Effect of process parameters of micro-plasma

arc welding on morphology and quality in

stainless steel edge joint welds
K. H. Tseng, S. T. Hsieh and C. C. Tseng

The effects of the process parameters of micro-plasma

arc welding (micro-PAW) on the morphology and
quality of stainless steel edge joint welds were
investigated. Micro-PAW was applied on type 304
stainless steels to produce an edge joint weld. Welding
experiments were carried out for various combinations
of arc current, welding speed, arc length, shielding gas,
and clamp distance, with all other operating conditions
held constant. The experimental results indicated that
the collimated shape of the low current plasma arc was
mainly responsible for the low sensitivity of the weld
morphology to variations in the nozzle standoff
distances. The arc voltage increased with increasing
quantity of added hydrogen in the argon shielding gas.
It was also found that satisfactory edge joint welds
can be formed using a clamp distance of 0.35 mm,
and that the edge joint penetration on a 0.1 mm 1 Comparison of gas tungsten arc welding (GTAW) and
thickness stainless steel is about 60% of the clamp plasma arc welding (PAW) processes
distance. STWJ/373

The authors are in the Welding Technology Section, welded to enable heat to be generated in forming
Metal Processing Research and Development Depart- the arc roots.
ment, Metal Industries Research and Development (ii) The nozzle constricts the plasma to form a
Centre, 1001 Kaonan Highway, Nantzu, Kaohsiung, columnar arc which, compared with the gas
Taiwan 811, Republic of China ([email protected]. tungsten arc, is more directional and exhibits a
org.tw). Manuscript received 18 November 2002; lack of puddle size sensitivity to arc length changes.
accepted 4 February 2003. (iii) Controlled penetration is achieved via the use of a
keyhole mode welding technique. The most sig-
# 2003 IoM Communications Ltd. Published by Maney for ni cant difference between the GTAW and plasma
the Institute of Materials, Minerals and Mining. welding arcs lies in the keyhole technique. This
technique produces a small hole that is carried
along the weld bead. During the welding process,
the hole progressively cuts completely through the
workpiece, with the molten weld pool owing
INTRODUCTION behind to form the weld under surface tension
Fabrication of metal structures in space will require forces.
extensive use of various metal joining processes. The (iv) Since the electrode is held within the constricting
plasma arc welding (PAW) process is one of several nozzle, it is not possible for the electrode to
methods being considered for this purpose. Plasma arc touch the weld pool. The possibility of tungsten
welding is an arc welding process in which coalescence, or inclusions in the weld metal is reduced and
the joining of metals, is produced by heating with a the period between electrode adjustments is
constricted arc between a non-consumable electrode and the extended.
workpiece (transferred arc) or the electrode and the The PAW technique can offer higher quality welds than
constricting nozzle (non-transferred arc).1 3 The PAW GTAW and at lower equipment cost than laser beam or
process is essentially similar in operation to the gas tungsten electron beam welding, and it may be the most effective
arc welding (GTAW) process in that the arc is used as a heat process for many applications. These include the welding of
source to fuse the joint and, when required, ller metal is stainless steel expandable bellows, where PAW is more
added to form the welds. In contrast to the GTAW process, tolerant to joint misalignment than laser beam welding and
because the electrode is recessed within the constricting gives better weld penetration than the GTAW technique.
nozzle (Fig. 1), the PAW process has several singular Consequently, the practical PAW process is widely accepted
operating characteristics as follows.4 7 in the aerospace, chemical, naval and shipyards, and
(i) A pilot arc can be formed between the electrode nuclear industries, etc.
and the constricting nozzle; since a pilot arc is Many instruments require welds of great accuracy. The
relatively inef cient as a heat source, an arc must be micro-plasma arc welding (micro-PAW) process, with its
transferred from the nozzle to the workpiece to be extremely ne control and high precision, provides the only

DOI 10.1179/136217103225009107 Science and Technology of Welding and Joining 2003 Vol. 8 No. 6 423
424 Tseng et al. Effect of micro-PAW process parameters on SS edge joint welds

2 Micro-PAW equipment used

3 Joint design and xturing tolerances on 0.1 mm thick-
viable practical possibility of producing such critical welds. ness stainless steel
Micro-PAW technology was developed to obtain a very
stable, controllable arc for joining ne sheet down
to 0.1 mm thickness, and wire and mesh sections, Micro-PAW was carried out with 31.5 mm outer diameter,
because of the precise control that is achievable in 0.1 mm thickness test specimens using the melt in welding
respect of arc initiation and input of heat to the work- mode. The programmable and transistorised direct current
piece. Typical applications of micro-PAW include joining power source, supplying a regulated arc current of 0.08 to
of aspiration needles, battery seals, catheter guide 50 A, was used to generate a low current needlelike arc on
wires, diaphragm bellows, enclosure lters, microrelay negative polarity. A machine mounted welding torch having
cases, etc. a standard 1.5% lanthanated tungsten electrode (1 mm
A needle plasma jet is obtained by passing a very low diameter) was used. The electrode tip con guration was a
current arc through a small diameter ori ce.7 ,8 The narrow sharp point with a 30 included angle. The preferred plasma
needlelike stiff arc generated at currents within the range 0.1 gas for low current plasma welding is argon because its low
to 15 A can prevent arc wander; the equivalent gas tungsten ionisation potential ensures reliable starting and a depend-
arc at comparable current suffers from instabilities because able pilot arc. In the present work, the ori ce gas owrate
of arc wander, and is much more diffuse. Compared was 0.25 L min2 1 through a 0.8 mm diameter ori ce and
with a gas tungsten arc, the micro-plasma jet has improved pilot arc current was xed at 5 A. Argon is also a suitable
low current arc direction and stiffness, increased arc shielding gas for welding stainless steels, but it does not
stability, and greatly reduced sensitivity to changes in arc necessarily produce optimum joining results. The welding
length. It is possible to vary the arc length over a processes were performed using pure argon and Ar H2
comparatively wide range (up to 20 mm) without adversely mixed gases. The shielding gas owrate was 7 L min2 1 for
affecting stability and, owing to the nature of the all compositions.
constricted plasma, without causing excessive spreading To achieve uniform fusion in thin sheet material, it is
of the arc.4 crucial that the component edges are accurately machined
The successful joining of foil thickness metal sheets using and that the clamping provides a uniform heat sink. The
the micro-PAW technique depends on the availability of all joint design and xturing tolerances in the present work are
necessary equipment (power source, cooling unit, torch, shown in Fig. 3. For a high precision edge joint on 0.1 mm
plasma gas, and shielding gas), precise control systems thickness stainless steel, a maximum joint gap of 0.05 mm is
(orientation and motion), adequate joint design and permitted. The maximum allowable joint mismatch is
xturing, and careful cleaning procedures. The process 0.1 mm to ensure that both joint edges are in continuous
characteristics and the operating parameters, which contact and that both edges melt simultaneously to fuse
can in uence weld morphology and quality, become together into a smooth fusion weld. Additionally, clamp
rather signi cant in such low current foil welding applica- distances of 2 to 4 times the thickness of the sheet are
tions. In the present work, detailed experiments were required to ensure correct joint tup and prevent warping of
conducted to investigate the effect of micro-PAW process the joint edges. The chilled clamps that held test specimens
parameters on morphology and quality in stainless steel in contact were manufactured from brass to prevent heat
edge joint welds. buildup.
Strict attention to the initial cleaning of the component
surfaces is necessary because oxides or grease lms affect
EXPERIMENTAL the surface tension of melted joint edges and the degree of
Austenitic stainless steel 304 having the chemical composi- melting attained at a given heat input level when thin metals
tion and mechanical properties given in Table 1 was used. are welded. Cleaning of the grease lms was necessary to
The equipment used in the present work is shown in Fig. 2. prevent the formation of porosity. With respect to cleaning,

Table 1 Chemical composition (wt-%) and mechanical properties of type 304 stainless steel used

C Si Mn P S Cr Ni Fe Yield strength, MPa Elastic modulus, MPa Poissons ratio

0.07 0.47 1.12 0.02 0.03 18.5 8.25 Bal. 287 191 0.25

Science and Technology of Welding and Joining 2003 Vol. 8 No. 6

 Tseng et al. Effect of micro-PAW process parameters on SS edge joint welds 425

5 Comparison of gas tungsten arc and micro-plasma arc

static volt ampere characteristic curves
4 Schematic illustration of weld bead geometry of edge
joint
Effect of hydrogen in argon shielding gas on arc
characteristic and weld morphology
the number of defects can be kept low using an ultrasonic Generally, arc power equals the heat lost from the plasma
method. Before welding, all test specimens in the present arc column, and the arc characteristic is a function of the
work were cleaned using a conventional ultrasonic cleaner heat loss.9 Figure 6 shows the characteristics of the plasma
with acetone solvent to remove surface impurities, and the arc in the micro-PAW process for various amounts of added
cleaned parts were stored in plastic bags until ready for hydrogen in the argon shielding gas at a certain arc length.
welding. The arc current was maintained at a constant value, and it
Figure 4 shows a schematic illustration of the weld was found that the arc voltage increases as the amount of
bead geometry. Optical microscopy was used to measure the hydrogen added to the argon atmosphere increases.
bead width and weld penetration. A minimum of three Figure 7 shows the thermal conductivity of each of the
measurements were made and the average values are gases. The diagram shows that the thermal conductivity
reported. All metallographic specimens were prepared by of hydrogen is much higher than that of argon at
mechanical lapping, grinding, and polishing to 0.3 mm operating temperatures typical of the welding arc because
nish. of the dissociation and recombination effects. Conse-
quently, more heat is lost from the plasma arc column

RESULTS AND DISCUSSION

Static volt ampere characteristic of low current arc
welding process
The relationship between arc voltage and arc current at a
certain arc length is known as the static volt ampere
characteristic. The gas tungsten arc and plasma arc static
volt ampere characteristics are compared in Fig. 5 for arc
currents below 5 A in argon shielding gas. The voltage of an
argon shielded tungsten arc increases from 15 to 23 V, a
change of about 53%, as the arc current is reduced from 5 to
1 A. It can be seen that the negative slope of the gas
tungsten arc volt ampere characteristic curve is fairly steep
at extremely low current. In a low current PAW process, the
slope of this volt ampere characteristic curve is almost at,
with the arc voltage increasing from 21 to 22 V, i.e. a 5%
change, as the arc current is again reduced from 5 to 1 A. In
contrast to the GTAW process, the PAW process combines
a continuously operating pilot arc within the torch and
arc constriction to provide a needlelike plasma jet, which
is continuous and stable at very low currents. It thus
provides greater arc stability than a gas tungsten arc.
To reiterate, in the present work a needle plasma arc
welding technique was developed to provide an arc that
is stable at very low currents for melt in welding of thin
gauge metals. 6 Arc characteristics for various shielding atmospheres

Science and Technology of Welding and Joining 2003 Vol. 8 No. 6

426 Tseng et al. Effect of micro-PAW process parameters on SS edge joint welds

7 Thermal conductivity of possible shielding gases as

function of temperature (after Ref. 10)

as hydrogen is added to the argon shielding gas. The

loss of energy must be balanced by the gas in the plasma
arc column; the power generated in the arc, the
product of arc current and arc voltage, is therefore also
increased. These effects increase the arc voltage at a given
arc current as hydrogen is added to the pure argon
atmosphere.
The composition of the shielding mixture for arc welding
depends primarily on the type of base metal to be welded.
Notably, for high alloy stainless steels, the use of shielding
mixtures consisting of argon and about 1 5% added
hydrogen is preferred, with the risk of porosity in the welds
occurring at greater hydrogen contents.8 ,1 1 ,1 2 Figure 8
shows the macroscopic sections of solidi ed welds
obtained using the micro-PAW process with various
shielding mixtures at a constant arc current. The arc
voltages varied owing to the different shielding gases used.
As hydrogen is added to the pure argon atmosphere,
increasing the arc voltage and consequently the arc power,
the arc heats up and heat transfer to the fusion zone b
becomes more ef cient, melting a large quantity of the
base metal. a pure argon (arc voltage 19.5 V); b Ar 5%H2 mixture (arc
It can be seen in Fig. 9 that a high quality weld was voltage 21.3 V)
obtained on stainless steels by adding a small amount of 8 Macroscopic view of edge joint welds produced using
hydrogen. Hydrogen is a reducing gas, and hence combines same arc current intensity of 2.5 A and welding speed
with oxygen and hinders the formation of oxide on the weld of 2.0 rev min2 1, but under different shielding mixtures
surface;1 1 1 4 therefore, high quality welds with a clean (optical)
surface can be obtained. Furthermore, the addition of
hydrogen assists in the promotion of good puddle uidity
and wetting action on the stainless steel welds.7 ,1 3 Accord- shown in Fig. 10. The experimental results clearly
ingly, a smooth weld metal can be obtained using an Ar H2 indicate a low sensitivity of puddle size to changes in
mixture. nozzle standoff distance. As shown in Fig. 10, the weld
depth/width ratio for a 3 A plasma arc current in an Ar H2
shielding gas varies from 0.85 to 0.81, a change of about 5%,
Effect of micro-plasma arc lengths on weld as the nozzle standoff distance increases from 0.2 to
morphology 0.8 mm. The needle plasma arc assumes a collimated shape
In the present work, the distance by which the electrode is and, as the micro-plasma arc length is varied within normal
recessed within the constricting nozzle was xed at 1 mm, limits, the heat transfer rate per unit area and the arc
whereas the distance from the outer face of the constricting current intensity are virtually constant. Thus, a needle
nozzle to the workpiece was varied. The effect of nozzle plasma arc can allow greater tolerance of variations in
standoff distance on the geometry of the weld bead is arc length.

Science and Technology of Welding and Joining 2003 Vol. 8 No. 6

 Tseng et al. Effect of micro-PAW process parameters on SS edge joint welds 427

10 Effect of nozzle standoff distance on weld bead geo-

metry during micro-PAW process

Bead width, weld penetration, and weld depth/width ratio

were selected as being most useful in characterising the
process. A note of caution is that the clamping design may
signi cantly affect weld bead geometry. Figure 14 shows the
effect of clamp distance on the weld bead geometry of an
edge joint at constant arc current, welding speed, and
plasma arc length. A greater clamp distance, allowing
more free space to form a weld, tends to increase the
b weld depth/width ratio. When the clamp distances are
below approximately 0.25 mm, the edge joint welds
a pure argon shielding gas; b Ar 5%H2 mixture cannot form a circular cross-section (Fig. 15a and b).
9 Effect of hydrogen addition to argon shielding gas on When the clamp distances are greater than 0.30 mm,
weld quality in type 304 stainless steel weld metals roughly circular edge joint welds (width>2t) can be
(optical) 640 formed (Fig. 15c and d). In the present work it was found
that satisfactory edge joint welds (width>2t and penetra-
tion>2t) could be formed using a clamp distance of
Effect of process parameters on acceptable welding 0.35 mm. However, warping of the joint edges starts to
zone
Figure 11 shows the in uence of micro-PAW process
parameters on the acceptable welding zone of the edge
joint. All the welds were produced using an arc length of
1.5 mm on 0.1 mm thickness stainless steels, using an Ar
5%H2 shielding gas mixture. At low arc currents, the welds
were unfused because the heat input was insuf cient to
melt the specimens (Fig. 12a); however, at higher arc
currents, smooth and uniform welds could be obtained
(Fig. 12b). A further increase in the current tended to lead
to formation of uneven or intermittent weld surfaces
because of the excessive heat input, which results in
incomplete weld pro les (Fig. 12c and d). Figure 11
indicates that an acceptable welding zone (shaded region)
is formed within a particular range of arc current and
welding speed at a given arc length. The bead width is
proven to be proportional to the weld penetration. The
quality of the weld is satisfactory if the bead width is
between 2.1 and 2.6t, which corresponds to a weld
penetration of 1.7 to 2.2t, where t is the thickness of the
test specimen. In this acceptable welding zone, a sound weld
bead geometry in an edge joint can be achieved, as shown
in Fig. 13.

Effect of clamp distance on weld bead geometry 11 Effect of process parameters of micro-PAW on accep-
To characterise the welding process, one or more basic table welding zone for edge joint on 0.1 mm thickness
features that are a function of the process must be selected. stainless steels

Science and Technology of Welding and Joining 2003 Vol. 8 No. 6

428 Tseng et al. Effect of micro-PAW process parameters on SS edge joint welds

13 Macroscopic section (optical) of edge joint weld

within acceptable welding zone (arc current 2.6 A,
welding speed 1.5 rev min 1, plasma arc length
1.5 mm, and clamp distance 0.35 mm)

occur at clamp distances greater than 0.40 mm. Addition-

b ally, Fig. 14 also shows that using the micro-PAW
technique, the roughly circular edge joint penetration
produced on a 0.1 mm thickness stainless steel is about
60% of the clamp distance.

Process application: welded bellows

Bellows are expected to provide a hermetic seal with a low
resistance to mechanical motion. They are used in vacuum
systems, semiconductors,medical applications,petrochemical

12 Weld con gurations of a unfused, b smooth and

uniform, c uneven, and d intermittent welds 14 Relationship between weld bead geometry and clamp
(optical) 625 distance

Science and Technology of Welding and Joining 2003 Vol. 8 No. 6

 Tseng et al. Effect of micro-PAW process parameters on SS edge joint welds 429

a b

c d

15 Macroscopic sections (optical) showing effect of clamp distance of a 0.20, b 0.25, c 0.30, and d 0.35 mm
on weld bead geometry of edge joint (arc current 2.7 A, welding speed 2.2 rev min2 1, and plasma arc length
1.5 mm)

applications, instrumentation industries, and other con- CONCLUSIONS

texts. Welded bellows have been successfully manufactured 1. In the low current plasma welding process, the volt
using the present micro-PAW technique. A stainless steel ampere characteristic curve is almost at. Consequently,
strip having a 0.1 mm thickness wall was stamped to this technique can provide an arc that is stable at very low
produce washer shaped diaphragms having concentric currents for welding thin metals.
circular ripples (Fig. 16a). Pairs of diaphragms were 2. The collimated shape of the low current plasma arc is
welded along the inside edge to form convolutions. mainly responsible for the low sensitivity of the weld
These convolutions were then stacked together and their morphology to variations in the nozzle standoff distances.
outside edges were welded. The resulting product was a 3. As hydrogen is added to the argon shielding gas,
welded bellows, which is a highly exible metal tube increasing the arc voltage and consequently the arc power,
(Fig. 16b). The results of the present work indicate that the the ef ciency of the heat transfer to the fusion zone
1.5 mm length, 2.8 A needle plasma produced clean, increases, melting a larger quantity of base metal.
smooth, and uniform edge joint welds at a speed of 4. In PAW of stainless steel in an Ar H2 mixture, a
1.5 rev min2 1 . In these experiments, welding was performed stable process and high quality welds having very smooth
in an Ar 5%H2 mixture shielding gas. surfaces can be achieved.

Science and Technology of Welding and Joining 2003 Vol. 8 No. 6

430 Tseng et al. Effect of micro-PAW process parameters on SS edge joint welds

5. Satisfactory edge joint welds can be formed using

a clamp distance of 0.35 mm, and for the present micro-
plasma welding technique on 0.1 mm thickness stainless
steel, the roughly circular edge joint penetration produced is
about 60% of the clamp distance.

REFERENCES
1. s. e. barhorst, e. h. daggett, s. a. hilton, j. t. perozek and
e. spitzer: in Welding handbook, (ed. R. L. OBrien), 8th edn,
Vol. 2, 330 350; 1991, Miami, FL, American Welding Society.
2. s. kou: in Welding metallurgy, 1st edn, 3 27; 1987, Toronto,
Ont., Canada, John Wiley & Sons.
3. h. b. cary: in Modern welding technology, 4th edn, 70 106;
1998, Englewood Cliffs, NJ, Prentice-Hall.
4. w. lucas: in TIG and plasma welding, 1st edn, 80 108; 1990,
a Cambridge, UK, Abington Publishing.
5. c. d. lundin and w. j. ruprecht: Weld. J., 1974, 53, (1), 11 19.
6. y. m. zhang and s. b. zhang: Weld. J., 1999, 78, (2), 53s 58s.
7. e. f. gorman: Weld. J., 1969, 48, (7), 547 556.
8. r. l. o9brien: Weld. Res. Counc. Bull., 1968, 131, 1 37.
9. j. f. lancaster: in The physics of welding, 2nd edn, 146 227;
1986, Oxford, UK, Pergamon Press.
10. c. m. cho, h. c. chiang and c. n. dong: Proc. 1st National
Conf. on Welding technology, Taipei, Taiwan, December
1987, Welding Assoc. of the Republic of China, 159 172.
11. w. lucas: Weld. Met. Fabr., June 1992, 60, 218 225.
12. b. bennett: Weld. Met. Fabr., July 1990, 58, 335 336.
b 13. d. hilton: Weld. Met. Fabr., July 1990, 58, 332 334.
14. m. onsIEN, r. peters, d. l. olson and s. liu: Weld. J., 1995, 74,
16 Samples of a diaphragms and b welded bellows (1), 10s 15s.

Science and Technology of Welding and Joining 2003 Vol. 8 No. 6