Accepted Manuscript

Title: Laser cutting using off-axial supersonic rectangular

nozzles

Authors: A. Riveiro, F. Quintero, J.del Val, M. Boutinguiza, R.

Comesana, F. Lusquinos, J. Pou

PII: S0141-6359(17)30159-9
DOI: http://dx.doi.org/doi:10.1016/j.precisioneng.2017.07.013
Reference: PRE 6623

To appear in: Precision Engineering

Received date: 16-3-2017

Revised date: 20-7-2017
Accepted date: 25-7-2017

Please cite this article as: Riveiro A, Quintero F, Val Jdel, Boutinguiza M,
Comesana R, Lusquinos F, Pou J.Laser cutting using off-axial supersonic rectangular
nozzles.Precision Engineering http://dx.doi.org/10.1016/j.precisioneng.2017.07.013

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Laser cutting using off-axial supersonic rectangular nozzles.

A. Riveiro1, F. Quintero1, J. del Val1, M. Boutinguiza1, R. Comesaa2, F. Lusquios1, J. Pou1

1
Applied Physics Department, University of Vigo, EEI, Lagoas-Marcosende, Vigo, 36310, SPAIN.

2
Materials Engineering, Applied Mechanics and Construction Dpt., University of Vigo, EEI,
Lagoas-Marcosende, Vigo, 36310, SPAIN.

Graphical abstract

Highlights

A new assist gas injection system for laser cutting has been proposed.
Off-axial supersonic rectangular nozzles improve the efficiency of current assist gas
injection systems.
Suitable values for the impingement point, stand-off, and inclination of the supersonic
rectangular nozzle were determined.

1
 Dross free cuts, with low roughness (Ra=1 m), and negligible HAZ were obtained during
laser cutting of a 2024-T3 alloy.
Microstructure in the cut edge remains unaltered.

Abstract

Laser cutting is a well-established process in the industry, mainly for cutting metallic materials
such as steels. However, the utilization of this laser technique to process advanced materials,
such as ceramics or some aluminium alloys, is restricted due to the formation of a large heat
affected zone (HAZ), rough cut edges, presence of dross, or thermal cracks. These defects are
closely related to the performance of the assist gas during the removal of molten material.
Previous works have demonstrated the impressive improvement in cut quality with the
utilization of off-axial supersonic axisymmetric nozzles to inject the assist gas. This is due to the
higher removal of molten material. However, these results can be even improved if the
geometry of the gas jet is tailored to the cut kerf. In the present work, a cutting head with an
off-axial supersonic rectangular nozzle was developed to improve the efficiency of the current
assist gas injection systems. The performance of this system was compared with that of a
converging (coaxial) nozzle, and an off-axial supersonic axisymmetric nozzle during the
processing of Al2024-T3 sheets. Results demonstrated the superior performance of the new
assist gas injection system in terms of cut quality and productivity (cut edge roughness was 6.5
times lower, and the cutting speed is 1.98 times higher as compared to the results found for a
converging coaxial nozzle).

Keywords
Laser cutting, supersonic nozzle, rectangular nozzle, aluminium alloy, molten material

1. INTRODUCTION

Laser cutting was one of the first applications of lasers in the industry, and now is a consolidated
processing technique to cut sheet metal parts. The main laser cutting technique, called laser
fusion cutting, is based on the melting, and removal of molten material by an auxiliary high-
pressure gas jet [1]. This technique involves the localized melting, and partial evaporation of the

2
workpiece throughout in depth, and the simultaneous injection of an assist gas jet in order to
remove the molten material [2, 3]. The coupling of the laser beam energy with the workpiece,
and the removal of the molten material determine the success of this technique. However,
several works pointed out that the limiting factor in this process is fundamentally the removal
of molten material by the assist gas [4, 5]. Indeed, a correct removal of molten material is
required to maximize the cut speed but also the quality.

Theoretical analyses performed by Vicanek et al. [6] showed that the removal of molten material
by the assist gas jet is a consequence of two main actions: the shear stress, and the pressure
gradient exerted by the assist gas on the molten material along the cut front. It was found that
both factors are of the same order of magnitude, and they increase with the gas velocity.
Consequently, the improvement of the cut quality requires their maximization. This is especially
important when cutting certain materials. Ceramic materials or glasses are prone to cracking
due to thermal stresses consequence of the poor removal of molten material [7, 8]. In this case,
the high viscosity of the melt makes the removal of the molten material very hard, and cut edges
with dross, resolidified material, and a large heat affected zone (HAZ) are obtained [9]. For other
materials, the final application of the workpiece demands the maximization of the removal of
molten material to obtain high cut quality, and/or a reduced HAZ. In this sense, cutting
composites is a complex task due to the large HAZ typically formed in laser cutting [10]. Cutting
aluminium alloys for aerospace applications demands cuts with high quality and an extremely
reduced HAZ to meet the exigent mechanical performance of the workpiece [11, 12]. Therefore,
an improvement of the current assist gas injection systems, or new developments are required
to enhance the current cut quality.

In this work, the main phenomena related with the interactions of the assist gas jet with the
workpiece are briefly reviewed. Then, a new assist gas injection system to overcome the current
limitations is proposed. Finally, this system is experimentally assessed by means of cutting trials,
and the results are compared with those obtained with two of the most commonly used assist
gas injection systems, i.e. converging (coaxial) nozzles, and supersonic axisymmetric nozzles.

2. AERODYNAMIC INTERACTIONS DURING LASER FUSION CUTTING

2.1 Cutting with converging nozzles

High-pressure inert gas laser cutting partially alleviates the previously cited limitations in cutting
difficult materials, namely, presence of dross or resolidified material, and a large HAZ; however,

3
several phenomena related to the compressibility of the assist gas degrade the jet quality, and
then, the removal capacity. These prevent the application of this approach to cut materials for
mechanically demanding applications where an outstanding cut quality is required [13]. The root
of these limitations arises due to the structure of the jets expelled by the converging nozzles
(commonly used in laser cutting), and they are aggravated due to the interaction with the
workpiece.

The aerodynamic interactions of the assist gas jet during conventional high-pressure gas laser
cutting have been studied in the past by different authors [14-16]. In this injection system, a
converging (or also called sonic) nozzle, coaxial with the laser beam, and positioned close to the
workpiece is used to inject the assist gas into the cutting zone (see Fig. 1). The assist gas is
normally supplied at pressure ratios (or underexpansion ratios, i.e. the ratio between the static
pressure P in the flow just at the nozzle exit, and the ambient static pressure P0) P/P0>1.89, which
produce underexpanded jets [17]. This induces a severe shock wave pattern in the jet;
decelerates the flow, and reduces both the shear stress, and the gradient pressure along the cut
front. Then, the removal of molten material is severely compromised. On the other hand, the
normal impingement of the jet onto the surface of the workpiece reinforces this shock wave
pattern, and a normal shock wave, called Mach Shock Disk (MSD), stands over the inlet kerf [18].
The strength of this shock depends on the assist pressure, and the nozzle-workpiece distance
(also called stand-off distance) [19].

The MSD reduces even more the kinetic energy of the jet available to remove molten material
from the cut kerf. Furthermore, the static pressure of the gas increases after the MSD, produces
an underexpansion of the jet, and the generation of expansion waves just into the inlet kerf.
These waves, after being reflected in the jet boundaries, form compression waves which can
build up a single stronger branch. The impact of this compression oblique shock wave on the cut
front seriously disturbs the flow of the assist gas over the cut front. The laminar boundary layer
formed by the assist gas on the molten material can be separated due to the adverse gradient
pressure imposed by the interaction with this compression wave. The assist gas flow
downstream the separation is unsteady, chaotic, and transformed into a turbulent regimen;
therefore, decreasing the capacity of the gas to remove molten material from the kerf [20]. The
boundary layer separation (BLS) is easily detected in the cut edge because a rough surface finish,
recast material, and strong dross attachment in the lower cut surface is observed [21].
Furthermore, some swirls can be formed after the boundary layer separation [22-24]. These can
trap ambient air which can penetrate into the cut kerf, and react with the molten material. The
oxygen from the atmosphere can produce an undesirable oxidation of the cut front during laser

4
cutting using inert gases. On the contrary, nitrogen or other impurities from the atmosphere can
also reduce the cutting performance during laser cutting using oxygen because of the reduction
in the gas purity [25, 26].

These phenomena are also aggravated due to the strong choking suffered by the jet in the
entrance of the cut kerf. La Rocca et al. estimated the area blockage of the jet to be around 90%
[27]. This promotes a strong turbulence, and the formation of a stable vortex ring just at the
inlet kerf [14, 28]. These swirls can produce the contamination, and reduction in purity of the
assist gas due to the suction of ambient gases. This contamination, and the higher residence
times of the gas can also promote the formation of plasma [29].

2.2. Supersonic and off-axial nozzles.

Several strategies were followed to solve the aforementioned problems. The utilization of
supersonic (also called converging-diverging or De Laval) nozzles were proposed to obtain jets
free of shock waves, and to avoid the energy dissipation [30-33]. These are axisymmetric (i.e.
with circular cross-section), and have a converging-diverging internal geometry. They supply a
high velocity assists gas jet, but avoiding the reduction in kinetic energy because the jet is absent
of shock waves. Cutting experiments with supersonic axisymmetric nozzles performed on
several materials, ranging from mild steel to wood, detected a remarkable improvement in the
cut quality, and processing speed [31, 32, 34].

On the other hand, off-axial nozzles, supplying the assist gas jet tilted certain angle with regard
to the laser beam, were proposed to avoid the problem of the choking. In this way, the area
blockage of the jet is greatly reduced, because more gas can enter into the cut kerf. Ketting et
al. [35] found the enlargement in the parameter range leading to dross-free cuts during high-
pressure laser cutting of an Al-Mg aluminium alloy when the laser beam is placed in front of
centre of nozzle as compared to on-axis cutting.

Quintero et al. merged the precedent approaches using an off-axial supersonic nozzle. Using this
injection system, dross free cuts were obtained during Nd:YAG, and CO2 laser cutting of thick
sectioned mullite-alumina workpieces [36, 37]. Using the same experimental setup, Riveiro et
al. confirmed the superior performance of this assist gas injection system during the CO2 laser

5
cutting of 2024-T3 and 7075-T6 aluminium alloys [11, 38]. Dross-free cuts, and smooth cut edges
were obtained (with an average roughness close to Ra=1 m) [39]. Also a high processing rate,
and a noticeable reduction of the HAZ, as compared to samples processed by means of a cutting
head assisted with a converging nozzle, were observed.

3. IMPROVEMENT OF THE CUT QUALITY. NEW ASSIST GAS INJECTION SYSTEM

Nozzles used for laser cutting have an axisymmetric geometry (i.e. with a circular cross-section),
and the typical exit diameter ranges from 1 to 3 mm. Therefore, they produce jets with a circular
geometry in a millimetric range. On the other hand, the shape of the kerf is approximately
rectangular, and with a typical width as small as tens of microns. In consequence, the partial
blockage of the jet is unavoidable. Several authors have proposed the utilization of microjets
with dimensions similar to the kerf width to avoid the blockage of the jet [40, 41]. The main
drawback of this approach is the low mass flow entering into the kerf [41]. This reduction in
mass flow decreases the removal of molten material, and the cut quality is compromised.

It seems reasonable that the most suitable shape of the assist gas jet should be rectangular to
avoid the precedent drawbacks. In this way, the jet can naturally adapt to the geometry of the
cut kerf, and the choking would be even smaller. Furthermore, the jet can be thin in one side (to
mimic the kerf width), but large on the other side to avoid the reduction in mass flow (see Fig.
2). In this regard, Son et al. studied the impingement characteristics of jets from off-axial
supersonic rectangular nozzles with a simulated cut kerf using the Schlieren technique, and Pitot
pressure measurements [42]. They compare the results with those previously observed for
converging coaxial nozzles. They found a marked reduction in the strength of the MSD, and
higher Pitot pressures on the kerf surface. These results suggest a favourable aerodynamic
interaction of this kind of jets during laser cutting.

According to these considerations, we developed a cutting head with a supersonic nozzle with a
rectangular cross-section (i.e. non-axisymmetric) to inject the assist gas into the cut kerf. Figure
2 depicts a schematic drawing of this experimental setup. The supersonic rectangular nozzle was
designed in accordance with the basic principles of the steady one-dimensional compressible
flow with friction. There are three main positioning parameters for the jet, as depicted in Fig. 2.
These are: the distance between the cut front and the impinging point of the gas jet (X), the
stand-off distance from the supersonic nozzle to the workpiece (Z), and the angle between the
laser beam axis and the axis of the gas jet (). They were adjusted by means of cutting

6
experiments to obtain the higher optimum removal of molten material by the gas jet. The results
have been compared with those obtained using a converging nozzle (coaxial with the laser beam
and with axisymmetric geometry), and an off-axial supersonic nozzle with a circular cross-
section (i.e. with axisymmetric geometry).

4. MATERIALS AND METHODS

4.1 Laser processing parameters

The performance of the new assist gas injection system was determined by means of cutting
trials. The base material used in this study was a 3 mm thick Al-4.5%Cu alloy (2024-T3). This is
one of the 2XXX-series heat treatable aluminium alloys, and it is typically used for structural
applications in aircraft construction (such as in fuselage structures, wing tension members, shear
webs and ribs, and structural areas where stiffness, fatigue performance and good strength are
required). The excellent properties of this material rely on the heat treatment called age or
precipitation hardening, which produces finely dispersed precipitates. This alloy is susceptible
to hot cracking.

Cutting experiments were performed with a 3.5 kW CO2 slab laser (Rofin-Sinar), the laser mode
being a TEM00. The cavity of the laser source was isolated with a beam bender mirror with a
multilayer coating to absorb the backreflected laser radiation due to the high reflectivity of the
aluminium to this laser wavelength. This laser source delivers a maximum output power of 3.5
kW in continuous wave (CW) mode at a wavelength of 10.6 m. The laser power used during
the experiments ranged from 1.0 kW to 3.5 kW (in CW mode). The cutting speed was varied
from 100 mm/min to 15000 mm/min. In all the experiments the laser beam was focalized with
a ZnSe lens, with f=127 mm in focal length. The focus was positioned just onto the surface of the
workpiece. The spot size, and the Rayleigh length were 110 m (in diameter), and 1.48 mm,
respectively.

4.2. Assist gas processing parameters

A supersonic rectangular nozzle was used to inject the assist gas during the cutting experiments.
This was designed to obtain fully expanded jets under the assumption of one-dimensional steady
compressible flow. The design pressure (P), Mach number (Me), and main geometrical

7
characteristics of the nozzle are summarized in Figure 3. This nozzle operates at a Mach number
of Me=1.97, and a fixed pressure ratio of P/P0=8 (P0=atmospheric pressure). The stand-off
distance (Z) was selected by cutting trials to ensure that the supersonic core of the jet is
impinging on the cut front, and the spreading of the jet is not excessive. Moreover, the
impingement point (X) and incidence angle () were chosen to avoid choking, and shock
formation into the kerf which can induce the boundary layer separation (BLS) [43].

The cutting results obtained by means of this new injection system were compared with those
obtained with a cutting head with a converging coaxial nozzle, and with a cutting head
incorporating an off-axial supersonic axisymmetric nozzle. The assist gas processing parameters
for both systems are summarized in Table 1. The nozzle exit diameter of the converging nozzle
was D=2 mm, and the stand-off distance was adjusted to Z=1.5 mm. The supplying pressure was
P=8 bar in all the experiments involving the converging coaxial nozzle. On the other hand, the
nozzle exit diameter in the supersonic axisymmetric nozzle was D=1.7 mm, the nozzle angle
W=55, and the stand-off distance was Z=4 mm in all the experiments. This nozzle was designed
to operate at a Mach number of Me=1.97, and a fixed pressure ratio of P/P0=8 (P0=atmospheric
pressure); therefore, the supplying pressure was also P=8 bar.

Argon 4.8 (99.998 % purity) was used as assist gas in all the assist gas injection systems.

4.3. Sample characterization

The maximum cutting speed, and cut quality (measured in terms of the average roughness of
the cut edge, and extension of the HAZ) were determined after the cutting trials.

The cutting edge of selected samples were characterized by means of and optical stereoscopic
microscope (Nikon SMZ-10A) with a photographic system to record and store the images.
Moreover, the cross-sectional view to the edge of the kerf was characterized by means of
scanning electron microscopy (Phillips XL30).

The roughness was measured by means of a roughness meter (Taylor-Hobson Form Talysurf
Plus) in accordance with the recommendations specified in the International Standard ISO
4288:1996. A Gaussian filter corrected in phase in three locations of cut edges (upper, middle
and lower part) was applied. Then, an average value was extracted from the data to characterize
the finishing of cut edges.

8
5. RESULTS

5.1. Influence of the assist gas positioning parameters on the processing with the supersonic
rectangular nozzle

The performance of the cutting head with the supersonic rectangular nozzle depends on three
positioning parameters: the impingement point (X), the incidence angle (), and the stand-off
distance (Z). Several laser cutting trials were performed to determine their most suitable value.
This process was performed by varying only one factor at a time, while keeping others fixed;
therefore, interactions between factors were discarded to simplify the analysis.

First, the influence of inclination for the assist gas jet (processing parameters, and experimental
results are summarized in Table 2) was analysed. This parameter has a clear influence on the
cutting results, as depicted in Fig. 4. Cutting speeds and qualities in a range of 20% the
maximum cutting speed, minimum roughness, and minimum HAZ is obtained for nozzle angles
around =50-55 (see Figs. 4a-c). No separation of the assist gas flow along the cut edge was
detected under this range, as depicted in Fig. 4d. Also, clinging dross is avoided.

The cutting trials also revealed a strong influence of the impingement point on the cut quality
(processing parameters and experimental results are summarized in Table 2). Cutting speeds
and qualities in a range of 20% the maximum cutting speed, minimum roughness, and
minimum HAZ is obtained for a distance between the cut front and the impinging point of the
gas jet around X=3.5 mm (see Figs. 5a-c).

Some results of these tests are presented in Fig. 5d, showing some images of the cut edge of the
as made cuts. It is obvious that the difference among them lies in the recast layer adhered to
the cut edge. Recast material covering almost the whole cut edge is clearly noticeable for small
distances (X=0.5 mm), whereas the amount of this recast layer is gradually decreasing as the X
distance increases. No recast layer is observed for distances around X=3.5 mm. For higher
distances, the recast layer is again observed. Quintero et al. observed the same trends during
the processing with supersonic axisymmetric nozzles [44]. They related this behaviour with a
change in the gas flow inside the cut kerf from turbulent into laminar. For small X distances
(X=0.5 mm), the jet impinges on the top corner of the cut front. This deflection of the jet induces
the formation of a steady detached shock wave, and provokes a strong wave pattern at the inlet
kerf leading to a quick detachment of the gas flow. This strongly reduces the capacity of the jet

9
to remove molten material. As the X distance increases, this detached shock wave disappears
and the gas flow is deflected by the cut front through an oblique shock wave. This produces a
weak compression of the flow, and a reduction in velocity. However, if this oblique shock wave
impinges on the cut front provokes a reversal of the pressure gradient which leads to the
boundary layer separation (BLS), and its change into a turbulent flow. For a sufficiently long X
distance (in the current system X=3.5 mm), this oblique shock wave does not impinge on the cut
front. In this case, a laminar flow is fully developed along the cut front, sweeping away the
molten material. This is corroborated by the absence of the recast layer in the sample presented
in Fig. 5d, and corresponding to X=3.5 mm. However, these explanations should be corroborated
with the visualization of the assist gas jet, for example, using Schlieren or shadowgraph
techniques.

The stand-off distance does not significantly affects the maximum cutting speed as depicted in
Fig 6a; however, it has certain influence on the quality of the cut edge (processing parameters
and experimental results are summarized in Table 2). The increment of this parameter makes
the cut edge rougher (see Fig. 6b). This trend can be explained taking into account the evolution
of the cross-sectional shape of the jet with the distance, as depicted in Fig. 7. The cross-sectional
shape was measured using a pressure sensitive film (UltraLow Pressurex film from Sensor
Products) for distances from the nozzle exit in a range from 1 to 10 mm. As expected, the jet
evolves from a rectangular to a circular cross-sectional shape. This finding is in accordance with
the results found in the literature. The work performed by Seiner [45] showed that the length
of the core is shorter in rectangular jets as compared to equivalent circular jets. Therefore, the
utilization of supersonic rectangular nozzles restrict the working distance as compared to
axisymmetric nozzles. Furthermore, several authors experimentally confirmed the higher spread
of the rectangular jets as compared to those with an axial symmetry, in particular for the
supersonic ones [46-52]. The spreading is particularly observed in the shorter axis of the jet,
which produces an evolution of the jet shape from a rectangular to a circular profile [53, 54].
Separation of the nozzle must not be excessive to avoid a large divergence of the jet, and
reducing the choking in the inlet kerf. Stand-off distances around Z=2-3 mm are in a range of
20 % the maximum cutting speed, and minimum roughness (see Figs. 6a-b). This range is good
enough to avoid the large spread of the jet, but also to protect the nozzle from splashes, or back
reflections which could damage it.

10
The preceding analyses revealed a suitable range of values of the positioning parameters for the
supersonic rectangular nozzle. Cutting speeds and qualities in a range of 20% the maximum
cutting speed, minimum roughness, and minimum HAZ is obtained for the values of the
positioning parameters summarized in Table 3. Taking into these results, we selected a nozzle
angle inclination of =50, an impingement point of the jet of X=3.5 mm, and a stand-off
distance of Z=3 mm, as summarized in the last column of Table 3. These were used in the cutting
trials performed to evaluate the performance of this assist gas injection system as compared to
the processing with a converging (coaxial) nozzle, and a supersonic axisymmetric nozzle.

5.2 Comparison of the performance of assist gas injection systems

The performance of the new assist gas injection system was assessed in comparison to the
processing with a converging (conventional) nozzle, and a supersonic axisymmetric nozzle.

The cutting trials revealed that the maximum cutting speed attained with the supersonic
rectangular nozzle is smaller than that achieved with the supersonic axisymmetric one (see Fig.
8). In both cases, the cutting speed is higher than that obtained with a converging nozzle working
at the same assist gas pressure (P=8 bar). In particular, the cutting speed for the supersonic
rectangular nozzle is 1.98 times higher than for the converging nozzle for a laser power of 3000
W.

The cut quality is greatly improved with the new assist gas injection system. Both supersonic
arrangements give cuts with a roughness greatly reduced as compared to conventional systems
(see Fig. 9). The average roughness of the cut edge reached a minimum value of Ra=1 um for the
new assist gas injection system, i.e. 6.5 times lower than for the converging nozzle for the same
processing conditions (see Fig. 10). This result is a clear evidence of the improvement of the
assist gas performance.

In contrast with samples processed with the converging nozzle, the roughness of samples
processed with both supersonic nozzles (axisymmetric and rectangular) tends to decrease with
the laser power (see Fig. 10). Boundary layer separation (BLS) is not detected in the cut edge of
samples processed with both supersonic nozzles (see Figs. 9b,c). On the contrary, the BLS is

11
observed in the lower part of the cut edge obtained with the converging nozzle (see Fig. 9a). In
this case, the inclination of the striations is higher after the BLS, and the cut edge becomes
rougher.

These results can be explained taking into account the flow of molten material along the cut
front. Using a similar arrangement as that described in Ref. [55], the influence of the assist gas
injection system on the flow of molten material during laser cutting was visualized. Laser cutting
of glass plates, 6 mm thick, was recorded from a lateral perspective with a high-speed camera
(Photron FASTCAM-1024 PCI) to observe the cut front, cut edge, and the flow of molten material
(see Fig. 11). The molten material (bright areas) is observed to flow along the cut front, and cut
edge during the processing with the converging nozzle. Moreover, the formation of the MSD is
clearly identified in the inlet kerf. This normal shock seriously degrades the assist gas jet. On the
other hand, the detachment of the boundary layer (BLS) avoids the confinement of the molten
material over the cut front. Then, this flows along the cut edge. Part of this material is not
entirely removed by the gas, and will remain adhered to the cut edge after the cutting process.
This explains the higher roughness during the processing with the converging nozzle. On the
contrary, the molten material is confined on the cut front for the process assisted with the
supersonic axisymmetric nozzle. This confinement avoids the flow of molten material along the
cut edge; then, no recast layer is formed, and a superior cut quality is obtained. In the case of
the supersonic rectangular nozzle, the confinement is maximized (due to the normal
impingement of the jet over the whole cut front) and no shock phenomena are observed. A
lower temperature in the cut edge is expected in this case; however, pyrometry, or high-speed
thermal imaging should confirm this hypothesis. This fact is extremely relevant for the
processing of heat sensitive materials. The lower temperatures would also explain the slight
reduction in cutting speed observed in the Fig. 8. This is the first time that the phenomenon of
confinement has been discussed in the literature, and further investigations are required to
elucidate its impact on the laser cutting process.

The influence of the assist gas injection system was also assessed from a metallurgical point of
view. The inspection of the cross-section to the cut edge, as depicted in Fig. 12, reveals that the
microstructure is almost unaltered in the case of the processing with both supersonic nozzles.
On the other hand, a clear grain refinement is observed in the cut edge of the samples processed
with the converging nozzle. This indicates a higher heat input into the cut edge. Thus, the
mechanical properties will be largely reduced in this region due to the dissolution and

12
coarsening of the precipitates that provide the excellent mechanical properties to this
aluminium alloy.

The extension of the thermal input into the cut edges was also estimated in terms of the
intermetallic particles. These are the S-phase (rich in aluminium, copper and magnesium), the
-phase (rich in aluminium and copper), and the -phase (rich in aluminium, copper, iron,
manganese and silicon). These appear as bright features in backscattered electron micrographs
and are characterized by a low melting point. For example, the melting point of the -phase was
estimated to be around 500 [56]. Then, they provide an excellent marker to determine the
heat penetration depth in the cut edge. The processing assisted with the converging nozzle (see
Fig. 12a) removes these particles in the first 70 m of material. This value is reduced up to 5 m
for the samples processed with the supersonic axisymmetric nozzle (see Fig. 12b). However,
these particles remain unaltered for locations adjacent to the cut edge in the samples processed
with the supersonic rectangular nozzle (see Fig. 12c). Indeed, some particles at 1.5 m from the
cut edge were found in those samples processed with this assist gas injection system. This
confirms that the processing assisted with a supersonic rectangular nozzle does not modifies the
microstructure, and mechanical properties of the material in the cut edge.

).

These results are an indirect proof of the lower temperatures reached in the cut edge with the
supersonic rectangular nozzle. They also suggest that this new assist gas injection system is the
most suitable system to cut materials prone to be metallurgical affected by the laser cutting
process, or to achieve an outstanding cut quality required in mechanically demanding
applications.

6. CONCLUSIONS

The analysis of the aerodynamic interactions of the assist gas jet with the cut front suggests a
way to improve the current laser cutting window in terms of cut quality, and processing speed.
In this paper, a new assist gas injection system based on the utilization of an off-axial supersonic
rectangular nozzle has been proposed. This assist gas system has been designed taking into
account the aerodynamic interactions of the assist gas with the workpiece. The results were
compared with those obtained with the most commonly used assist gas injection systems: a
converging (coaxial) nozzle, and an off-axial supersonic axisymmetric nozzle.

13
The characteristics of the jet exhausted by these nozzles must be considered to optimize the
cutting results. The higher spreading of the jet, and the reduced length of the core restricts the
useful stand-off distance of the nozzle up to Z3 mm. Furthermore, the results point out a clear
necessity of adjustment of the impingement point of the jet, and the inclination of the nozzle
with regard to the laser beam. A proper selection of these positioning parameters was noticed
to be extremely relevant to guarantee a laminar flow along the whole cut front, and to avoid the
boundary layer separation. The most suitable value for these parameters was observed to be
W=50 and X=3.5 mm. These values maximize cutting speed, and quality. This new assist gas
injection system makes possible to obtain dross free cuts, with reduced roughness, and
negligible HAZ. Furthermore, the microstructure of the workpiece remains unaltered. These
results are in marked contrast to those obtained with a cutting head with a converging nozzle.
The explanation of the better performance of the new assist gas injection system can be
explained by the higher confinement of the molten material on the cut front, and the most likely
lower temperatures reached in the cut edge during the processing.

Acknowledgments

The authors wish to thank the technical staff from CACTI (University of Vigo) for their help with
sample characterization. This work was partially supported by the Government of Spain
(MAT2015-71459-C2-P), and by Xunta de Galicia (ED431B 2016/042, and Plan I2C Grant Program
POS-A/2013/161, ED481D 2017/010, and ED481B 2016/047-0).

14
References

[1] Steen WM. Laser material processing. 3rd ed. ed. London ; New York: Springer, 2003.
[2] Ready JF, Farson DF. LIA handbook of laser materials processing. Orlando, FL: Laser Institute
of America ; Magnolia Pub., 2001.
[3] Caristan CL. Laser cutting guide for manufacturing. Dearborn, MI: Society of Manufacturing
Engineers, 2004.
[4] Olsen FO. Fundamental mechanisms of cutting front formation in laser cutting. Laser
Materials Processing: Industrial and Microelectronics Applications. 1 ed. Vienna, Austria:
SPIE, 1994. p. 402-13.
[5] Olsen FO. An evaluation of the cutting potential of different types of high power lasers.
Proceedings 25nd International Congress on Application of Lasers and Electro-Optics,
ICALEO 2006. Scottsdale: Laser Institute of America, 2006. p. 188-96.
[6] Vicanek M, Simon G. Momentum and heat transfer of an inert gas jet to the melt in laser
cutting. Journal of Physics D: Applied Physics. 1987;20:1191-6.
[7] Quintero F, Pou J, Lusquios F, Boutinguiza M, Soto R, Prez-Amor M. Quantitative evaluation
of the quality of the cuts performed on mullite-alumina by Nd:YAG laser. Optics and Lasers
in Engineering. 2004;42:327-40.
[8] Yang LJ, Wang Y, Tian ZG, Cai N. YAG laser cutting soda-lime glass with controlled fracture
and volumetric heat absorption. International Journal of Machine Tools and Manufacture.
2010;50:849-59.
[9] Sun S, Brandt M. Laser beam machining. Nontraditional Machining Processes: Research
Advances2013. p. 35-96.
[10] Cenna AA, Mathew P. Evaluation of cut quality of fibre-reinforced plastics - A review.
International Journal of Machine Tools and Manufacture. 1997;37:723-36.
[11] Riveiro A, Quintero F, Lusquios F, Pou J, Prez-Amor M. Laser cutting of 2024-T3 aeronautic
aluminum alloy. Journal of Laser Applications. 2008;20:230-5.
[12] Chen C, Gao M, Zeng X. Relationship between temperature at cut front edge and kerf quality
in fiber laser cutting of AlCu aluminum alloy. International Journal of Machine Tools and
Manufacture. 2016;109:58-64.
[13] Riveiro A, Quintero F, Pou J. Laser fusion cutting of difficult materials. In: Lawrence J, Low
DKY, Pou J, Toyserkani E, editors. Advances in Laser Materials Processing Technology: CRC
Press & Woodhead Publishing 2010.

15
[14] Fieret J, Terry MJ, Ward BA. Aerodynamic interactions during laser cutting. Proceedings of
SPIE-The International Society for Optical Engineering. Quebec: SPIE, 1986. p. 53-62.
[15] Fieret J, Terry MJ, Ward BA. Overview of flow dynamics in gas-assisted laser cutting. Proc
SPIE. 1987;801:243-50.
[16] Zefferer H, Petring D, Aachen E. Investigations of the gas flow in laser beam cutting. DVS-
BERICHT. 1991;135:210-4.
[17] Liepmann HW, Roshko A. Elements of gasdynamics. New York: John Wiley & Sons, 1957.
[18] Kamalu JN, Steen WM. The importance of gas flow parameters in laser cutting. The
Metallurgical Society of AIME, 1981.
[19] Chen K, Yao YL, Modi V. Gas jet-workpiece interactions in laser machining. Journal of
Manufacturing Science and Engineering, Transactions of the ASME. 2000;122:429-38.
[20] Leidinger D, Penz A, Schucker D, Deinzer G, Geiger M, Hnsel A, et al. Nozzle design and
simulation of gas flow for the laser cutting process. Proceedings of SPIE, 1994. p. 469-79.
[21] Wandera C, Kujanpaa V. Characterization of the melt removal rate in laser cutting of thick-
section stainless steel. Journal of Laser Applications. 2010;22:62-70.
[22] Horisawa H, Fushimi T, Takasaki T, Yamaguchi S. Flow characterization in a laser cut kerf.
CLEO, Pacific Rim'99: IEEE, 1999. p. 358-9.
[23] Horisawa H. Gas flow characteristics in a laser cut kerf. Initiatives of precision engineering
at the beginning of a millenium: 10th International conference on precision engineering
(ICPE): Springer, 2001.
[24] Kovalev OB, Yudin PV, Zaitsev AV. Formation of a vortex flow at the laser cutting of sheet
metal with low pressure of assisting gas. Journal of Physics D: Applied Physics.
2008;41:155112.
[25] Powell J, Ivarson A, Kamalu J, Borden G, Magnusson C. The role of oxygen purity in laser
cutting of mild steel. Proceedings of ICALEO '92: Laser Institute of America, 1992. p. 433-
42.
[26] Ivarson A, Powell J, Broden G, Kamalu J, Magnusson C. The effects of oxygen purity in laser
cutting mild steel: a theoretical and experimental investigations. Lasers in Engineering.
1993;1:299.
[27] LaRocca AV. Nozzle design to control fluid-dynamic effects in laser cutting. SPIE.
1994;2207:169-80.
[28] Fieret J, Ward BA. Circular and noncircular nozzle exits for supersonic gas jet assist in CO2
laser cutting. Third Conference on Lasers in Manufacturing. 1986:1-11.
[29] Grigoryantis AG. Basics of laser material processing. Boca Raton: Mir Publishers ; CRC Press,
1994.

16
[30] Thomassen FB, Olsen FO. Experimental studies in nozzle design for laser cutting.
Proceedings of the 1st International Conference on Lasers in Manufacturing. Brighton, East
Sussex, Engl: IFS (Publ) Ltd, 1983. p. 169-80.
[31] Mukherjee K, Grendzwell T, Khan PAA, McMillin CW. Gas flow parameters in laser cutting
of wood - nozzle design. Journal of Forest Products. 1990;40:39-42.
[32] Semrau H, Tonshoff HK. Laser gas cutting using Laval nozzles. Proc SPIE. 1990;1988:449-58.
[33] Man HC, Duan J, Yue TM. Dynamic characteristics of gas jets from subsonic and supersonic
nozzles for high pressure gas laser cutting. Optics and Laser Technology. 1998;30:497-509.
[34] Boutinguiza M, Pou J, Lusquios F, Quintero F, Soto R, Prez-Amor M, et al. CO2 laser cutting
of slate. Optics and Lasers in Engineering. 2002;37:15-25.
[35] Ketting HO, Olsen F. High pressure off-axis laser cutting of stainless steel and aluminium.
Proceedings of LAMP'92. Nagaoka1992. p. 607-12.
[36] Quintero F, Pou J, Lusquios F, Boutinguiza M, Soto R, Prez-Amor M. Comparative study of
the influence of the gas injection system on the Nd:yttrium-aluminum-garnet laser cutting
of advanced oxide ceramics. Review of Scientific Instruments. 2003;74:4199-205.
[37] Quintero F, Pou J, Lusquios F, Boutinguiza M, Soto R, Prez-Amor M, et al. Comprehensive
assessment of the CO2 laser cut quality of ceramics with different assist gas injection
systems. Journal of Laser Applications. 2004;16:212-20.
[38] Riveiro A, Quintero F, Lusquios F, Comesaa R, Pou J. Effects of processing parameters on
laser cutting of aluminium-copper alloys using off-axial supersonic nozzles. Applied Surface
Science. 2011;257:5393-7.
[39] Riveiro A, Quintero F, Comesaa R, Del Val J, Boutinguiza M, Lusquios F, et al. The influence
of the assist gas flow regime on the laser cutting of aeronautic aluminium alloys. 4th Pacific
International Conference on Applications of Lasers and Optics, PICALO 2010. Wuhan,
China2010.
[40] Min SK, Yu DO, Lee Y, Cheong JS. Study of micro-supersonic impinging jets and its application
to the laser machining. Transactions of the Korean Society of Mechanical Engineers, B.
2009;33:93-100.
[41] Jackson MJ, Robinson GM, Gill MDH, ONeill W. The effect of nozzle design on laser micro-
machining of M2 tool steels. Journal of Materials Processing Technology. 2005;160:198-
212.
[42] Son S-H, Jun D-Y, Lee Y. Experimental Study of Characteristics of Assist Gas in Laser
Machining Using Supersonic Rectangular Nozzle. Transactions of the Korean Society of
Mechanical Engineers B. 2012;36:233-40.

17
[43] Brandt AD, Settles GS. Effect of nozzle orientation on the gas dynamics of inert-gas laser
cutting of mild steel. Journal of Laser Applications. 1997;9:269-77.
[44] Quintero F, Pou J, Fernndez JL, Doval AF, Lusquios F, Boutinguiza M, et al. Optimization
of an off-axis nozzle for assist gas injection in laser fusion cutting. Optics and Lasers in
Engineering. 2006;44:1158-71.
[45] Seiner JM. Fluid Dynamics and Noise Emission Associated with Supersonic Jets. Studies in
Turbulence. New York: Springer-Verlag, 1992. p. 297-323.
[46] Trentacoste N, Sforza P. Further experimental results for three-dimensional free jets. AIAA
Journal. 1967;5:885-91.
[47] Sfeir AA. The velocity and temperature fields of rectangular jets. International Journal of
Heat and Mass Transfer. 1976;19:1289-97.
[48] Krothapalli A, Baganoff D, Karamcheti K. On the mixing of a rectangular jet. Journal of Fluid
Mechanics. 1981;107:201-20.
[49] Tsuchiya Y, Horikoshi C. On the spread of rectangular jets. Experiments in Fluids.
1986;4:197-204.
[50] Grinstein FF. Self-induced vortex ring dynamics in subsonic rectangular jets. Physics of
Fluids. 1995;7:2519-21.
[51] Zaman KBMQ. Spreading characteristics of compressible jets from nozzles of various
geometries. Journal of Fluid Mechanics. 1999;383:197-228.
[52] Mohamed A, Hamed A, Lehnig T. Supersonic Rectangular Over-expanded Jets of Single and
Two-phase Flows. Proceedings of the 19th ISABE2003.
[53] Von Glahn UH. On some flow characteristics of conventional and excited jets. NASA
Technical Memorandum1984.
[54] Von Glahn UH. Rectangular nozzle plume velocity modelling for use in jet noise prediction.
Proceedings of the 25th Joint Propulsion Conference. 1989.
[55] Riveiro A, Quintero F, Lusquios F, Comesaa R, Pou J. Study of melt flow dynamics and
influence on quality for CO2 laser fusion cutting. Journal of Physics D: Applied Physics.
2011;44.
[56] Reiso O, verlie H-G, Ryum N. Dissolution and melting of secondary Al2Cu phase particles
in an AlCu alloy. Metallurgical Transactions A. 1990;21:1689-95.

18
Figure 1. Aerodynamic interactions during laser cutting using a converging (sonic) coaxial nozzle
at underexpansion ratios P/P0>1.89.

Figure 2. Schematic of the new assist gas injection system. (Positioning parameters: X - distance
between the cut front and the impinging point of the gas jet, Z - stand-off distance from the
supersonic nozzle to the workpiece, - angle between the laser beam axis and the axis of the
gas jet).

19
Figure 3. Design and operating parameters for the supersonic rectangular nozzle (P: jet exit static
pressure, P0: ambient static pressure, P/ P0: pressure ratio, Me: Mach number at the nozzle exit,
Ae/At: exit-to-throat area ratio, and De: width of the jet in the nozzle exit).

3000 20
Maximum cutting speed (mm/min)

18
2500
16
14
2000
12
Ra (m)

1500 10
8
1000
6
4
500
2
0 0
35 40 45 50 55 60 65 70 75 35 40 45 50 55 60 65 70 75

- Nozzle angle () - Nozzle angle ()

a b

120000

100000
HAZ area (m2)

80000

60000

40000

20000 1 mm
0
35 40 45 50 55 60 65 70 75

- Nozzle angle ()

c d

Figure 4. a) Maximum cutting speed, b) average roughness of the cut edge (Ra), c) area of the
HAZ, and d) optical images of the cut edge as a function of the nozzle angle ().

20
 3000
20

Maximum cutting speed (mm/min) 2500 18

16
2000 14
12

Ra (m)
1500 10
8
1000
6
4
500
2
0 0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7

X - Impinging point (mm) X - Impinging point (mm)

a b

120000 X=0.5 mm X=3.5 mm X=6.5 mm

100000
HAZ area (um2)

80000

60000

40000 1 mm
20000

0
0 1 2 3 4 5

X - Impinging point (mm)

c d

Figure 5. a) Maximum cutting speed, b) average roughness of the cut edge (Ra), and c) area of
the HAZ, , and d) optical images of the cut edge as a function of the impingement point of the
jet (X).

3000 20
Maximum cutting speed (mm/min)

18
2500
16
14
2000
12
Ra (m)

1500 10
8
1000
6
4
500
2
0 0
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 10
Z - Stand-off (mm) Z - Stand-off (mm)

a b

Figure 6. a) Maximum cutting speed, and b) average roughness of the cut edge (Ra) as a function
of the stand-off distance (Z).

21
Figure 7. Cross-sectional shape of the jet at different locations downstream the nozzle exit (axial
distance from nozzle exit: A=2 mm, B=3 mm, C=4 mm, D=5 mm, 1E=6
mm
mm, F=7 mm).

12000 Sup. Rectangular Nozzle

Maximum cutting speed (mm/min)

11000 Sup. Axisymmetric Nozzle

10000 Converging Nozzle
9000
8000
7000
6000
5000
4000
3000
2000
1000
0
0 500 1000 1500 2000 2500 3000 3500
Laser power (W)

Figure 8. Comparison of the maximum cutting speed as a function of the laser power during the
processing with a converging nozzle, supersonic axisymmetric nozzle, and a supersonic
rectangular nozzle (assist gas pressure P=8 bar).

BLS

1 mm 1 mm 1 mm
a b c

Figure 9. Cut edges after the processing with the: a) converging nozzle, b) supersonic
axisymmetric nozzle, and c) supersonic rectangular nozzle (processing conditions corresponding
to the maximum cutting speed for each injection system for a laser power of 3000 W).

22
 10 Sup. Rectangular Nozzle
9 Sup. Axisymmetric Nozzle

8 Converging Nozzle

7
6
Ra (m)

5
4
3
2
1
0
0 500 1000 1500 2000 2500 3000 3500

Laser Power (W)

Figure 10. Comparison of the average roughness (Ra) as a function of the laser power during the
process with the converging nozzle, supersonic axisymmetric nozzle, and supersonic rectangular
nozzle.

Nozzle Cut
MSD edge

Cutting 1 mm
front
a b c

Figure 11. High-speed images of the cutting process of glass plates (thickness=6 mm) using a
cutting head with a a) converging nozzle, b) supersonic axisymmetric nozzle, and c) supersonic
rectangular nozzle. Yellow arrows approximately delimit the boundaries of the molten material.
(Processing conditions: Laser power=1400 W, cutting speed=1000 mm/min).

23
 Intermetallic
particle

a b

Intermetallic
particle

Figure 12. SEM micrographs of cross-sections of laser cut edges obtained with the a) converging
nozzle, the b) supersonic axisymmetric nozzle, and the c) supersonic rectangular nozzle
(processing conditions corresponding to the maximum cutting speed for each injection system
for a laser power of P=3000 W

24
Table 1. Assist gas processing parameters for the off-axial supersonic axisymmetric nozzle, and
for the converging coaxial nozzle used during the experiments.

Nozzle P (bar) D (mm) Me () Z (mm) Gas

Off-axial supersonic
8 1.7 1.97 55 4 Argon 4.8
axisymmetric nozzle
Converging coaxial
8 2 1 0 1.5 Argon 4.8
nozzle

25
Table 2. Influence of the positioning parameters (nozzle angle (), impingement point (X), and
the stand-off distance (Z)) on the cutting speed, and quality during the processing with the
supersonic rectangular nozzle.

Influence of the nozzle angle ()

P(W) () X(mm) Z(mm) Vc,max(mm/min) Ra(m) HAZ area (m2)
1500 40 6 3 1300 8.53 61473
1500 45 6 3 1400 8.01 56362
1500 50 6 3 1800 3.4 0
1500 55 6 3 2200 3.89 0
1500 60 6 3 1800 5.48 78534
1500 70 6 3 1400 7.27 98471
Influence of the impingement point (X)
P(W) () X(mm) Z(mm) Vc,max(mm/min) Ra(m) HAZ area (m2)
1500 50 0.5 3 1200 15.99 58291
1500 50 1.5 3 1400 12.6 42101
1500 50 2.5 3 1700 4.32 21059
1500 50 3.5 3 1800 3.40 22211
1500 50 4.5 3 2000 3.13 48144
1500 50 5.5 3 1600 4.51 -
1500 50 6.5 3 1000 6.92 -
Influence of the stand-off distance (Z)
P(W) () X(mm) Z(mm) Vc,max(mm/min) Ra(m) HAZ area (m2)
1500 50 6 2 1200 6.15 -
1500 50 6 3 1500 6.83 -
1500 50 6 5 1000 8.64 -
1500 50 6 7 1200 12.7 -

26
Table 3. Suitable assist gas positioning parameters for the supersonic rectangular nozzle.

20% max. Vc 20% min. Ra 20% min. HAZ Selected Value

() 50-60 50-55 50-55 50

X (mm) 2.5-5.5 3.5-4.5 2.5-3.5 3.5
Z (mm) 2-3 2-3 - 3

27