Experimental investigation on the effects of the gas mixtures
in plasma arc cutting of austenitic stainless steel
M Boselli2, L. Ceschini1,2 ,V Colombo1,2, E Ghedini1,2, M Gherardi1, F Rotundo2, P Sanibondi1,
S Dallavalle3, R Fazzioli3, M Vancini3
Alma Mater Studiorum-Università di Bologna
Department of Industrial Engineering (D.I.N.)
2
Industrial Research Centre for Advanced Mechanics and Materials (C.I.R.I.-M.A.M.)
V.le Risorgimento 2, 40136 Bologna, Italy
1
3
Cebora S.p.A., via Andrea Costa 24, 40057 Cadriano di Granarolo, Italy
e-mail: [email protected]
Abstract: The effects of different Ar/H2/N2 plasma gas mixtures on the performance of
stainless steel plasma cutting has been investigated experimentally using a plasma arc cutting
torch working at 400A. Results of the microstructural and compositional analysis of the kerf
surface using SEM - EDS show that the presence of N2 or Air in the plasma arc column can
affect the chemical composition and microstructural characteristics of the kerf surface,
influencing machinability and weldability.
Keywords: Plasma arc cutting, austenitic stainless steel cutting, Plasma cut quality.
1. Introduction
Plasma Arc Cutting (PAC) is a process for cutting
metallic sheets with thickness in the range of 1 - 160 mm
in which a thermal plasma arc is ignited between a
thermionic cathode that is enclosed in the plasma torch
and the workpiece which works as the anode [1, 2]. In the
last decades significant increase in the quality of the cut,
on the maximum cutting speed and in the reduction of the
wear rate for torch consumables have been achieved [3-5].
The development of high current PAC torches (from 400A
to 1600A) has further increased the cutting speed, leading
to the possibility of cutting thickness up to 50 mm at a
speed of 0.5 m/min for a 400A torch.
PAC technology is particularly competitive for
Stainless Steel (SS) cutting, since it does not rely on metal
oxidation for severing the workpiece, while oxyfuel (one
of the PAC main competitor) is limited only to steel and
other ferrous metals supporting oxidization. However the
extremely high temperatures reached during the PAC
process can affect the composition and microstructure of
the kerf surface and in turn the weldability and
machinability of the workpiece.
Surface quality evaluation is usually performed by the
end users by means of qualitative observation of the kerf
surface color and appearance and only few preliminary
studies have been published concerning the effect of
plasma cutting on stainless steels microstructure [6]. The
aim of this work was to correlate the external appearance
of the kerf surface to the results of more accurate
compositional and microstructural analyses. These will
include the measure of the remelted surface layer and the
evaluation of the effect of heating on material
microstructure and atomic element concentration.
It is well known that an important role on the
optimization of the PAC process is played by the choice
of the working gas mixtures that can be tuned depending
on the cutting material, thickness and operating current to
obtain the best performance. Since oxidation is not the
main contribute for the cutting, the plasma gas choices for
SS cutting relies mainly on non-oxidizing Ar/H2/N2
mixtures which have also the advantage to allow using
tungsten cathodes, greatly increasing consumables life
time with respect to the hafnium cathodes used in O2
cutting of mild steel. PAC vendors usually provide tables
of operating conditions for different gas mixture that
typically including the commercial standard H35 (65% Ar
- 35% H2), pure N2 and variable mixtures of H35 - N2. For
this reason, this work will focus only on combinations of
such mixtures and their effects on the kerf surface for
different plate thicknesses and cutting speeds.
2. Materials and methods
2.1 Material
The stainless steel used in the experimental work is the
austenitic AISI 304L steel, whose nominal composition is
reported in Table 1. This material is characterized by
excellent corrosion resistance as well as good weldability.
Tests have been performed on 15, 25 and 50 mm plate
thickness.
2.2 PAC operating conditions
PAC process has been performed using a prototypal
CEBORA plasma torch equipped with a gas mixing unit
able to provide arbitrary mixtures. All tests have been
performed at 400 A current using pure N2 as shielding gas.
Operating conditions are summarized in Table 2.
Table 1. Nominal chemical composition (wt.%) of the AISI
304L austenitic stainless steel.
C
Si
Mn
P
S
Cr
Ni
Co
N
0.018
0,32
1,61
0,029
0,001
18,3
8.1
0.13
Plasma gas
Shield gas
A
B
C
D
E
F
H35
H35 30% + N2 70%
H35 70% + N2 30%
H35
H35
H35 30% + N2 70%
N2
N2
N2
N2
N2
N2
AH_13
AN_13
AQ_13
G_13
C_13
Z_13
3. Results
0.07
Table 2. PAC operating conditions.
Sample
Metallographic samples were polished by means of grit
papers and diamond paste up to 1 µm, then chemically
etched with Beraha II reactant (50% HCl, 0.8-1.0 K2S2O5,
1% NH4F·HF) [7] for the observation of the cross sections
with optical microscopy (OM) and SEM.
Cutting speed Plate thickness
[mm/min]
[mm]
2000
15
2000
15
2000
15
1125
25
900
30
520
50
2.3 Kerf surface characterization
The microstructural characterization of the kerf surface
has been performed by means of multi-focal optical
microscopy, scanning electron microscopy (SEM) with
energy dispersive microprobe (EDS) for localized
elemental analysis. Cross sections of two specimens (A,
B) were cut and mounted in epoxy resin for
metallographic analysis.
A
3.1 Microstructural characterization
Depending on the different cutting conditions, the kerf
surface exhibited a variety of distinguishable colours (Fig.
1), related to the physical properties of surface oxide
layers [8]. It is known that the thin film interference effect
can influence the appearance of a semi-transparent oxide
coating when the coating thickness is on the order of the
wavelength of the light incident on a specimen (400-700
nm for visible light). In particular, constructive
interference is likely to occur when the oxide thickness is
equal to λ/4n where n is the index of refraction for the
oxide [9]. Apparently, the use of plasma cutting gas rich
in nitrogen gave rise to darker surfaces (Fig. 1 A and F).
Sample F, in particular, showed a completely dark
appearance, possibly related to the presence of a thicker
oxide layer also due to the lower cutting speed and
consequent prolonged heating of the kerf surface.
Low magnification SEM analyses of the surface
morphologies (Fig. 2) showed generally smooth surfaces,
to except for the case of sample F.
A
B
B
C
C
D
F
E
F
Fig. 1 Macrographs (by multifocal optical microscopy)
showing portions of the surface of AISI 304L samples cut at
different operating parameters.
Fig. 2 SEM micrographs showing the surface of different AISI
304L samples cut at different operating parameters (according
to Table 2).
(about 1 µm) oxide layer could form either during cutting,
due to the entrance of atmospheric air in the arc column,
or during the cooling phase of the kerf surface.
Table 3. Surface composition (by large area EDS analyses) of
AISI 304L samples cut at different operating parameters.
Sample
Element (wt. %)
A
B
C
F
C
-
8.4
7
8.2
N
-
6.1
-
-
O
13.4
5.4
13
31.3
Si
1.0
0.6
0.4
0.2
Cr
16.3
16.2
15.6
13.3
Mn
2.2
1.8
1.9
6.9
Fe
59.6
54.7
55.7
37.7
Ni
7.5
1.5
6.7
6.4
Other
-
0.8
0.2
0.2
Fig. 3 AISI 304L base material microstructure (optical microscope).
Based on EDS analysis (Table 3), the surface
composition after PAC process was generally found to be
rich in oxygen, with weight ranging between 5.4% for
sample B and 31.3% for sample F. Only in sample B a
high content of atomic N (6.1 wt. %) was found on the
surface, as a consequence of using a 70% N2 rich plasma
gas that also to prevents surface oxydation. The higher
oxydation of sample F is caused by the lower cutting
speed that increases the plasma residence time and the
material overheating. It is well known that it is important
to limit the presence of oxide on the kerf surface to
guarantee its weldability, as the presence of oxide
inclusions would compromise the joint quality and
strength [10].
3.2 Cross sections
The base material shows a typical austenitic
microstructure with a relevant presence of sulphide
inclusions (Fig. 3), preferentially elongated along the
rolling direction. After being exposed to plasma the kerf
surface exhibits the presence of a remelted surface layer.
Sample A (H35) shows a layer with thickness ranging
between 5 - 35 µm along the kerf height (Fig. 4) while
sample B (30% H35 – 70% N2) shows a layer between
160-180 µm (Fig. 5). ). A dendritic microstructure
characterized the melted zones (Fig.5). Both samples are
15 mm thick and have been cut at the speed of 2000 mm/s.
For the same cutting speed a higher content of N2 as
plasma gas causes a thicker remelted steel layer. This is
probably caused by the higher thermal conductivity and
heat capacity of the sample B cutting mixture that has a
reduced amount of monoatomic species (Ar) in favour of
diatomic species (N2 and H2) with respect so sample A.
The higher oxygen concentration of the surface layer was
measured for both sample through EDS profiling (Fig. 6)
and mapping (Fig. 7) showing that the oxidation layer has
a peak near the kerf surface rapidly decreasing moving
inside the workpiece. Although non-oxidizing gas
mixtures were used in both A and B cases, this very thin
Fig. 4 Cross section of sample A (top: optical microscope,
bottom: SEM)
Fig. 5. Cross section of sample B (top: optical microscope,
bottom: SEM).
melted
melted
m
Fig. 6. EDS oxygen concentration profiling (a.u.) of the sample
melted layer A.
strongly affected by the overheating.
The darkest surface colour is found for the 50 mm thick
sample F characterized by a very high oxidation level.
However, comparisons between samples A and B shows
that sample B, with lower oxygen amount but higher
nitrogen content, has a darker colour with respect to
sample A. For this reason, it can be concluded that the
surface colour cannot be directly related only to the
oxidation level, but more generally to the chemical
composition and thickness of the external layer. A
qualitative analysis of the colour surface can be used to
evaluate general overheating but cannot be used as a
standalone evaluation to distinguish between different
oxidation levels and remelted layer thicknesses.
Future work will include micro-hardness profiling as to
determine the extension of the heat affected zone (HAZ).
Acknowledgements
Funding from C.I.R.I.-M.A.M./Cebora S.p.A. research
contracts is acknowledged.
Fig. 7. EDS oxygen concentration mapping (a.u.) of the sample
B melted layer.
The formation of intergranular Cr-rich carbide due to
stainless steel sensitization, which would deteriorate the
corrosion properties of the material, was not detected at
this stage.
4. Conclusions
In the present work, the effects of different Ar/H2/N2
plasma gas mixtures on the microstructural modifications
of the AISI 304L stainless steel in high current (400A)
plasma cutting has been experimentally investigated.
The kerf surface always shows a presence of a remelted
layer whose thickness is proportional to the heating
capacity of the plasma cutting gas mixture. A higher
percentage of diatomic species (N2 and H2) increases the
deepness of the layer for samples A and B, from 5 to 180
m. A thin oxygen rich layer of about 1 m is always
formed on the kerf surface and it is responsible of the
characteristic colour of the surface. For 15 mm plate
thickness and fixed cutting speed, microstructural analysis
of samples A (H35), B (H35 30% - N2 70%) and C (H35
70% - N2 30%) showed that the amount of atomic oxygen
on the kerf surface can be significantly reduced if a
plasma cutting gas very rich in N2 is used, with the
drawbacks of a thicker remelted layer and the presence of
atomic nitrogen on the surface.
The composition of the material on the kerf surface is
References
1. V.A. Nemchinsky, W.S. Severance, J Phys D Appl
Phys 39 (2006) R423–R438
2. V. Colombo, A. Concetti, E. Ghedini, S. Dallavale, M.
Vancini, Plasma Sources Sci. Technol. 18 (2009)
023001
3. V. Colombo, A. Concetti, E. Ghedini, S. Dallavalle,
Plasma Sources Sci. Technol. 20 (2011) 035010
4. V. Colombo, A. Concetti, E. Ghedini, F. Rotundo, S.
Dallavalle, Plasma Sources Sci. Technol. 19 (2010)
065023
5. F. Rotundo, C. Martini, C. Chiavari, L. Ceschini, A .
Concetti, E. Ghedini, V. Colombo, S. Dallavalle, Mat.
Chem. Phys. 134 (2012) 858-866
6. A. Gullu, U. Atici, Materials and Design 27 (2006)
1157–1162
7. ASTM E407 Standard Practice for Microetching
Metals and Alloys, 2011
8. D.P. Adams, V.C. Hodges, D.A. Hirschfeld, M.A.
Rodriguez, J.P. McDonald, P.G. Kotula, Surface &
Coatings Technology 222 (2013) 1–8
9. J.D. Rancourt, Optical Thin Films User's Handbook,
McGraw-Hill, New York, 1987.
10. A.R. Galvis E, W. Hormaza, Engineering Failure
Analysis 18 (2011) 1791–1799