Click to go back

Harris and Brandt

Materials Forum (2001) 25, 88-115

Laser Cutting of Thick Steel Plate

J. Harris and IVI. Brandt

Cooperative Research Centre for Intelligent Manufacturing

Systems and Technologies

Industrial Laser Applications Laboratory

IRIS, Swinburne University of Technology
PO Box 218 Hawthorn, Melbourne
VIC 3122 Australia

ABs-rRACT

Of the many machining processes used in manufacturing to cut metals

and non-metals, laser cutting offers unique advantages in terms of cut
quality, speed, absence of tool wear and minimal or no clamping of parts.
In industry, the lasers used for cutting are CO 2 (carbon dioxide) and
Nd:YAG (Neodymium-doped Yttrium Aluminium Garnet) lasers each with
their own characteristic output properties such as wavelength, power)
mode of operation and beam quality. These properties, together with the
optical and thermophysical properties of workpiece material and
workpiece handling system, control and determine the cutting
performance of any material. Presented in this paper are laser, materials
and system properties, and parameters influencing the cutting of metals
in particular. The operation of the CO 2 and Nd:YAG lasers and their
dominant features is also discussed. The cutting mechanism is described
in terms of the energy balance within the workpiece. This is then used to
show the difficulties with cutting thick (>10mm) steel plate with a laser,
leading into a number of novel laser methods explored for the cutting of
thick plates. Finally, recent work supported by the CRC for Intelligent
Manufacturing Systems and Technologies (CRCIMST) on laser cutting
with a "spinning laser beam" is presented and its potential for cutting thick
steel plate discussed.

© Institute of Materials Engineering Australasia, Ltd (2001)

88
 Harris and Brandt

1. INTRODUCTION through almost any material by erosion. The

kerf is rather large ('"'"' 1 mm) and the cutting
Many methods are available to cut materials speeds are modest compared to the thermal
and metals in particular. Among these are cutting methods.
the high energy processes of laser, electron
beam, plasma, water jet and oxy-fuel. Oxy- The use of the laser as a tool for materials
fuel or flame cutting, is still largely used for processing began as early as 1963 and is
cutting very thick workpieces of metals, still developing. The first experiments in
typically 100 mm to 500 mm thick (Rickfalt laser cutting with a gas assist nozzle were
1995). The edges of the cut are oxidised made in May 1967 (Hilton 1997) at the
and the kerf produced is large. The heating Welding Institute in Cambridge, UK, using
fuel is usually propane, natural gas, a a 300W CO 2 laser and oxygen assist gas. A
hydrocarbon mixture or acetylene, and is 2.5 mm thick plate was cut with remarkable
delivered through a concentric nozzle with accuracy. The idea of using the laser as a
the oxygen nozzle, so that the oxygen jet is cutting tool for metals belongs to Peter
surrounded by the fuel gas. Houldcroft who was then the deputy
scientific director at The Welding Institute.
The development of plasma tools began in He thought of it when confronted with
1941 in the US aircraft industry for welding cutting trials using a plasma torch for body
applications. This process uses an electric panel trimming at the British Motor
arc between an electrode and the metal, and Company in 1965. The plasma cutting was
a gas shielding around the arc, blown not accurate enough and produced burning
through the electrode. An innovation by of the edges.
scientists at Union Carbide's welding
laboratory in the early 1950s reduced the With the first industrial laser cutting system
gas nozzle opening. The result was that the supplied in 1970, the progress of the new
electric arc and the gas were constricted to a technology has been fast. In the last three
small volume in space and the gas velocity years some 18,000 lasers for metal sheet
increased dramatically, together with the arc cutting have been installed world-wide and
temperature and voltage. A cutting tool was this number is increasing yearly by some
thus obtained and a new technology 6000 units (Belfore 2001). The value of
developed that has been continuously these laser cutting systems in 2001 is
improved over the years. Plasma jets can cut estimated at about US$1.0 billion. This high
through thick materials, but not as thick as level of adoption of lasers for cutting is
those cut with the oxygen flame. They associated with a number of advantages the
possess very high temperatures and thus can technology offers compared with the
be used to cut materials with very high competitive cutting technologies described
melting points that cannot be cut with other above (Steen 1998). These include:
techniques.
• The non-contact nature of the machining
Unlike the two cutting technologies
process requiring light or no clamping of
mentioned above, abrasive water jet cutting
the workpiece. This absence of cutting
does not melt the material. The operating
forces also allows precise cutting of light
principle is the acceleration of abrasive
or flimsy materials.
particles by mixing them in a stream of
water. Passing through a nozzle, a highly • The absence of cutting tool wear and
focused jet of water is obtained that can cut tool chatter as a result of no contact.

89
 Harris and Brandt

• Machining operations are not governed Given the long time that commercial laser
by workpiece hardness, with lasers able cutting technology has been available to
to cut many hard materials. industry and the significant advancements
achieved in that time, laser cutting is now
• Minimal noise pollution as the laser
considered a mature technology.
cutting process is quieter than the
Consequently, cutting technology and
competitive processes of water jet and
knowledge of the cutting mechanism for
plasma.
thin and medium thickness «10mm) steels
• The narrow kerf width, typically 0.1 to have remained basically unchanged during
1.0 mm, requires little or no radial the last ten years and are considered well
compensation for beam diameter. understood. There is, however, significant
Furthermore, the possibility of nesting of demand by industry to increase cut
components in production allows thickness and the quality of cut at those
minimal wastage of raw material. thicknesses. This paper discusses issues
• The clean nature of the cut allows, in associated with the laser cutting of thick
most cases, no subsequent cut face (> 1Omm) steel plate and methods currently
processing. Molten or oxidised material being examined to cut it. Presented in
is removed by assist gas during the Section 2 are general aspects of laser cutting
cutting process or falls away during the and in Section 3 a brief description is given
separation of the part at the completion of the laser systems used for cutting.
of the cut. Section 4 summarises the chemistry of laser
cutting and Sections 5 and 6 address the
• Given that the laser cutting mechanism issues of, and novel laser technologies for
is fast and the area heated is small, there the cutting of thick plate.
is a minimal heat affected zone and
consequently minimal heat distortion.
2. GENERAL ASPECTS OF LASER
CUTTING
The disadvantages of laser cutting compared
to other cutting technologies are: The process of laser cutting may be
considered as a sequence of the following
mechanisms:
• The process effectiveness reduces as the
workpiece thickness increases. • absorption of laser radiation by material;
Commercially, workpiece thicknesses
greater than about 15 mm are generally • heating and melting of the material, and
not cut with a laser. • removal of the molten material by the
coaxial assist gas jet. If the assist gas is
• Laser cutting produces a tapered kerf oxygen the exothetmic chemical
shape. This is a result of the divergence interaction with the molten metal will
of the laser beam and is more accelerate the cutting process.
pronounced in thicker materials. The
taper can be reduced by positioning the
The key features of laser cutting are
focus below the surface of the
illustrated in Figure 1. In laser cutting, the
\vorkpiece.
workpiece is placed on a CNC table and
• The heat affected zone can be a problem moved at a predetermined speed and path
in some applications because of the designed on a CAD system and converted to
changed microstructure. machine code instructions. The laser beam

90
 Harris and Brandt

is typically focused to a diameter of 0.1 to • means of producing movement of

0.3 mm on the workpiece surface (or just workpiece relative to cutting nozzle,
below it) with a lens or a lens system held in generally CNC controlled.
a specially designed nozzle. The most
popular focal length range is 100mm to
(a)
150mm giving a good balance between spot
size and depth of field. Incident laser power
Cutting direction
densities are of the order of 106 W/cm2.
Once the laser beam has penetrated through
the material, an erosion front is established
in the cutting direction. The erosion front is .....--_ Assist gas
covered by a layer of molten material which
is continually removed by the assist gas at
the lo\ver surface of the cut and created by
the absorption of energy from the laser
Cutting nozzle
beam (Schuocker 1993). The assist gas, gas Narrow kerf

pressure, nozzle shape and position above

workpiece are all critical process parameters
and affect strongly the removal of material
from the kerf (Powell 1993). Nozzle
diameters vary from 0.8 mm to 3 mm
depending on the material being cut and the
nozzle position above the workpiece, which
influences gas pressure in the kerf. This
position is carefully chosen because of the
strong variation of the melt removal rate (b)
with the complex pressure patterns
produced in the kerf due to the supersonic
speed of the assist gas (Fieret 1987; Na
1989).

3. LASER SYSTEMS USED FOR

COMMERCIAL CUTTING

A typical laser system used for cutting

consists of:

• the laser beam source along with its

utilities and control unit,
• a beam guidance arrangement from the
laser beam source to the focusing optics,
• the focusing optics with cutting nozzle
Figure 1. Illustrated in (a) is laser cutting
and process gas supply system,
arrangement and key cutting features and in (b)
• the workpiece and its support plus fume CO 2 Jaser cutting a component (courtesy Rofin
extractors for vapors, and Sinar) ,

91
 Harris and Brandt

Table 1. Principal characteristics of C02 and Nd: Y AG cutting lasers.

Active Wavelength Laser power Beam delivery and Electrical efficiency

Medium focussing

CO2 10.6 Jlm ~ 4kW (pulsed or Mirrors and lenses 10-15%

molecule CW)
1.06Jlm Mirrors~optical fibres 4% (lamp pumped)
Ndion ~ 2 kW (pulsed or and lenses
CW) 10% (diode pumped)

The lasers used for cutting include CW CO2 energy to the CO2 molecule and enabling it
lasers, pulsed Nd:YAG lasers and more to remain in the upper laser leveL The
recently CW Nd: YAG lasers. Typical helium cools the gas mixture through
characteristics of these two lasers are shown collisions and transfer of stored energy from
in Table 1. The CO2 laser is by far the the CO2 molecule. The CO2 laser produces
most widely used laser for cutting because output at a wavelength of 10.6 flm (invisible
of its high power, beam quality and part of the spectrum) at power levels from
reliability. Typically, for cutting tenths of a watt to forty five kilowatts
applications the CO2 lasers range in power (Convergent) at an electrical efficiency of
up to 4.0kW. For the cutting of relatively about 10%. For conventional cutting
small, detailed and delicate parts a pulsed applications lasers with powers up to 4kW
Nd:YAG laser is normally used (van Dijk are used. Higher power lasers normally have
1993). At power levels up to 500W this type inferior beam quality which adversely
of laser yields very low thermal input into affects the cutting process.
the material and produces kerf widths of 0.1
mm to 0.2 mm compared to the 0.2 mm to The output power and beam quality of a
1.0 mm kerf widths produced with a CO2 CO2 laser are primarily determined by the
laser. Some typical cutting parameters for method of gas flow. The flow method
the two types of lasers are illustrated in determines how quickly the carbon dioxide
Figures 2 and 3. can be removed from the optical cavity so
that new ground state carbon dioxide can be
3.1 The C02 Laser introduced for excitation and stimulation.
The CO 2 lasers used for cutting can be
The CO2 laser is a gas laser, whose active classified as axial flow and no flow devices.
medium is a mixture of about 5% carbon In slow flo\v axial lasers the gas moves
dioxide, 10% nitrogen and the balance slowly in the direction of the laser beam
helium. There are several different designs through a glass tube (typically 10 to 14 mm
of CO 2 lasers depending on the method of in diameter) which is surrounded by
exciting and cooling the gas mixture in the another, co-axial water or oil-cooled tube.
resonant cavity, however, the principle of The length of the laser tube is typically 1 m
the lasing action is the same for all CO2 and it produces about 50 to 70 W of laser
lasers (Laos 1983). The active laser species power. The advantages of this type of
is the carbon dioxide molecule. The design include very good beam quality for
nitrogen acts as a catalyst, transferring focusing, high peak power when pulsed and

92
 Harris and Brandt

(a) (b)
14~--------------------------~ 7

13

12 • -·-700W 6 • -'-1700W
-.-1200W

\.,\
11
-·-1200W
-e-1700W
5

9 :" ':' . '-: • ': -:. '":' - • 3" \

\

8
'" • \,
"~~
"0
~ 7 Q)
Q) Q)
g. 6 0..
rJJ 3 --~-~
.....
Ol 0>
E 5
c:
E:;) :'-.. .
() 4
:;)
() 2 . ; '-...:
3

o~~~~~~~~~~~~~~~ O~~L-~~~~~~~-L~~

o 2 3 4 5 6 7 8 9 10 11 12 1314 1516 o 2 3 4 5 6
Material thickness (mm) Miterialthickness (nm)

Figure 2. Representative cutting speed as a function of material thickness for C02 laser cutting of (a)
mild steel with oxygen and (b) stainless steel with nitrogen assist gas.

(a) (b)
7 3.0 , . - - - - - - - - - - - - - - - - - - - - - - - - - - -.....

A - . - 500 W, pulsed, 0.6 mm fibre

6 - . - 500 W pulsed -1..- 1000 W, pulsed, 0.6 mm fibre

\\~.- 1000 W pulsed

2.5 -a- 500 W, pulsed, 0.4 rom fibre

5 ..
'2 "2 2.0

:€ "E
.s 4 1
'"0
""0 Q) 1.6
Q)
Q.) Q)
0.. a.
rJJ 3 tJi
0) 0)
c

..
c
1.0
E::J 1E
()
2
t.)
-:"'" ".--- ~

':~.
0.5
1-

o~~~--~~~~~~--~~~ 0.0
0 234567 B 9
o 1 234 5
Material thickness (mm) Material thickness (mm)

Figure 3. Representative cutting speed as a function of material thickness for Nd:YAG laser cutting of
(a) mild steel with oxygen and (b) stainless steel with nitrogen assist gas and optical fibres.

93
 Harris and Brandt

reduced system complexity and the tube it is cooled with a heat exchanger.
maintenance costs. The main disadvantage Laser tube length is considerably reduced
is the laser size and the maximum output compared to slow flow tubes resulting in a
power. Lasers employing this approach are more compact device. The advantages of
large because of the necessarily long laser fast axial flow lasers include, in addition to
tubes. The maximum power generated is the compact size, high output power and an
about 2kW from some 20 laser tubes placed output which can be electronically pulsed.
in series. This number of laser tubes can The main disadvantage is that the Roots or
present a challenge in keeping the optics centrifugal compressor needed to circulate
aligned. the gas is complex and costly to maintain.
CroQl1d potenllil
In no-flow or diffusion cooled lasers the gas
mix in the optical cavity is cooled by
conduction to the walls of the optical cavity.
Shown in Figure 5 is a schematic of a
diffusion cooled laser. The design involves
two flat copper plates placed close together
that act as both the RF electrodes and a heat
sink. Since the gas transport system is not
Guhdd
required the unit size is very compact.
These devices are now available in the
Figure 4. lllustrated is a schematic of a RF
excited fast axial·flow C02 laser (courtesy power range from 1 kW to 3.5 kW with
Rofin Sinar). very good beam quality for cutting
applications.

3.2 The Nd: YAG Laser

The Nd:YAG laser is a solid state laser,

Rear mIrror
usually in the shape of a rod, operating at
1.06 J!ffi (Koechner 1988). The active
species are neodymium ions present in
small concentrations in the YAG crystal.
Both continuous wave and pulsed laser
outputs can be obtained at an overall
efficiency in the 3 to 5% range. The laser is
Figure 5. The basic design of a diffusion used in industry because of its efficiency,
cooled C02 laser (courtesy Rofin Sinar). output power and reliability compared to
other solid state lasers. The crystal is grown
using the Czocbralski crystal growing
In fast axial flow lasers (see Figure 4), as technique (Dawes 1995) which involves
the name suggests, high powers are slowly raising a seed Nd:YAG crystal from
achieved by transporting the gas very the molten crystal constituents to extract a
quickly through the laser tube using Roots Nd:YAG boule. A single boule typically
or a centrifugal compressor thus reducing yields several laser rods. The concentration
the time it spends in the heated volume. of Nd ions in the boule is carefully
The gas travels at close to the speed of controlled and is not greater than about
sound, and before it is recirculated through 1.10/0. Increasing the Nd doping further in

94
 Harris and Brandt

order to increase the laser po\ver produces placed close together at the axis. The inside
unacceptable strain in the crystal and leads surface of the cavity is normally coated with
to a dramatic reduction in laser power. gold in order to maximise the coupling of
lamp light into the rod. Some laser
manufacturers also manufacture ceramic
cavities which produce more uniform
pumping of the rod but at the expense of
lower efficiency (some 5% lower) compared
to that of the gold-coated cavities.

For continuous operation, krypton arc lamps

are most widely used while for pulsed
operation high pressure xenon and krypton
flashlamps are used. Lamp lifetime
dominates the service requirement of
modern Nd:YAG lasers. For arc lamps the
lifetime ranges between 400 and 1000
Figure 6. Schematic of a Nd:YAG laser
(courtesy Rofin Sinar).
hours while for pulsed lasers it is about 20
to 30 million pulses depending on operating
conditions.
Laser rods are typically 6 mm in diameter
and 100 mm in length with the largest
commercial size rods being 10 mm in
diameter and 200 mm in length. As a
consequence of the crystal's small size
Nd:YAG lasers tend to be much more
compact than CO2 lasers. Illustrated in
Figure 6 are the main components of a
single-rod Nd:YAG laser.

Laser action is achieved by optically

exciting the crystal by lamps placed in close
proximity to it. The lamps have an emission
spectrum which overlaps the absorption Figure 7. Rofin Sinar 2.5 kW Nd:YAG laser
bands of the Nd: YAG crystal at 700 nm and with four pump chambers.
800 nm. In order to couple the maximum
amount of lamp light into the rod and
extract the maximum laser power from it, As only a fraction of the emitted spectrum is
the rod and the lamp are enclosed in absorbed by the laser crystal the rest of the
specially designed and manufactured emitted light is dissipated as heat in the
cavities. The two most common pump cavity and has to be removed for efficient
cavity configurations are elliptical and close laser operation. This is usually achieved by
coupled. In the case of elliptical cross- flowing deionised water around the rod and
sections the rod and the lamp are placed lamp in a closed loop cooling system. The
along the nvo foci, and in the case of close- loop is coupled to a heat exchanger for
coupled cavities the rod and the lamp are efficient heat removal.

95
 Harris and Brandt

To increase the laser power above 500 to 1999). This technology is the high power
600 W typically obtained from a single rod, laser diode and its main advantage lies in
requires an increase in the laser volume. having a very narrow spectral output
However, increasing the rod volume has compared with that of the lamp. The diode
fundamental limitations. Heat generated output is matched to the absorption bands of
within the rod causes large thermal the Nd: YAG laser rod thus increasing
gradients which lead to variations in the considerably the efficiency of the laser
refractive index, lowering beam quality, as system. Diode-pumped Nd:YAG lasers
well as large mechanical stresses, which can have much better beam quality because of
cause rod fracture. To obtain higher laser lower induced thermal stresses, are more
powers involves using multiple laser rods compact, require smaller chillers and have
(see Figure 7). The rods are ananged in much longer lifetimes compared to that of
senes and located either within the the lamp pumped systems. Rofin-Sinar and
resonator or some are placed outside the Trumpf are now offering commercial 6 kW
resonator to act as amplifiers. These diode-pumped Nd: YAG lasers with
configurations are discussed and described guaranteed 15,000 h diode operation.
in more detail by Emmelmann (1995).
There are now several systems on the 3.3 Bealn DelivelY and Focusing
market all giving in excess of 2 kW of laser
power with the highest power commercial To guide a CO2 laser beam to the
device producing 6 k W from 8 cavities workpiece, mirrors with precise mechanical
(Trumpf). guides are used to direct the beam along
lightweight, rigid, protective tubes to the
optical components near the work surface.
Different mirror arrangements are used for
2D and 3D motion. To focus a CO2 laser
beam both reflective and transmissive optics
are used. Focusing minors tend to be made
from copper as it is highly reflective at the
laser wavelength and can withstand high
energy densities without damage. Copper
mirrors are usually cooled with water to
minimize thermal distortion.

Transmission lenses can be made from

Figure 8. 2D CO 2 laser cutting system gallium arsenide, potassium chloride or zinc
manufactured by Laser Lab. selenide. Today, the most common material
is zinc selenide. Typically, reflective optics
are used with powers in excess of 4kW in
While lamps have been an integral part of order to minimise losses and imaging
the Nd:YAG laser technology to date and problems associated with the thermal
will remain so for the foreseeable future distortion of the zinc selenide lenses.
because of their relatively low cost, another
technology is now emerging for high power In the case of Nd: YAG lasers, mirrors and
laser applications both as a pumping source lenses made from borosilicate crown glass,
for Nd:YAG lasers and as a laser source in designated BK7, are used to guide the beam.
its own right (Bachmann 1998; Emmelmann This material is relatively cheap and has

96
 Harris and Brandt

excellent thetmal and optical properties. while giving good positioning accuracy and
The components are coated to minimize repeatability.
reflection losses. The Nd:YAG laser beam,
in addition to being able to be directed by
mirrors, can also be directed to the
workpiece through an optical fibre
(diameters typically 0.3 - 1 mm). The laser
may be easily switched between optic fibres
to share one laser between several
workstations or to share between
workstations concurrently. Fibres for
Nd:YAG transmission are typically of step
index design, meaning a high refractive
index core surrounded by a cladding of low
refractive index materiaL The change of
refractive index ensures total internal (and
highly efficient) reflection. The optic fiber
allows flexible and indeed remote operation
by robots with losses of 8% with no end
coatings and as little as 2% occurring in the
fibres with quartz windows (Rofin 2000).
Fibres as long as 100 m can be used
effectively (Ishide 1990). Fibres are Figure 9. 3D C02 laser cutting system.
surrounded by a metallic sheath to provide
mechanical protection and typically contain
continuity detection in the case of accidental An example of a 2D system is shown in
bum through, whereupon the laser is Figure 8 where a 2 axis hybrid system using
automatically turned off. a CO2 laser and flying optics can achieve
speeds exceeding 5Om/min and
3.4 Motion Systems accelerations exceeding Ims·2 with an
accuracy of ±0.01 mm. This approach
The focused laser beam is positioned on the offers great flexibility in processing large
workpiece either by moving the optics and sheets and allows complex shapes with
so the beam itself (flying optics, optical intricate detail to be cut. The main
fibres, or galvo optics) or by movement of drawback of these systems is the changing
the workpiece or a combination of both. focal spot position as it traverses the
One-dimensional systems process only in working area. Even though this change is
one direction and are typic ally used in the slight it is sufficient to affect the spot
manufacture of pipe, strip or section. Minor diameter on the workpiece and hence
process adjustments occur through process efficiency.
additional linear positioning axis. Two-
dimensional systems are used for the Three-dimensional systems use 5 or 6 axis
processing of flat surfaces. By moving the gantries or robot systems. For gantry
beam and beam guiding components less systems, positioning is performed using
inertia needs to be overcome than by three linear axes (i.e. X,Y,Z); beam
moving the workpiece. This allows high guidance can be by flexible arm or optical
speeds and high acceleration to be achieved fibre. Often the optics are moved in 2 axes

97
 Harris and Brandt

with the third performed by movement of vapour or the material vapour pressure
the workpiece. The laser optics are also itself shears and ejects molten materiaL
equipped with two rotational axes to ensure This technique generates narrow kerfs
the working surface is perpendicular to the and high quality surfaces for thin
beam path. Figure 9 shows a 3D cutting sections.
system with a 5 axis CNC hybrid system
using a CO 2 laser with mirror beam • Fusion cutting where the incident laser
guidance. Such systems can achieve speeds beam melts the workpiece and high-
exceeding 15m/min and accelerations pressure inert gas blows the molten
exceeding Ims·2 with an accuracy of ±O.Ol material out of the kerf. This method
mm. gives increased cutting rate over
sublimation cutting but includes the
formation of striations on the cut surface
and the adherence of dross to the lower
cut edge. The use of inert gas (typically
at pressures between 1 and 2 MPa)
allows the generation of non-oxidized
cut surfaces.

• Reactive fusion cutting where a reactive

gas such as oxygen is used in
conjunction with the laser heating of
Figure 10. Robotic Nd:YAG laser cutting at material. The resultant exothennic
Volvo (courtesy Rofin Sinar). reaction aids the cutting process and
high cutting rates can be achieved.

Robot systems are more effective for the

processing of 3D components· (e.g. in the
automotive industry) and are now becoming
increasingly common because of the ease of
manipulating the processing head. Robot
solutions are typically less accurate than
gantry systems but are considerably cheaper
to implement. Figure 10 shows a robot
cutting system using a 6 axis robot and
Nd:YAG laser coupled to the robot using
optical fibre.

4. THE LASER CUTTING Figure 11. Laser cut edge produced in mild
MECHANISM steel in the presence of oxygen assist gas. Note
the oxide layer.
The three dominant ways to cut materials
commercially are (Steen 1998): 4.1 Reactive Fusion Cutting ofMild Steel

• Sublimation cutting where the focussed Reactive fusion cutting is normally used
laser beam evaporates the material and when cutting mild steel. The interaction
the co-axial assist gas removes the between molten iron and oxygen produces

98
 Harris and Brandt

an additional source of heat due to the series of steps shown in Figure 12 (Powell
exothennic nature of their interaction. The 1993). The cutting mechanism involves the
exothermic reaction typically provides 40% following steps:
(Ivarson 1991) of the overall heat input to
the cut. This is done by the conversion of (i) The laser beam and co-axial oxygen
approximately 50% of Fe removed to FeO move to the edge of the steel sheet and
and small amounts ofFe203. the area of beam illumination on the
plate increases and temperature rises to
The exothennic reaction between oxygen a point where initially ignition and
and iron is represented by: subsequently melting will occur.
(ii) The melt front once established moves
Fe+~02--+ FeO away from the center of the illuminated
MI = -257.58 kJ/mol (at2000K) area until reaching the outer regions of
illuminated area.
2Fe + 3/2 O2 --+ Fe203 (iii) Having left the energetic area at the site
of illumination, the melt front cools and
AH = -826.72 kJ/mo! (at 2000K) extinguishes.
The process involves melting the material in (iv) The laser then initiates ignition and
the upper section of the workpiece by the melting in the next area and the process
focused beam from the laser. This molten repeats itself.
material transfers its heat to lower areas~
which in tum also melt. All the molten
material is free to react with oxygen present
as the assist gas creating further heat by the
exotheffi1ic reactions shown above. Molten
material is removed through the bottom of
the cut by the shear with the assist gas. As butrJing extinction reinitiation striation genuation
there is continuous interaction between the
energy provided by the focused beam and Figure 12. The periodic nature of the reactive
the workpiece during cutting, a dynamic fusion cutting of steels.
steady state condition is established. The
parameters that influence the cutting process
include laser power (continuous or pulsed), This model clearly explains the regularly
focus position, surface condition of spaced striations left on the cut edges. In the
material, assist gas pressure and cutting top section, the cut striations are more
speed. The interaction between each of regular, however, in the lower half of the
these variables is not completely understood cut random ripples are created by the flow
and the influence of lesser factors (eg beam of the molten material out of the cut zone.
mode, material temperature (Powel1 1993), This random pattern increases with
nozzle clearance (O'Niell 1992») has also increasing material thickness reflecting the
been shown to be significant. An example greater disturbance of the liquid metal
of a reactive fusion cut surface is shown in through the cut zone. The final cut surface is
Figure 11. therefore the result of predominantly
thermodynamic interactions near the top of
The cutting mechanism is shown to be the cut and fluid dynamic interactions
periodic in nature and can be described by a towards the base (powell 1986).

99
 Harris and Brandt

4.2 The Effect of Oxygen Purity

The purity of the oxygen assist gas is a

critical aspect of the reactive fusion process.
The relationship between oxygen purity and
-Omm
cutting speed is shown in Figure 13. 0.11& • ••• 3mm
Studies by Ivarson (1993) show that a build
,
,. I
--- 6mm
up of impurity gas at the gas-molten metal 0.84 , : ;.. " ,
,
h,",,,"'''''''' ._.................' - - IOmm
-.-20mm
interface during cutting is the main factor
responsible for the observed reduction in 0.8
J 1.5 1 2.5 J
cutting speed. It can be seen from Figure 13
Nuult' dtamfttr (mm)
that a 2% reduction in oxygen purity results
in a 50% reduction in cutting speed,
demonstrating the sensitivity of the process Figure 14. Theoretical nonnalised oxygen
to oxygen purity. concentration at the cut front as a function of
nozzle diameter \vith cut front location depth as
a parameter (O'Nie1l1992).
The build up of impurity gases
(pre40minantly nitrogen) at the cutting face
is a consequence of the turbulent mixing of Given that a critical oxygen purity threshold
the oxygen jet with the surrounding is approximately 98% (Figure 13), then a
atmosphere (O'Niell 1992). This process is Imm nozzle is effective to a steel thickness
called entrainment. This effect increases of 6nun, 1.5mm nozzle to a thickness of
with increasing distance downstream of the 10mm and 2.5mm nozzle to a thickness of
nozzle exit. A 2D numerical model and at least 20 mm. Further, the use of lower
subsequent experimental studies found that pressure assist gas such as that described in
changing oxygen concentrations occur at Powell (1987) presents a less turbulent or
different distances along the cut front when possibly laminar jet with lower
using various nozzle diameters as illustrated consequential entrainment. As a result, the
in Figure 14. effect of entrainment of ambient gases on
reducing the oxygen purity is a serious
inhibitor of cut efficiency when cutting
thick steel plate.

4.3 Laser Cutting ofStainless Steels

Reactive fusion cutting of stainless steels is

similar to that of mild steel, however, in
- -1 - - -- addition to the oxidation of Fe, oxidation of
I
Cr in particular plays a part (Ivarson
1991,1993). The oxidation of these two
elements contributes some 40% of the
90
overall energy input to the cut zone.
Oxygen purity {%}
Approximately 30% of removed Fe in
stainless steel is oxidized to Fe203 and 30%
Figure 13. Cutting speed as a function of of Cr to Cr203 with negligible formation of
oxygen purity (Ivarson 1993). NiO. The reactions involved are given as:

100
 Harris and Brandt

2Fe + 3/2 O2--''''' Fe203 stainless steel fusion cut with high pressure
Llli =·826.72 kJ/mol (at 2000K) nitrogen.

2Cr + 3/2 O2 --..... Cr203

Llli - 1163.67 kllmol (at 2000K)

Ni +1/2 O2 --..... NiO

Llli = • 248.23 kJ/mol (at 2000K)

The presence of Cr also hinders the cutting

process. The formation of Cr203 on the
surface of the melt makes further oxidation
Figure 15. Laser cut surface produced in
difficult because of the inability of oxygen stainless steel in the presence of nitrogen assist
to diffuse through it (Powell 1993). gas.
Furthermore, the increased surface tension
of the Cr203 when compared to mild steel
inhibits melt removal and is frequently the 5. THE CUTTING OF TIDCK STEEL
cause of significant dross on the lower PLATE
edges of the kerf. During the resolidification
of the outer surface of the kerf a chromium The laser cutting of steel sheet and steel
rich oxide layer (5-10 ~m thick) and a plate up to 15mm in thickness is now a ,veIl
deeper chromium depleted layer (50-150 established machining process (Heidenreich
~m) is fonned. This oxide layer and 1996; O'Niell 1995). The ability to CNC a
decreased chromium layer at a laser cut laser, its cutting accuracy, clean cut edges
edge can lead to corrosion or wear issues. and low heat input mean that if thicker steel
plate could be reliably cut, the laser would
While higher material removal rates are become a serious competitor to plasma and
achieved through reactive fusion cutting due oxy-fuel cutting technologies. This is
to the exothennic reactions, the fusion evidenced in industries such as ship
cutting process produces a better cut surface building where a laser was used to cut
quality and dimensional accuracy of the precisely machined ship steel up to 16mm
kerf. In addition, the surface can be welded thick to DIN 2310 Grade II quality
because of the absence of an oxide layer. In (Heidenreich 1996). The successful use of
fusion cutting of stainless steels, high laser cutting allowed considerable savings
pressure nitrogen is used to eject the molten in time, as less post processing of the parts
material through the kerf. The nitrogen was required. With careful control of system
pressures usually range from 1 to 2 MPa. parameters it is possible to cut mild steels to
The reduction in cut energy (due to the 40 mm but with a significant decline in cut
absence of reactive fusion) and increased quality and reproducibility (O'Nie1l1992).
heat losses due to forced convection of the
high pressure nitrogen gas yield cutting When cutting thick steels with a fixed laser
speeds 25% lower than those of reactive power it has been observed that, in order to
fusion under similar cutting conditions. achieve a given quality of surface finish, the
Illustrated in Figure 15 is the cut surface of width of the kerf remains approximately

101
 Harris and Brandt

constant, however, the cutting speed does MPa of N2 . When using oxygen assist gas
not reduce proportionately to the thickness however, increasing the oxygen pressure
(Powell 1987). Figure 16 shows the cutting with increasing material thickness results in
speed as a function of material thickness for too much oxygen being present within the
two different qualities of cut (maximum kerf, which leads to excessive burning of
cutting speed is often achieved at the the steel due to the exothennic nature of the
expense of cut quality). It can be observed reaction. Consequently, to counter this
from the graph that a reduction in cut effect, oxygen pressure is reduced with
efficiency occurs with increasing material increasing material thickness. For example}
thickness. 8mm mild steel can be cut at a speed of
1800 mmlmin using an oxygen pressure of
0.2 MPa but when cutting 20 mm thick plate
80 at 1000 mmlrnin the oxygen pressure is
~'
- . - Highest speed
0' 70
- . - Optimal speed
reduced to 0.08 MPa (Powell 1987).
E 60 However, the lower the oxygen pressure,
.§.
-g 50
.\\ the more difficult it becomes to shear the
.~ ~\~
Gl
g. 40 molten material from the kerf thus limiting
en .. •
E
::s
30 the cut thickness.
o 20
10
.
. :
.
~,
;~.
... Close control of oxygen pressure is essential
o~~~~~~~~~~~~~~ to prevent uncontrollable burning of the
o 2 4 6 8 10 12 14 16 18 20 22 24
steel away from the heated area. Studies
Cut thickness (mm)
(Gabzdyl 1992) indicate that when cutting
Figure 16. Influence of material thickness on mild steel an increase of only 5 kPa in
cutting speed using a 2.5 kW CO2 laser (Po\vell oxygen pressure from the optimal pressure
1987). can result in excessive bUlning and excess
dross formation at the base of the cut.
Similarly, work by Heidenreich (1996)
This reduction in efficiency as the material using PrandtI boundary flow theory on thick
becomes thicker is attributed to a reduction ship steel plate demonstrated that by careful
in the effectiveness of the assist gas as a selection of oxygen assist gas pressure
melt remover and the loss of heat at the excessive burning could be controlled. Here
bottom of the cut due to conduction. a kerf width of more than 0.5 mm was
Further, as the material thickness increases produced and oxygen pressure of 0.12 MPa
the gas stream travelling downwards was used when cutting 12mrn thick steel.
experiences greater "fanning out" as a result However, solutions for satisfactory cutting
of the greater reflection off the inclined conditions using parameters of power,
cutting front thus progressively reducing the cutting velocity and oxygen assist gas
melt shearing capability of the assist gas pressure show that these three parameters
(Powell 1987). must have certain values and cannot be
varied independently (e.g. cut speed cannot
When cutting thick steels by fusion, the be altered without altering beam power and
assist gas pressure must be increased to oxygen pressure). Further to this,
allow the ejection of molten materiaL A Heidenreich (1996) also shows that changes
typical reported gas pressure (Beyer 1990) in oxygen pressure of less than 0.02 MPa
for fusion cutting of 15mm thick stainless are sufficient to change the cutting process
steels with a 5 kW CO2 laser source is 2 from metal burning to dross adherence.

102
 Harris and Brandl

In order to increase the thickness of the in tum leads to less disturbance of the gas
material that can be cut, one approach is to stream penetrating the cut. Material removal
increase laser power and hence use the occurs as a result of melt shear by the gas
power of the laser to overcome control stream from the oxy-acetylene cutter. An
difficulties of the reactive fusion process. additional gas nozzle is also used and
While there are a number of advantages to directed into the kerf to further assist in the
this approach there are also significant removal of molten metal and prevention of
challenges. At higher laser powers (3.5 kW dross adherence or sealing of the kerf at
and higher) the beam quality beCOlnes high cutting speeds. The resultant quality of
poorer, lifetime of optical components is cut is DIN 2310 grade II with a kerf width
reduced due to thennal loading, equipment of 2.5mm and a HAZ comparable to that of
and running costs are high and cutting an oxy-acetylene cut. The cutting speed is
precision deteriorates. In view of these greater than that for conventional flame or
disadvantages, a number of alternative laser cutting alone. Experiments conducted
approaches are being investigated for on 20 rom and 30 mm thick mild steel plates
cutting thick (> 10mm) steel plate. demonstrated that they could be cut at 1200
mmlmin and 900 mm/min respectively
6. INNOVATIVE PROCESSES TO using a 2 kW CO2 laser, a 150 mm focal
CUT THICK STEEL PLATE length lens and the focus positioned 2.5 rom
ahead of the cutting front (Heidenreich
The following describes several of the 1993).
innovative methods to further increase the
depth of laser cut while still maintaining Though successfully used to cut thick mild
acceptable cut quality. Many of the steel, the method has issues of the
processes described use novel laser directionality of laser/nozzle assembly that
techniques to overcome the limitations of may be overcome by CNC control of an
thick steel cutting already discussed. additional axis. The process also involves
high capital outlay, high gas consumption
6.1 Laser Flame Cutting and lacks a niche application.

Thennal flame cutting is an effective cutting 6.2 Cutting with Dual Focus Lenses
process capable of cutting very thick steels
(Heidem'eich 1993). Its main drawback, The most effective energy source shape for
however, is that it is a slow process ("'" 400 cutting is a line source (O'Neill 1999),
mmlmin for 19 mm thick mild steel (Powell because it can distribute the energy
1993). Lasers on the other hand cut at high uniformly throughout the depth of the cut.
speed depending on the material thickness. During reactive fusion cutting of steel, the
By combining the two techniques the exothennic interaction of the melt with the
advantages of each can be exploited. narrow oxygen jet as it travels down the
kerf does assist in the creation of an ideal
In the augmented laser oxy-acetylene line energy source. For mild steels greater
technique the material is heated by an oxy- than 10 rom in thickness the focal point is
acetylene flame that is inclined at an angle typically positioned at the top surface of the
of 85° to the metal sheet with the laser beam workpiece to allow a significant intensity of
heating a small workpiece area in front of light on this surface to initiate the cutting
the oxy-acetylene cutting nozzle. The laser process and to achieve process stability with
heating creates a larger molten zone, which acceptable cut quality. For fusion cutting of

103
 Harris and Brandt

stainless steels, however, this focus is

typically positioned well into the kerf. This mner
allows maximum energy intensity to occur outer
at the central zone of the cut where the dual focus lens
molten material must accumulate enough
heat energy to conductively melt material
below it (powell 2000).
upper focus
With increase in material thickness it
becomes increasingly difficulty to cut as it
is not possible to maintain adequate energy lower focus
intensity at all positions through the depth
of the kerf. To help counter this problem,
higher laser powers and longer focal length
lenses are used. This gives a longer depth of
focus, however, this means an increase in Figure 17. Schematic of dual focus meniscus
spot size and commensurate decrease in lens. For illustration purposes the laser beam is
power density within the spot. Further, divided into an inner section (clear) and an
higher powers degrade beam quality, outer section (shaded).
leading to beams not suitable for cutting
(Nielsen 1997).
The dual focus method shows improved
A new lens developed for CO2 lasers has cutting of thicker material (30 mm 40 mm
t\vo focal spots, one situated above the compared to 15mm - 20mm carbon steel for
other. This produces focused energy deep a 3 kW CO2 laser) improved piercing with
within the cut but still allows high energy little surface eruption (Neilsen 1997;
density at the top of the cut in order to O'Neill 1999), improved quality of cuts
sustain the cutting process (Powell 2000). (dross free edges) and improved cutting
Previously, these types of optics were based speeds. When cutting stainless steel of 12
on multi-beam principles (i.e. combination mm thickness dual focus lenses produce a
of two laser beams). The new method uses 30% narrower kerf with more perpendicular
an inner and outer annulus within a single edges and less dross while cutting 23 %
lens, each with a different focal length. The faster than single focus lenses (Powell
inner annulus being flatter compared to the 2000). The use of dual focus lenses
outer annulus has a longer focal length to demonstrates a more efficient method of
focus the beam deeper \vithin the cut as cutting thick steels with more energy
shown in Figure 17. The area of lens entering the cut, producing higher cut
dedicated to each focus governs the temperatures and consequently giving lower
proportion of power distribution of the laser viscosity melts. This results in the removal
beam. The first focussed beam is maintained of less material and reduction in dross and
at the upper surface of the metal to create dross attachment. The disadvantage of the
suitable ignition conditions while the second technique is the high cost of dual focus
focused beam, near the lower surface, lenses. Also, they are designed for a
creates a lower viscosity melt resulting in a particular material thickness making them
cleaner kerf throughout the thickness of the less flexible when compared to conventional
material. lenses.

104
 Harris and Brandt

6.3 CO2 Beam Sawing with Adaptive Optics thickness without a decrease in cut quaHty
using a 2.2 kW CO2 laser.
This technique involves the oscillation of Using a focal length of 160 mm the initial
the focus position, by several millimetres, in focus position was Imm above the surface
the same direction as the material thickness of the workpiece. The maximum focus
as a result of one of the reflective mirrors of travel distance was 4.5 mm after which the
a CO2 laser being plastically deformed by a beam intensity at the top of the cut was
piezo-electric actuator (Geiger 1996). An reduced sufficiently to allow fonnation of
external voltage controls the defonnation excessive dross. Results show that
frequency from 0 Hz to the experimental striations in the top section of the cut reflect
limit of 100 Hz. This method has been the frequency of the oscillating beam as is
shown to cut mild steel up to 16 mm shown in Figure 18.

h! -meosur 119
.) mOl rJi~l()if
the tou sid~:

5 mrn

OXj'gen ossisled loser beam sawing 25H1 leser beam sowing tomb
Il)~erbeam Gutting

Figure 18. Cut surfaces at various frequencies (Geiger 1996).

laser beam cuU!n

Rz -measuring
~ 3mm below
the top side
~ 1.0 mm
~~~115mm
cutting gas: 01 pressure: 1.1*105 Po
focol shift: 4-.5 mm velocity: 0.7 m/min

Figure 19. Cut surface profile showing reduced amplitude of striations with increasing sawing
frequency (Geiger 1996).

105
 Harris and Brandt

Rectangularity of the cut edges is also

Cl'ntral ox,gcn input
affected with improvement reported in
Central nOlzle samples cut at frequencies greater than 75
Hz. No mention is made by Geiger (1996)
of dross removal issues except a comment
that another gas nozzle system is required to
optimize the dross problem.
{h~'fD burrrr th)gtfl buffer
in,.ot ltll.'U
6.4 Laser Cutting using a Co-axial Nozzle

The flow of assist gas through the laser

nozzle comprises an inner stream in which
the pressure and velocity remains constant,
Ring nOlzle
(llatr
' " Burrer jet and an outer boundary where momentum}
Mainjct material and heat diffusion take place.
Narrow nozzles are shown to induce more
(a)
mass transfer with the surrounding
environment within a small distance of the
c 1.00
~ nozzle exit (O'Niell 1992, 1995). Transfer
g 0.96 is more pronounced with turbulent flow, jets
cG)
(.)
c typically becoming turbulent several nozzle
0 0.92
(.)
cG)
diameters downstream. As the nozzle
C')
0.88 - - Buffer - theoreIicaI diameter is increased the effects of the
~ - . - Buffer - experirnerla
0
"0 J. No buffer - theoretk::aI
momentum transfer are shifted further into
G) 0.84
.!:'l -e- No buffer - experirnerla the materiaL Typically, turbulence cannot
'iii
§ 0.80 be avoided unless low Reynolds numbers
0
z are used. The most important issue for
0 5 10 15 20 25 reactive fusion laser cutting is the reduction
rxstance dO'Nn cut front (rrni
of oxygen concentration through the
entrainment of surrounding environmental
(b)
gases such as nitrogen. As discussed in
Figure 20. Illustrated in (a) is layout of buffer Section 4.2, small losses in oxygen
nozzle and (b) theoretical and experimental concentration have been found to
oxygen concentrations down the cut face for a significantly reduce cutting speed.
1.5 rom diameter nozzle (O'Neill 1995).
To combat the effects of entrainment of
ambient gases 0 'Niell (1992, 1995)
Significantly, the roughness of the cut suggested the use of a co-axial nozzle. A co-
surfaces is shown to be improved over those axial nozzle illustrated in Figure 20
of conventional cutting at frequencies comprises two concentric nozzles. The
exceeding 25 Hz, with maximum larger diameter, inner, nozzle can be used to
improvement occurring at 100 Hz as can be maintain oxygen concentration further into
seen in Figures 18 and 19. At the maximum the cut. As previously discussed, larger
frequency of 100 Hz, upper striations reflect diameter nozzles ensure larger flows of
the oscillation of the focal position but oxygen resulting in a reduction and shifting
lower area striations are more random of entrainment to larger distances from the
reflecting the liquid flow from the kerf. nozzle. The second, sacrificial, "entrainment

106
 Harris and Brandt

buffer " is created using an outer annulus to The device shows promise for commercial
allow a turbulent mixing zone. This allows exploitation because of the ease of
the primary cutting jet to experience less integration with standard cutting head
mixing with the atmosphere (i.e. the designs. The main drawback of the device
entrainment process still occurs but in the is the increase in oxygen consumption rate.
outer jet). A combined theoretical and
experimental oxygen concentration with and 6.5 Dual-bealn CO 2 Laser Cutting of Thick
without buffer is shown in Figure 20(b). Metallic Materials

Results of cutting 10 mm, 16 mm and 20 When focusing laser light onto a workpiece
lnln grade 43A mild steel plate with and there may be coupling problems if plasma is
without the buffer jet demonstrate formed at the surface. By using two laser
significant improvement in cut quality for a beams separated by a small distance there
given cut speed when the buffer jet is used. can be improvement in coupling efficiency
There is little change in cut quality for the due to minimization of the laser plasma
10 nun plate because at a depth of 10 nun interactions. There is also an advantage due
there is only a 1% reduction in oxygen to the increased absorption associated with
concentration for a single nozzle. There is, the molten surface (Molian 1993). The dual
however, a significant improvement in cut beam approach was adopted by Molian and
quality for the 16 and 20 mm thick plate used in laser cutting experiments. The
using a central pressure of 0.05 MPa and an process involves beams from two CO2
outer flow greater than 30 Vmin. lasers of po\ver 1.5 kW each delivered
through a single beam duct. Both focused
Processing with the co-axial nozzle has beams were 0.1 mm in diameter, with foci
allowed the cutting of thick steel plate positioned 6 mm apart and set 1/3 of the
producing clean cut edges and effective material thickness below the top surface.
removal of dross as a result of increased Nozzle diameter was 10 mm with oxygen
exothermic energy at the base of the cut. pressure set at 137 kPa (Malian 1993).

G\.\on
\lOt;>°f\l

Cuttinq direction, speed V

.-----
Cutting Erosion front
depth, thi(kne~S,
Oct- De
t
Dri Iling Molten melt/slog
depth,
Dd~_""""""-'--L---.t'!'!!'

Figure 21. Schematic of dual-beam laser cutting (Malian 1993).

107
 Harris and Brandt

The first beam partially penetrates the and analytically demonstrated that the
moving workpiece and fonns a blind maximum cutting speed increases as beams
cylindrical keyhole. The second laser are separated in the direction of the cutting
impinges on the molten region created by velocity, the optimum depth could be gained
the first beam and further heats it, by combining the two beams into one
vaporizing some material and superheating single, more powerful beam. Technical
the remainder. The oxygen assist gas aids difficulties were reported however in
the cutting process with its exothermic achieving a stable beam coincidence within
reaction and in the ejection of molten one spot diameter (0.1 mm). Evidence of cut
droplets. A schematic of the technique is quality was not presented, the paper instead
given in Figure 21. being significant in the development of and
verification of heat flow models through
experimentation.
40
35
•
Again, this approach demonstrates an ability
~ 30 to increase the thickness of steel plate that
S 25 can be cut but its major disadvantage is that
'tI
III
III
Co
1/1
20 it has directional dependence thus making it
m
c 15 difficult to use for profile cutting.
E:;:,
10
0
5

0
6 8 10 12 14 16 18 20
Thickness (mm)

Figure 22. Effect of single (1.\) and dual (0)

1.5kW C02 laser beams on cutting speed and
material thickness. Focal length of lenses was
127 mm and oxygen assist gas pressure was 413
kPa.

The workpiece preheating improves the

'deep penetration effect' mainly due to the Figure 23. Schematic of Lasox cutting process
improvement in absorption and (O'NieU1998).
modification of conduction characteristics.
This method was tested by cutting mild steel
(ATST 1018) and superalloy (Hastalloy C) 6.6 Laser Assisted Oxygen Cutting - Lasox
plates. For a fixed laser power,
improvement in cutting speed and cut depth As discussed in Section 4.1, during the
was observed for the dual beam conventional reactive fusion cutting of mild
configuration when compared to single steels, approximately 40% of energy is
beam as shown in Figure 22. provided by the exothennic reaction and in
excess of 97% of combustion products is
A mathematical heat flow model for the FeD (the remainder being Fe2D3 and Fe2D4
dual beam cutting process was developed by (Ivarson 1993). Assuming that FeD is the
Molian and shown to fit well with only reaction product produced and with AH
experimental data. It was experimentally 257.58 kJ/mol then if complete

108
 Harris and Brandt

cOlnbustion of the iron takes place, the and controlled by the oxygen assist gas
power generated in such a process is given delivered co-axially with the beam.
in Table 2. Hence, large power is available
to the cutting process. However, to generate The Lasox method employs a nozzle
this power there must be sufficient oxygen assembly and a short focal length lens
delivered to the kerf. It is well documented where the focal length is within the
that during the cutting of thick steel nozzle/lens housing. This results in a highly
sections, side burning becomes an issue. To divergent laser beam exiting the nozzle with
minimize this effect oxygen pressure 1S a co-axial oxygen gas stream. The nozzle
reduced to less that 100 kPa. diameter is such that the laser interaction
area is greater than the gas jet interaction
In the Lasox process (see Figure 23) a laser area but the laser interaction still generates
beam is employed to assist the oxygen temperatures well above the critical 900°C
cutting process. The laser is used to preheat (O'Niell 1998) required for initiation of the
the material while the cutting is performed exothermic reaction.

Table 2. Calculated exothermic power generated in thick section laser cutting (O'Niell 1998).

Cut Depth Kerf Width Cutting Speed Exothermic Power

(rom) (mm) (mmfmin) (kW)

20 1.5 700 12.5

30 2 500 18.5

40 2.5 400 23.6

50 3 300 21.52

Table 3.. Plate thickness and corresponding cutting speeds (O'Niell1998).

Plate Thickness (nun) Optimum Speed Speed Range

(mmfmin) (mmlmin)

20 300 200 500

30 230 200 350

50 175 150 - 220

109
 Harris and Brandt

The laser/oxygen jet interaction ensures that this system to existing lasers making its
a reaction can be initiated across the whole implementation attractive.
of the oxygen jet, maximizing reaction yield
and ensuring that the Lasox condition is
met. Heating of the workpiece occurs at the
surface and the ensuing exothermic reaction
occurs through the depth of the cut. If the
above laser/gas condition is not met, a poor
quality cut is generated as a result of
uncontrolled metal burning. It is essential
to realize that the laser is depositing very
little energy into the depth of the cut and
that oxygen is the facilitator of melt
production. Results of cutting trials using a
2 kW CO2 laser and 20 mm, 30 mm and 50
nun mild steel plate (O'Niell 1998) sho\v
that the cutting action is driven almost Figure 24. Lasox profile cuts in 40 mm thick
exclusively by the oxygen jet reaction and EN9 steel (O'Neill 1998).
that the laser contributes little to the overall
energy of the process. lliustrated in Table 3
are the cutting speeds obtained. 6.7 Thick Plate Cutting with a Spinning
Laser Beam
It was observed that as the cut speed
increases above optimum) dross levels Another approach that has shown potential
increase dramatically. Under optimum for cutting thick section steels involves
cutting conditions the cut taper was shown spinning the laser beam while performing a
to be less than 1 degree with the top edge conventional laser reactive fusion cut. The
square as compared to conventional oxy- feasibility of this approach was
flame cutting which produces a round edge. demonstrated by Arai (1997) who used a 1.8
The HAZ was of similar thickness to flame kW CO2 laser to cut mild steel and stainless
cutting in the body of the workpiece but steel plate up to 25 mm thick while
without any increase near the top surface as maintaining high cut quality. It was also
produced by flame impingement in flame demonstrated that there was minimal HAZ
cutting. Widening of the HAZ occurs at the and distortion. To produce the desired
base of the cut if optimum conditions are effect, a window is placed in the beam path
not sustained due to a reduction of melt at an angle to the beam axis as can be seen
flow velocity. Also, piercing is found to be in Figure 25. This produces a lateral shift of
possible along with trepanning though some focal position. Rotating the window causes
amount of taper is expected. The cutting of the laser beam to form a spiral path as it is
profiles was shown to be feasible as moved across the workpiece.
illustrated in Figure 24.
While Arai demonstrated that relatively lo\v
Consequently~ Lasox is a relatively simple power lasers can be used to cut thick steel
process that shows promise for cutting of plate with acceptable quality, little
thick section steel plate using a controlled infotmation was presented as to the
ignition process and only moderate laser mechanisms responsible for the increase in
power. It may also be feasible to retrofit thickness. In order to understand the

110
 Harris and Brandt

processes occuning during spinning beam in an increase in maximum cut depth from
laser cutting, a project was initiated to 12 to 15 mm when compared to
determine the key process controlling conventional cutting. This is illustrated in
factors and determine the maximum Figure 28 where a comparison between
material thickness that can be cut with a conventional cutting and spinning beam
fixed laser power. The project is part of a cutting of the various plate thicknesses is
larger project involving the cutting of thick shown. Attempts at cutting the 20 mm thick
steel plate with laser beams supported by plate were pal1ially successful. While the
the Co-operative Research Center for cut depth was increased, excessive dross
Intelligent Manufacturing Systems and accumulated at the base of the kerf and
Technologies (CRCIMST). there was excessive burning of the sides.
From the figure it can be observed that a
continuity of the conventional cutting
Lens characteristics extends into the spinning
beam area. That is, as cutting progresses
into those thicknesses cut by the spinning
beam, the one characteristic curve can
describe both processes for a given laser
power.

Figure 25. Spinning beam apparatus (Arai

1997).

A spinning beam apparatus \vas designed

and constructed as shown in Figure 26.
This was attached to a fibre-optic delivered
2.5 kW Nd:YAG laser. This allowed a
maximum offset ( eccentricity) of beam of
450 Jlm at a maximum spin speed of 3000
RPM. In preliminary trials, the apparatus
demonstrated the spiral cut path as shown in
Figure 27. Here a 200W Nd:YAG beam
was used to mark the surface of a mild steel
plate from left to right in (a) conventional Figure 26. Spinning beam apparatus cutting 12
mode and (b) spinning beam mode. mm thick mild steel plate.

In cutting trials involving the device and

mild steel plate thicknesses of 10 mm, 12 The figure also demonstrates that although
mm) 15 mm and 20 mm, it was there is improvement in cut depth with the
demonstrated that the spinning beam results spinning beam, there is only modest

111
 Harris and Brandt

400um 920um

L
r t
(a) (b)

Figure 27. Beam marking at 200W representing (a) a conventional cut and (b) spinning beam cut at
440 RPM} 220 ~m offset and 150 mmlmin cutting speed.

improvement in cut speed. When cutting experiences an increase in effective cut

the 12 mm mild steel plate at 1488 watts speed and the other a corresponding
incident on the workpiece and a cutting decrease.
speed of 700 mmlmin an average kerf width
of 1.2 mm was produced for conventional
2400
cutting and 2.4 rom for a 450 Jlm offset 2200 -·-1339W
beam at 2000 RPM. Hence, during the 2000 .. -.t.-1637W
C -y-1769W
cutting process similar process limitations to :§ 1800
-'-1339W SpIn
cut speed are being imposed even though E 1600 -+-1637W Spin
.§. 1400 -_.- 1769 W Spin
cutting with the spinning beam produces a -l(-

far larger kerf resulting in a larger volume

"
~
~
Q.
1200
(J)
1000
of material being removed and a larger 11
0 800
amount of oxygen entering the cut. 600
400
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Current research into the possible causes of Plate Thickness (mm)
limitations to cut speed for any given power
indicate that the two main contenders are (i) Figure 28. Dependence of cut speed on plate
limitations to the amount of energy thickness for different incident Jaser powers.
available per mole of material removed, and
(ii) amount of oxygen available per mole of
material removed. Distinct thresholds Illustrated in Figure 30 are cut surface
appear to exist for both parameters for profiles obtained using a spin speed of 2000
successful cuts to be perfonned. RPM and conventional cutting conditions
(no spin). It is noted that the spinning beam
The spinning of the beam also affects the uses larger nozzle diameters consequently
shape of the kerf as seen in Figure 29. The when the spinning beam is stationary it does
spinning beam kerf is non-symmetric and not imitate the conventional cut.
results from the difference between the Furthermore, the offset of the beam when
rotational velocity of the beam and the stationary approximates a poor nozzlelbeam
translational velocity of the workpiece. alignment which is a crucial cut quality
Consequently, one side of the kerf parameter.

112
 Harris and Brandt

Figure 29. Kerf shape for (a) spinning and (b)

conventional laser cut in 12mm mild steel plate.

age of ever increasing energy consciousness

this method of cutting thick steel plate may
prove an advantage in the next generation of
more efficient laser cutting machines.
Further work is being carried out to
understand the mechanisms responsible for
the observed increase in cut thickness.

2000 RPM
7. SUMMARY

Today, the laser cutting of 2D and 3D steel

components is a well established and
flexible non-contact machining process
producing narrow kerfs and minimal heat
affected zones. Currently, commercial lasers
cut steel plate up to 15mm thick with
relative ease. There is, however, industry
Stationary (no spin) interest to increase this thickness which
would allow lasers to be a significant
competitor to plasma and oxy-fuel
Figure 30. Changes in cut surface characteristic
technology in thick plate cutting.
using spin speed of 2000 RPM and stationary
beam for 12mm thick mild steel plate. Laser
power 2000W; cut speed 600 mm/min and The cutting of steel plate using lasers is
oxygen assist pressure 40kPa. achieved by the continuous interaction of
the energy supplied by the focussed beam,
From the research conducted to date, the the feeding of the workpiece and assist gas.
spinning beam cutting approach is showing Using the reactive fusion technique oxygen
itself to be significant in increasing the provides momentum transfer for shearing of
depth of cut that can be achieved, but not the melt as well as for the exothermic
significant in improving the cut speed. In an reaction with the molten steel. There are

113
 Harris and Brandt

limitations inherent with the reactive fusion Belforte, D. A. (Jan. 2001). "The new year
cutting of thick steel plate. These limitations repeats the old H • Industrial Laser
are primarily attributed to a reduction in Solutions, 13 18.
shear forces produced by the assist gas as Bharti, A. and Sivakumar, R. (1996).
the cut depth increases and a need to reduce "Mechanism of material removal in laser
oxygen assist gas pressure to control fusion cutting". Lasers in Engineering, ~,
burning. These two complementary factors 87-105.
appear to limit the thickness of material that Convergent web site~ www.convergent-
may be cut. To overcome these factors a energy. com.
number of innovative cutting techniques Dawes, D. G. (Dec 1987). "Growing
described above are being investigated. The crystals for solid state lasers"', Lasers and
spinning beam method shows potential for Optronics.
cutting thick steel plate, demonstrating Duly, W. W., Semple, D. 1., Morency, J ..P.
significant increase in cutting depth when and Grovel, M. (1979). "Coupling co-
efficient for CW CO 2 laser radiation on
compared to conventional cutting at the
stainless steer'. Optics and Laser
same laser power. Further work is being
Technology~ 281.
undertaken to fully understand the
Emmelmann, C. and Lunding, S. (1995).
mechanisms responsible for this increase in
"Introduction to industrial laser material
cut thickness. processing". by Rofin Sinar Laser
GmbH, Hamburg, Germany.
ACKNOWLEDGEMENTS Emmelmann, C. and Piening, A. (June
1999). "Diode-pumped solid-state lasers
The authors wish to thank CRCIMST for for industrial laser material processing".
funding the spinning beam project. The Proceedings 32 ISATA, Vienna, 359 -
CRCIMST was established under and is 366.
supported in part by the Australian Engblom, G. and Falck, K. (1990). "Quality
Government's Cooperative Research classification of thermally cut surfaces -
Centres Scheme. comprehensive review of different
standards". Welding in the World, 28, no.
REFERENCES 11112,26 - 37.
Fieret, J. et a1. (1987). "Overview of flow
Arai, T. and Riches, S. (1997). "Thick plate dynamics in gas-assisted laser cutting".
cutting with spinning laser beam". Proceedings of SPIE, 801,243 -250.
Proceedings of ICALEO 97, part B, Gabzdyl, J. T. and Morgan, D. A. (1993).
Laser Inst. of America, 19 - 26. "Assist gases for laser cutting of steels".
Arata, Y., Maruo, H., Miyamoto, I. and Proceedings of SPIE, 1990, 443-447.
Takeuchi, S. (1979). "Dynamic Gieger, M., Shub erth, S. and Hutfless, J.
behaviour in laser cutting of mild steel", (1996). "C02 laser beam sawing of thick
Trans. JWRI, B, 15. sheet metal with adaptive optics".
Bachmann, F. (1998). "Applications of Welding in the World, no. 1, 5- 11.
high power diode lasers". Proceeding of Havrilla, D. and Anthony, P. (Feb. 1999).
ICALEO 98, Europe, Laser Inst. of "Process fundamentals of industrial laser
America. welding and cutting". Rofin Sinar Inc,
Beyer, E. and Petring, D. ( 1990). "State of Plymouth Michigan USA.
the art in laser cutting with CO2 lasers". Heidenreich, Band Sepold, G. (May 1993).
Proceedings of ICALEO 90, Laser Inst. "Laser flame cutting of thick steel
of America, 199-205. plates". Industrial Laser Review, 6 7.

114
 Harris and Brandt

Heidenreich, B., Juptner, W. and Sepold, G. M. (1992). "The dynamical behaviour

(1996). "Fundamental investigations of of gas jets in laser cutting". Proceedings
the burn-out phenomenon of laser cut of ICALEO 92, Laser Inst. of America,
edges". Lasers in Engineering, ~, 1-10. 449 - 458.
Hilton, P.A., (August 1997). "The early O'Niell, W.O., and Gabzdyl, J.T., (1995).
days of laser cutting". Industrial Laser Welding in the World, 35, no.1, 6-11.
Review, 14 - 16. O'Niell, W.O., Gabzdyl, J.T. and Stracey,
Ishide,T., Matsumoto,O., Nagura,Y. and R., (1998), "New developments in laser
Nagashima,T. (1990). "Optical assisted oxygen cutting". Proceedings of
Transmission of 2kW CW YAG Laser ICALEO 98, Laser Inst. of America, 88-
and its practical application to welding". 97.
Proceedings of EC03 Conference on O'Neill, W. (1999). "Laser cutting with dual
High Power Solid State Lasers and focus lenses". The Industrial Laser User.
Applications, 188. Issue 16, August, 27- 28.
Ivarson, A., Powell, J. and Magnusson, C. Powell, J., King, T. G., Menzies, I. A. and
(Fall 1991). "Role of oxidation in laser Frass, K., (1986). "Optimisation of
cutting stainless and mild steel". Journal pulsed laser cutting of mild steels". Proc.
of Laser Applications,~, no. 3,41-45. 3rd Int. Conf. on Lasers in
Ivarson, A. (1993). Doctorial Thesis "On Manufacturing, Paris, France, 67 -75.
the Physics and Chemical Powell, 1., Frass K. and Menzies, I.A.
Thennodynamics of Laser Cutting", (1987). "2.5kW laser cutting of steels;
Lulea University of Technology. Factors affecting cut quality in sections
Ivarson, A., Powell 1., Kamalu 1., Broden, up to 20 mm". Proceedings of SPIE ,
G. and Magnusson, C. (1993). "The 801, 278 -282.
effects of oxygen purity in laser cutting Powell, J. (1993). "C02 Laser Cutting".
mild steeP'. Lasers in Engineering, 1, 299 Springer-Verlag, New York, USA.
- 314. Powell, J., Tan, W. K., Maclennan, P.,
Koechner, W. (1988). "Solid-state Laser Rudd, D., Wykes, C. and Engstrom, H.
Engineering," 2nd Ed, Springer-Verlag. (2000). "Laser cutting stainless steel
Laos, 0. (1983). "Evaluating a CO 2 with dual focus lenses". Journal of Laser
Industrial Laser Systems". Proceedings Applications, 12, no 6, 224 -231.
First International Conference in Lasers Rickfalt, C. and Broden, G. (1995). "Oxy-
in Manufacturing, IFS Publications, fuel cutting of thick plates" . Welding in
Amsterdam, 21- 30. the World, 35 no A, 238 - 242.
Molian, P.A. (1993). "Dual-beam CO 2 Rofin Sinar (2000). "Introduction to
laser cutting of thick metallic materials". Industrial Laser Materials Processing".
Journal of Materials Science, 28 no. 7, Rofin Sinar Laser GmbH, Germany.
1738 - 1748. Schuocker, D. (1993). "The physical
Na, 1. et al. (1989). "Effect of shielding mechanism and theory of laser cutting".
gas pressure in laser cutting of sheet Industrial Laser Handbook, Penwell
metals". Journal of Engineering Books, 65-79.
Materials and Technology, Transactions Steen, W.M. (1998). "Laser Material
of ASME, 111, 314 - 318. processing" (2nd Ed.). Springer, London.
Nielssen, S.E. (April 1997). "Developments Ttumpfwebsite: www.trumpf.com
in laser beam cutting of thick materials", van Dijk, M. H. H. (1993). "Pulsed
Industrial Laser Review, 11 - 13. Nd:YAG laser cutting". Industrial Laser
O'Neill, W., Gabzdyl, 1. T. and Steen, W. Handbook, Penwell Books, 52-64.

115