 2.0 0.7 0.35
2013 2.5 1.1 0.44
2014 4.0 2.0 0.50

The asymmetric laser beam is generated by a cutting head consisting of two parts, a combination unit and a focusing unit
as shown in figure 2. Two optical fibers are attached to the combination unit. The light emitted from the fibers is
collimated and set to be close to each other. This pair of laser beams passes the same focusing lens. The cutting head is
designed for a total optical power of 8 kW, i.e. 4 kW per optical fiber, and fusion cutting of metals with Nitrogen as
process gas. The Nitrogen connectors are placed close to the nozzle oriented under 45 deg. The focusing lens is
exchangeable allowing for different spot diameters. For further test scenarios an additional telescope optic can be added
to the optical path. The parameters of the cutting head are summarized in figure 2 and table 2. A third component of this
cutting head, a rotation unit, which will be utilized for spot orientation adjustment, is under construction.

Proc. of SPIE Vol. 9348 934808-2

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/14/2015 Terms of Use: http://spiedl.org/terms

 Fiber connectors Beam combination unit Focussing unit

Figure 2: Cutting head with two fibers connected to the combination unit. The combined pair of laser beams passes the same
optical elements in the focusing unit, the units are in mm.

Table 2: Parameters of the cutting head with asymmetric laser beam.

Parameter Value
Fiber core diameter [µm] 400
Fiber numerical aperture, effective 0.12
Average beam parameter product [mm mrad] 55
Spot diameter range (FW 1/e²) [µm] 480 -1000
Working distance range [mm] 0.3 - 1

In this work we used two optical setups generating a focal spot diameter, just called spot diameter in following, of 620
µm and 1005 µm, while the beam parameter product is unchanged. The presented model below explains the choice of
these diameters. A measurement of the asymmetric beam caustic for the 620 µm spot diameter is shown in figure 3.The
laser beams are directed under an angle of 23.5 degrees to a focus spot. It has a slightly elliptically shape but can still be
treated as round according to DIN EN ISO 11146-1, since its ellipticity is above 0.87. The beam parameter product of
each single fiber output is 24 mm mrad, which is unchanged in the x-axis due to the cutting head optics. The measured
data are summarized in table 3. A range of 6 mm, which corresponds to the thickness of the used metal sheet, is also
marked in figure 3, together with the z direction of the laser beam propagation.

(1111poóöo00000 z
Figure 3: Measured beam caustic in grey scale representation. The two beams propagate in z direction. The focus position
and a 6 mm range are indicated.

Proc. of SPIE Vol. 9348 934808-3

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/14/2015 Terms of Use: http://spiedl.org/terms

 Table 3: Axes dependent beam parameter product (BPP) of the asymmetric laser beam for two optical setups.

Optical setup 1 Optical setup 2

Axes BPP [mm mrad] Diameter [µm] BPP [mm mrad] Diameter [µm]
X 24 580 24 1010
Y 85 660 78 1000
Mean 55 620 51 1005

As discussed in the following, the asymmetry of the laser beam has an influence of the cutting direction. For further
investigations of this effect both parts, combination unit and focusing unit, are separable and an additional rotation unit
can be added in between. This motorized unit rotates the laser beam around z-direction continuously.
Extended tests concerning industrial robustness, e.g. effects of acceleration forces on the mechanical stability were done.
The x- and y-axes of the laser machine were programmed to follow square patters with 10 mm and 60 mm side length
and circles with diameters of 5 mm and 60 mm. The maximum linear acceleration was 8.5±1 m/s² and the maximum
radial acceleration 11.0±1 m/s². These tests proofed the structural stability of the cutting head.

3. RESULTS
In this section the cutting results are presented, using stainless steel sheets (type 1.4301) of 6 mm thickness as sample
material. Since the laser beam is asymmetric the orientation of the beam with respect to the cutting direction has to be
taken into account. If not specially denoted, the tests were done with an orientation of 0 degree, meaning that the axis of
the two laser beams points in cutting direction as shown schematically in figure 5.The effect of the angle on the cutting
results is discussed afterwards. As process gas Nitrogen at 18 bar pressure was applied.
Both optical setups were used. The difference in the laser power originates from the availability of laser sources at the
time of the application tests. The quality of the cutting edges are evaluated using the roughness value Rz defined in DIN
EN ISO 9013, which is the arithmetic mean of the height variation of five single profile elements along the cutting
direction. The Rz values are measured at three height positions of the cutting edge samples as marked in the images.
While laser power and cutting velocity are comparable, the edge surface quality is better using optical setup 2, the 1005
µm spot diameter, where the Rayleigh length is a factor of three longer compared to setup 1.

Table 4: Cutting result for two sets of parameters.

Parameter Optical setup 1 Optical setup 2

Optical power [kW] 4.0 3.9
Spot diameter [µm] 620 1005
Velocity [m/min] 2.0 1.8
Rz top / center / bottom [µm] 60 / 13 / 81 30 / 17 / 32

Image of cutting edge -_ : rwwwpíomooksti'xisomiNif*?tWkií

Proc. of SPIE Vol. 9348 934808-4

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/14/2015 Terms of Use: http://spiedl.org/terms

 The next parameter under investigation was the angle dependence of cutting results with respect to the orientation of the
asymmetric beam. In table 5 the roughness values Rz for different angles are summarized. No significant trend of the
edge surface roughness is observable in the angle range from zero to 35 degrees. Even at 90 degree the surface is still
comparable but at angles bigger than 35 degrees dross formation at the bottom side takes place. When decreasing the
velocity this dross formation is reduced. The test pattern shown in figure 4 b can be processed by reducing the velocity to
80% when reaching the critical angel.

90° metal sheet

asym. beam

15°

Figure 4. a) Schematically top view of a metal sample, to indicate the orientation of the asymmetric laser beam versus
cutting direction. The laser beam keeps its orientation even when the cutting direction changes. b) Cutting pattern with
R = 20 mm and α = 110 deg.

Table 5: Quality measurements of the cutting edges.

Rz [µm]
Angle [deg.] top center bottom mean
0 26 16 95 46
5 59 18 76 51
15 82 27 64 58
25 31 13 91 45
35 25 22 57 35
90 58 14 53 42

4. PROCESS MODELLING
In this work the process of fusion cutting was used, other processes like oxygen cutting are described elsewhere. In a
fusion cutting process a cut front is formed under an inclination angle ϕ , which is shown schematically in figure 5. The
angle is determined in a simplified two-dimensional scheme using the caustic of the laser beam and the corresponding
beam radii at the top rt and bottom rb of the metal sheet,

⎛ d ⎞
ϕ =arctan⎜⎜ ⎟⎟ . (1)
⎝ rt + rb ⎠

Proc. of SPIE Vol. 9348 934808-5

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/14/2015 Terms of Use: http://spiedl.org/terms

 laser beam

cutting direction
Ir I

;HI
I

ro

Figure 5: Schematic side view of laser cutting process.

A part of the total laser power Plaser is reflected at the surface of the molten metal film due to Fresnel reflection [5,9]. The
other part Pheat is absorbed by the metal sheet, so that

Pheat = A ⋅ Plaser , (2)

with

A =1 − 0.5[RS (n, k , ϕ ) + RP (n, k , ϕ )] . (3)

The reflection coefficients Rs and Rp depend on the inclination angle ϕ , the refractive index n and extinction coefficient
k of the molten metal film. Furthermore, the absorbed power Pheat does only partly contribute to the melting process
power Pmelt, while the other part Ploss heats up the area surrounding the cut front [10],

⎡ Pe ⎛ Pe ⎞γ ⎤
Pheat = Pmelt + Ploss = 4ηd ⋅ (Tmelt − Tamb ) ⋅ ⎢ + ⎜ ⎟ ⎥ , (4)
⎣⎢ 2 ⎝ 2 ⎠ ⎦⎥
v⋅r r0 + rt + rb
and Pe = , r= . (5)
κ 3

Pe is called the Peclet number. As beam radius r, the mean value of the radii r0 at the focal plane, rt at the top and rb at
the bottom surface of the metal sheet are used. These formulas give a relation between the optical power at the output of
the cutting head, the beam parameter product of the laser beam and the cutting velocity.
The following analysis uses the material parameters accordingly to references [5,10], as they are listed in table 6.

Table 6: Model parameters.

Parameter Comment Symbol Value

Thermal conductivity [W/(m K)] Steel 1.4301 η 16
Thermal diffusivity [m²/s] Steel 1.4301 κ 4.00E-06
Exponent of power law Range [0.3 - 0.36] γ 0.30
Refraction index, iron Iron @ 1800 K and 1.07µm n 5.46

Proc. of SPIE Vol. 9348 934808-6

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/14/2015 Terms of Use: http://spiedl.org/terms

 Extension coefficient Iron @ 1800 K and1.07µm k 3.96
Melting temperature [K] Steel 1.4301 Tmelt 1694
Ambient temperature [K] Normal condition Tamb 298

In figure 6 the experimental results are shown together with the predicted data from the model. The model data are
plotted as errors bars, which are estimated by varying single parameters and measuring their influence on the velocity.
The experimental results are in good agreement with the model data, within these error bars. This indicates that the
assumptions for the model are valid, as they are: a) Material parameters taken for a single wavelength of 1.07 µm instead
of weighted values for the diode laser wavelength mixture between 0.8 and 1.7 µm. b) Use of a single BPP of 55 mm
mrad as mean value for the asymmetric laser beam. c) Use of an effective radius r. d) Use of refractive index and
extension coefficients of molten iron instead of stainless steel type 1.4301. However, in this work the model is not
thought to explain the physical process of laser cutting but to give hints to the best set of parameters which then have to
be proofed in applications. The data points of figure 6 are listed additionally in table 7. While the beam parameter
product is constant, the spot diameter has two values.
The linear dependence between the optical power and the cutting velocity has a gradient of 0.65 m/(min kW) and is
independent on the focus diameter.

2.2
620 p.m
2.0
620 µm -
1.8
C ,,-11 1005 p.m
É 1.6
\L`111-,

v> 1.2
C taa\eob
1.0
u
0.8
1005 µm
0.6 -
----1005 p.m
0.4
15 2 25 3 35 4 45
Laser power [kW]

Figure 6: Comparison of experimental and theoretical data.

Table 7: Data sets shown in figure 6

Laser power[kW] BPP [mm mrad] Spot diameter [µm] Cutting velocity [m/min]
1.8 55 1005 0.6
2.3 55 1005 0.8
3.7 55 620 1.8
3.9 55 1005 1.8
4.0 55 620 2.0

Proc. of SPIE Vol. 9348 934808-7

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/14/2015 Terms of Use: http://spiedl.org/terms

 The comparable large spot diameters in focal plane of 620 µm or even 1005 µm are unusual in cutting applications.
Taking the model, the expected cutting velocity in arbitrary units versus focal diameter for a given metal sheet thickness
of 6 mm and beam parameter product of 55 mm mrad is plotted in figure 8. The maximum velocity is reached with a spot
diameter of 750 µm. When planning the experiment this spot diameter was the aim but unfortunately the measurements
have to be postponed. With the available optical components the presented spot diameters of 620 µm and 1005 µm were
realized. Both optical setups 1 and 2 reach with comparable optical power of 3.7 kW and 3.9 kW the same velocity of 1.8
m/min. This result is in good agreement with the prediction in figure 7, where a comparable velocity for both spot
diameters of 96% is estimated. The next application step is to adapt the cutting head optics to generate a spot diameter of
750 µm, which is easily achieved by an exchange of the focusing lens.

1.1
1

1 I
1.0

.F 0.9 1

1
E
1
E 0.8
-o 1
w
á 1

m 0.7
c

c-)0.6
I

I I
0.5

0.4
02 04 06 08 1 12 14 16
Spot diameter [mm]

Figure 7: Dependence of cutting velocity on the focal diameter for constant BPP and laser power according to (4). The
two corresponding spot diameters of the applications are marked by the dotted lines.

5. SUMMARY
A new laser cutting head, generating an asymmetric laser beam, is presented. The design allows power scaling up to 8
kW by parallel mounting of two optical fibers and therefore the usage of two laser sources. The application tests of 6 mm
stainless steel processing were done with a cutting velocity of 2.0 meter per minute at an optical power of 4 kW. A
theoretical model is described and its reliability is proofed on the basis of the experimental results. The predictions of the
model give hints for further development of high power direct diode lasers for metal cutting applications. The
comparable large focal spot diameters of the asymmetric laser beams are close to the calculated optimum. This paper is
focused on the interaction of the laser beam with the material. Further important parameters, like pressure and flow rate
of the process gas as well as the cone geometry of the cutting nozzle are on the roadmap for future investigations.

6. ACKNOWLEDGEMENT
This work is funded by the German Federal Ministry of Education and Research (BMBF) with the program “Brilamet”
[FKZ: 13N12282]. It is one of several projects, which are patronized by the BMBF to facilitate the research in Germany
on high brightness lasers for material processing applications.

Proc. of SPIE Vol. 9348 934808-8

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/14/2015 Terms of Use: http://spiedl.org/terms

 REFERENCES

[1] Crump, P., Erbert, G., Wenzel, H., Frevert, C., Schultz, C. M., Hasler, K.-H., Staske, R., Sumpf, B., Maaßdorf,
A., Bugge, F., Knigge, S. and Tränkle, G., “Efficient High-Power Laser Diodes,“ IEEE J. Sel. Top. Quantum.
Electron., Vol. 19, No. 4, (2013).
[2] Havrilla, D., Koitzsch, M., Holzer, M. and Ryba, T., “Latest Advances in High Brightness Direct Diode Lasers
and their Applications,“ Proc. ICALEO, 303 (2014).
[3] Pocorni, J.K., Petring, D., Powell, J., Deichsel, E. and Kaplan, A.F.H., “Differences in Cutting Efficiency
between CO2 and Fiber Lasers when cutting mild and stainless steels,“ Proc. ICALEO, 905 (2014).
[4] Davim, J.P., [Nontraditional Machining Processes], Springer, 39, (2013)
[5] Mahrle, A. and Beyer, E., „Theoretical aspects of fibre laser cutting, “J. Phys. D: Appl. Phys. 42, 175507
(2009).
[6] Fahrenwaldt, H. J. and Schuler, V., [Praxiswissen Schweißtechnik], Ed. 3, Vieweg -Teubner, 75, (2009).
[7] Lentjes, M., Bruns, P., Voß, M., Reinecke, K. and Lissotschenko, V., “Process-Reliable & Cost-Effective Laser
Materials Processing,” Proc. ICALEO, 807 (2012).
[8] Wahab, H., Gröninger, J., Dickmann, K., Bruns, P., Voß, M., Kardosh, I., Meinschien and J., Aschke, L., “High
Brightness Diode Lasers as Direct Cutting Tools,” Proc. ICALEO, 903, (2014).
[9] Ready, J. F. and Farson, D. F., [LIA Handbook of Laser Materials Processing], Laser Institute of America, 175-
181 (2001).
[10] Schulz, W., Becker, D., Franke, J., Kemmerling, R. and Herziger, G., “Heat conduction losses in laser cutting of
metals,“ J. Phys. D: Appl. Phys. 26, 1357-1363 (1993).

Proc. of SPIE Vol. 9348 934808-9

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/14/2015 Terms of Use: http://spiedl.org/terms