Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
WELDING RESEARCH
G. GÖTT (g.goett@inpgreifswald.de) and D. UHRLANDT are with the Leibniz Institute for Plasma and Technology, Greifswald, Germany. A. GERICKE and
K.M. HENKEL are with Fraunhofer Application Center Large Structures in Production Engineering, Rostock, Germany.
WELDING RESEARCH
Fig. 1 — Setup with highspeed camera and spectrometer. Fig. 2 — Chemical compositions of SAWslags.
WELDING RESEARCH
Fig. 4 — Front view illustrating material transfer of the DCEP process with 600 A (Ref. 16, SOM2).
Element C Si Mn P S Al Ti B Fe
essary to keep the arc in constant focus cavern. If the shielding gas pressure is
Table 2 — Parameter Variation for the optical diagnostics. An overview too high, it will be injected into the
of the whole setup is shown in Fig. 1. It cavern’s atmosphere and influence the
Process Identifier Current Voltage consists of two HSCs and a spectrome- process. In the case of argon (Ar), it
DC+ 600 A 30 V ter. The welding was performed with an would change the process to a spray
DC −600 A −30 V inverter power source (Lincoln AC/DC transfer similar to the GMAW process.
AC 600 A 30 V 1000) with a maximum current of If the pressure is too low, the cavern
DC++ 1000 A 34 V ±1000 A. A constant current welding will shrink, which is visible in the weld
characteristic was chosen. joint profile.
ties to be able to change the resolu- In this paper, the single-wire SAW In addition, the tunnel tends to be
tion. The spatial distribution could be process was analyzed with four vary- clogged with debris. With a balanced
recorded with an optical system that ing parameter settings. The four pa- setting of the gas pressure, the influ-
contained spherical and several planar rameter changes that were observed ence on the process is minimized and
mirrors, an edge filter, and an ad- with the diagnostics are given in Table the view into the cavern is unobstruct-
justable aperture. Therefore, it was 2. The materials were not altered. The ed. The best results were achieved by
possible to distinguish between the wire was a Lincoln Electric L50M (EN using Ar at an overpressure of 25
different areas inside the cavern and ISO 14171 S3Si) with a diameter of mbar. Carbon dioxide (CO2) and Ar
determine where the different species 4 mm, and the base material was an were investigated as applicable shield-
were located. EN 10025 S355 J2+N. The main ing gases. None of the gases in the pre-
Optical emission spectroscopy chemical composition of the materials liminary trials changed the chemical
(OES) had been performed with a is listed in Table 1. The flux used was a composition of the weld deposit.
Spectromaxx by SPECTRO. The deter- Lincolnweld 8500 (EN 760 – S A FB 1) Nonetheless, changes in the chemical
mination of oxygen had been imple- with a basicity index of 2.9 and a neu- compounds of the molten slag, investi-
mented through carrier-gas melt ex- tral chemical behavior. The flux com- gated by x-ray fluorescence (XRF),
traction with a Bruker Elemental G8 position is listed in Table 3. The weld- were observed using CO2 — Fig. 2.
GALILEO ON/H analyzer. After finish- ing and wire-feed speed was constant. Furthermore, the measured short-
ing the welding process, the remaining The height of the pile of flux was kept circuit frequency changed from 3.6 Hz
droplets were collected and carefully constant as well. This was necessary to in the unaffected welding process to
cleaned from the remaining scale for keep the basic conditions steady. 4.2 Hz by injecting CO2 as a shielding
the carrier-gas melt extraction. The pressure that the flux applied gas. Short-circuit frequency was con-
In this setup, the welding head was to the cavern was about 0.05 g/mm². stant while using the inert gas Ar. This
fixed and the base material was moved The gas pressure that impinged on the indicated that the use of CO2 as a
by using a linear table with a constant cavern through the tunnel had to be shielding gas is more invasive to the
velocity of 1000 mm/min. This was nec- finely tuned to the pressure inside the process.
WELDING RESEARCH
Fig. 5 — Front view of DCEP process with 600 A and flux melt Fig. 6 — Side view behind the process showing the weld pool
ing into the droplet — see red marking (Ref. 16, SOM2). and the cavern ceiling (Ref. 16, SOM4).
Results and Discussion showed a stable arc behavior the entire wire. One is the kinking of the undu-
recorded time. The droplet transfer loid, a mathematical term used to de-
In the following subsection, some was turbulent and changed randomly scribe the geometry of the long molten
of the phenomenon observed with the between short circuiting dropping, ex- metal droplet that is still attached to
high-speed camera is presented. In or- ploding, and repelling — Fig. 4 and the upper electrode. Magnetic forces
der to understand the still frames, the Ref. 16, SOM1. drive the kinking and throw molten
supporting online material is recom- Most of the time, flux grains, and metal to the back. This effect can also
mended (Ref. 16). This will help to en- small metal and slag droplets, were be seen in GMAW processes with high
hance understanding on what each splattering through the cavern area. currents (Ref. 16, SOM3).
part in the frames represents, and it The analyzed videos show more or less The cavern was stable the entire ob-
will facilitate identification of those turbulent processes inside the cavern, served time, with just a few flux grains
parts. For a better understanding of although the SAW process is known falling from the walls. This means the
the findings and explanations, Fig. 3 for a high grade of stability and minimum internal cavern pressure
shows a snapshot of the clearly visible smooth weld joints. These are proba- was equal to the pressure applied by
moment in the DCEP process. bly a consequence of the slow solidifi- the flux on top of the cavern. The cav-
It is helpful to keep in mind from cation of the molten weld pool and the ern had a half-ovoloid shape with a
what perspective the process was ob- smoothing effect of the freezing slag. minimum width of 12 mm based on
served. In this case, it was from a low The reaction between flux and metal the given scale and visible wire diame-
angle just above the surface of the happens preferably at the contact ter. Figure 6 (SOM4) shows the rear
base material. The tunnel was the point between molten droplet and cav- part of the cavern where different ef-
outer-limiting part of each image, and ern wall in the welding direction, fects appear compared to the front
moving parts like the droplets, flux where flux is continuously molten and part. The view is mostly obscured by
grains, weld pool surface, and the wire absorbed by the droplet. the debris coming from the falling
were visible. The only light source was This reaction is clearly visible in flux. On the left side of this image, the
the arc itself, except for the hot sur- Ref. 16 (SOM2) as a front view, where sloping surface of the weld pool can be
faces that represented a very small the absorbed molten flux also leads to seen. In the center of the frame a part
part of the total emission. Therefore, a change of emissivity in the metal of the wire is visible, and left of the
particles in front of the arc are seen as droplet. This is probably the place and wire the molten cavern ceiling ap-
shadows. Particles next to or behind state in the welding process where the pears. It merges into the weld pool,
the arc are illuminated and can be seen most intensive chemical slag metal re- visible on the left end of the frame.
as bright spots. The surface of the liq- actions take place due to the spherical This part is where the cavern surface
uid metal has a low emissivity. There- droplet shape (positive ratio of absorb- in this area is mostly molten and
fore, it has a high reflectivity. It was ing-surface area to volume), the high migrates toward the metal surface.
perceived as a reflecting surface simi- reaction temperatures, and the con- As soon as the cavern surface gets
lar to mercury at room temperature. stant flux supply. In the lower part of in contact with the still molten weld
High-speed images of the DCEP the frame, the base material with joint, the cooling process starts be-
process with 600 A. The videos some flux and metal droplets can be cause there is no heat input any more.
showed different effects. In Fig. 5, the seen. Obscured by the base material Once the molten flux is cooled, it will
DCEP process, with its SAW basic pa- surface is a settling, which builds the peel off the weld joint as slag. This
rameters, is shown in a front view, and weld pool, respectively the emerging contributes to the high weld quality,
in the following section the general weld joint — Figs. 5 and 6. It is created since the cooling is slowed down and
findings are presented and discussed. by the arc pressure onto the liquid the atmospheric gases are held back
Later on, the characteristics of the weld pool. In the upper left corner of during this process. In contrast to the
other parameter sets are discussed in Fig. 5, parts of the melting tunnel are smooth surface of the slag, once
comparison to the DCEP process. The visible. cooled, it can be said that most parts
front view of the DCEP process Other effects take place behind the of the cavern wall are not as smooth.
WELDING RESEARCH
Fig. 7 — Side view shows the flux falling on the righthand side (Ref. 16, SOM5).
The inner cavern surface consists of cause this is a side view of the process. different compounds or crystal phases
solid flux grains, molten slag, and sol- Flux grains in different sizes are falling before melting of the flux grains. One
id particles merging into a molten through the cavern or along the cavern has to be aware that these compounds
stage (Ref. 16, SOM5). Within a short wall. These grains with a melting sur- have different, usually lower, melting
time of exposure to the heat source, face appear in the rear part of the cav- and solidification points compared to
the surface of the flux grains starts ern relative to welding direction as the separate components listed in
melting with visible outgassing or well, which is supposed to be the cold- Table 3 (Ref. 17).
even boiling on the surface. The over- est part because of its maximum For a first approximation of the in-
exposed region at the center is the arc distance to the burning arc. ternal cavern temperature close to the
with the hot wire tip. The emitted The flux used was agglomerated surface, these effects were disregard-
light illuminates the whole cavern. The and fluoride-basic. These fluxes typi- ed. Since the fusing process of the flux
arc contains mostly metal vapor and cally start melting at around 1200°– grains is visible in the high-speed im-
nonmetallic elements, like calcium 1400°C depending on chemical compo- ages, the cavern surface must rapidly
(Ca), according to the recorded spec- sition and the relation of mineral con- exceed the melting temperature of the
tra. On the right-hand side in each stituents (Refs. 17–19). There is the flux. This is possible due to the tem-
frame, the front wall can be seen be- possibility of reactions and forming of peratures of the arc and the liquid
Table 3 — Main Chemical Composition of Used Welding Flux and Melting Temperature Tmelt (Values in wt% and C, respectively)
Chem. Comp. SiO2 MnO MgO CaF2 Na2O Al2O3 CaO K2O TiO2 Metal Alloys
CONCN in % 13 1 30 24 2 19 8 1 1 1
Tmelt in C 1713 1650 2852 1423 1275 2050 2575 ± 5 2575 ± 5 1855 –
WELDING RESEARCH
WELDING RESEARCH
A B
the tunnel (Fig. 1). The upper part in gest that the main
the spectra represents the upper re- current path is situ- C
gion in the middle of the images and ated below the
the lower part in the spectrum, which droplet that is still
is the lower part in the middle of the attached to the wire.
image. Figure 10 shows frames from Therefore, the
Ref. 16 (SOM9). The images show droplet transfer in
both high-speed images of the process SAW has similarities
and synchronized spectra in combina- to a CO2 GMAW
tion. The recorded spectra were domi- process, although its
nated by iron (Fe), calcium (Ca), self- atmosphere is differ-
reversed sodium (Na) lines, and man- ent. In CO2, the
ganese (Mn) (compare to Fig. 9). main current path
Noticeable is the pair of Na lines at exits at the wire tip
the center with around 589 nm, which in contrast to the Ar-
appear as one dark stripe because they dominated GMAW
are strongly self-reversed. Most lines processes where it Fig. 10 — Three successive acquisitions of an AC process a few
below 580 nm are from Fe vapor. Both exits the wire above hundred µs apart. They go from positive phase via current zero
the line groups between 610 and 620 the liquid part of the to negative. Marked is the change in the spectrum from Mn
nm, and 643 and 652 nm, and with wire. This mecha- (Ref. 16, SOM9).
three lines each are from Ca. In the nism is necessary in
center image of Fig. 10 there is anoth- certain GMAW processes to achieve ements in the wire. As can be ob-
er set of spectral lines visible at globular and spray transfer (Ref. 21). served, slight changes occur by varying
around 601–602 nm, which are miss- The same effect was stimulated in SAW polarity with most melting loss of al-
ing in the frames before and after when the shielding gas introduced into loying elements in DCEN and AC
(see left and right image in Fig. 10). the tunnel was set to an excessively processes. This is especially the case
These lines originate from Mn (601.4, high pressure and Ar entered the cav- for alloying elements with a high affin-
601.7, and 602.2 nm), and are only ern. Under these circumstances, the ity to oxygen like carbon, aluminum,
detectable during the phase around droplet transfer changed to a constrict- and titanium.
current zero. ed spray transfer without any short cir- Oxygen is an important element in
This phase obviously has a lower cuits. This had to be avoided to main- welding metallurgy and can act both
arc temperature, and the composition tain a diagnostic method with as little positively and negatively on mi-
of the plasma allows these lines to be influence as possible. crostructure formation. In a balanced,
emitted. Just before the positive phase Chemical Analysis. The weld low amount, oxygen plays an impor-
at higher currents, there are many joints were analyzed by OES. The sam- tant role in nucleation and can sup-
more Fe lines visible and the Na line is ples for the OES of the weld metal port a fine-grained microstructure for-
also more intense. A similar situation were collected from bead-on-plate mation with improved toughness and
is visible in the negative phase. It welds with eight layers to avoid dilu- tensile strength. In interaction with ti-
might be because of the low boiling tion with the base material. Only a few tanium, boron, or other microalloying
temperature of Mn that its lines are changes were found in the chemical elements, this effect is enhanced (Ref.
visible at all. composition of the main alloying ele- 12). In contrast, a high amount of oxy-
Otherwise, the spectra show an Fe- ments within the varying processes gen in the weld joint leads to embrit-
dominated arc. This is consistent with (Table 1). This can be attributed to the tlement and porosity. Therefore, opti-
the earlier presented observation made chemically neutral character of the mized oxygen content is ideal for ade-
by the high-speed imaging. Both sug- flux and the low amount of alloying el- quate mechanical properties. In sub-
WELDING RESEARCH
WELDING RESEARCH