Journal of Materials Processing Technology 214 (2014) 123131

Contents lists available at ScienceDirect

Journal of Materials Processing Technology

journal homepage: www.elsevier.com/locate/jmatprotec

Wetting behavior of eutectic AlSi droplets on zinc coated

steel substrates
Marius Gatzen , Tim Radel, Claus Thomy, Frank Vollertsen
BIAS Bremer Institut fuer angewandte Strahltechnik GmbH, Klagenfurter Str. 2, 28359 Bremen, Germany

a r t i c l e i n f o a b s t r a c t

Article history: Transient spreading behavior, joint properties and metallurgical compositions are investigated for differ-
Received 4 June 2013 ent hot-dip and electro-galvanized zinc coatings. The main focus is set on the effect of coating thicknesses
Received in revised form 13 August 2013 and droplet size. While most of the droplets are observed on surfaces at room temperature, the case of
Accepted 15 August 2013
pre-heated substrates is also accounted for. Both the coating thickness and the droplet size have little
Available online 27 August 2013
effect on the resulting wetting angle compared to the effect of preheating or the absence of a coating.
The transient spreading behavior signicantly differs for different coating types. The coating thickness
Keywords:
affects heat transfer into the substrate during the initial stage of wetting. The metallurgical composition
Dissimilar materials
Wetting
shows that the coating is removed over a broad interfacial area, while it accumulates at the toe of the
Galvanized steel deposited braze metal most likely due to uid dynamic effects.
2013 Elsevier B.V. All rights reserved.

1. Introduction
Joining of dissimilar materials is of signicant importance
in many elds of industrial manufacturing. Combining the
lightweight properties of aluminium with the strength and forma-
bility of steel can provide increased safety and weight reduction,
e.g. in the automotive industry. Laser or laser-hybrid welding and
brazing gained signicant attention for joining of dissimilar mate-
rials in the case of aluminium/steel and aluminium/titanium joints.
An overview of this topic was given by Thomy (2009). To improve
corrosion resistance, low-carbon deep-drawing steel sheets are
typically hot-dip or electro-galvanized with zinc or zinc alloys in a
continuous process. An overview of the various metallurgical prop-
erties of hot-dip galvanized steel substrates is given by Marder
(2005). Song et al. (2012) showed that during hot-dip galvanizing,
the pool of liquid Zn typically contains a small amount of Al (less
than 1 wt.%), leading to an exothermal reaction resulting in a thin
layer (0.1 m) of Fe2 Al5 Znx between the steel substrate and the
actual zinc coating. The kinetics of these reactions was described
by Giorgi et al. (2005). In contrast, electro-galvanizing is a non-
thermal process resulting in a nearly pure zinc coating on the steel
substrate, the properties of which depend on parameters such as
the applied voltage and plating time, investigated, e.g., by Popoola
and Fayomi (2011).
Contrary especially to deep-penetration laser welding, where
Schmidt et al. (2008) showed that the low evaporation temperature
Corresponding author. Tel.: +49 421 218 58041; fax: +49 421 218 58063.
E-mail address: gatzen@bias.de (M. Gatzen).
of the zinc may cause signicant process instabilities and requires
careful welding parameter selection, in laser brazing zinc coatings
are generally considered to be helpful. As an example, Vollertsen
and Thomy (2009) showed that these zinc coatings favour the
wetting of aluminium on coated steel substrates. Numerous stud-
ies have been performed on the wetting and spreading process of
different liquid metals on coated and uncoated steel substrates.
Special attention was paid to the resulting wetting angle and wet-
ting length in the case of joining of aluminium to steel by brazing
aluminium on coated steel. Kreimeyer (2007) established that
wetting length is a crucial factor of inuence on the mechanical
properties of the joint of dissimilar materials. In an experimental
series presented by Moeller et al. (2011), wetting length and
mechanical properties meeting industrial requirements in terms
of tensile strength and intermetallic phase seam thickness were
achieved.
In all brazing processes solid material (the substrate) is dissolved
into the liquid phase (the braze metal). Especially during contact
of liquid aluminium with solid steel, dissolution of steel occurs,
resulting in the formation of brittle intermetallic phases after solid-
ication. In the case of very short interaction time (as is the case
in high-speed joining processes such as laser brazing) the inter-
action between liquid aluminium and solid steel is known to be
dominated by reaction phenomena rather than diffusion effects, as
Bouche et al. (1998) pointed out. Considering that the coating itself
is molten during contact with liquid aluminium, the interfacial
reaction can typically be described as a reactive wetting process.
To characterize the effect of different alloying elements in the zinc
pool on the wetting behavior during reactive wetting, isothermal
hot-dip experiments are widely used.

0924-0136/$ see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jmatprotec.2013.08.005
124 M. Gatzen et al. / Journal of Materials Processing Technology 214 (2014) 123131
Fig. 1. Experimental setup for droplet generation and dened substrate wetting under inert gas atmosphere. Photography and sketch of the process chamber and droplet
detachment sequence.
However, the wetting process during brazing or welding is not
isothermal. Previous experiments have revealed an accumulation
of zinc during spreading on galvanized steel in the outer bound of
the seam toe, e.g. by Zhang and Liu (2011) in MIG brazing. This
was observed for different liquid metals, wetting conditions and
processes, e.g. by Koltsov et al. (2010) in case of copper droplets
wetting zinc-galvanized steel and by Peyre et al. (2007) in laser
beam overlap brazing of zinc coated steel with liquid aluminium
or by Tan et al. (2013) for magnesium alloy. The microstructure of
these zinc-rich zones has been investigated by Agudo et al. (2007)
and found to contain between 20 at.% and 60 at.% of zinc.
Up to now, little is known about the actual uid dynamics inside
the molten metal drop during the wetting of a zinc coated sur-
face in brazing. It is not clear how the distribution of zinc in the
propagating liquid is determined by uid dynamics inside the liq-
uid and/or diffusion and solubility effects. However, Thomy and
Vollertsen (2012) have shown in their experiment that the zinc
coating is crucial to the brazing process, both with respect to the
stability of the brazing process as well as to the properties of the
joints. Consequently, for the case of non-isothermal wetting at very
short interaction times, an experimental data base on the temporal
thermal and dynamic behavior of the liquid melt on zinc coated
surfaces is required.
The aim of this study is to characterize the thermal and dynamic
behavior during the wetting of liquid aluminium droplets on differ-
ent zinc-coated steel substrates together with the resulting wetting
angle and metallurgical appearance. To this end, an experimental
setup is developed to produce droplets with reproducible temper-
ature and size from a commercial aluminiumsilicon ller wire
(AlSi12). Special attention is paid to the inuence of the coating
thickness on the transient spreading process, droplet cooling and
resulting wetting angle.
2. Experimental methods
2.1. Experimental model system and procedure
An experimental model system was designed to generate liquid
aluminium droplets of reproducible size and properties dripping
onto a galvanized steel substrate (at room temperature (RT) or pre-
heated (PH)) under argon gas atmosphere to prevent oxidation. It
consists of a process chamber that contains an adjustable nozzle
for the aluminium wire and a clamping device for the coated steel
sheets (Fig. 1). A slightly defocused Trumpf TruDisk 8002 solid state
laser beam (wavelength 1030 nm) is irradiating the tip of the wire
with a xed power of 900 W, thus generating a constantly grow-
ing drop of liquid aluminium while the wire is fed with a speed
of 25 mm/s. As the diameter of the laser beam is approx. 1.5 mm in
the axis plane of the protruding wire, the complete droplet remains
irradiated during its growth. After a short, adjustable irradiation
time tir the wire is drawn back into the nozzle, stripping off the liq-
uid at the tip of the wire. Subsequently, the droplet is falling down
onto the substrate from a dened height of hd = 22 mm.
The falling process, the impingement and the subsequent wet-
ting process are recorded at 2000 fps with a high-speed camera
through an observation window in the process chamber. In addition
to the camera, a 2-color pyrometer is aligned to the area of impact.
It acquires the highest temperature in an area of about 56 mm
in diameter (see Fig. 2) at a frequency of 500 Hz. The peak tem-
perature is measured when the droplet enters the radiation cone
of the slightly inclined pyrometer spot (see Fig. 2). The measured
peak temperature is dened as the impinging temperature Td,im .
The temperature then typically decreases rapidly until it drops
below the value of 500 C. Although this is not exactly the solid-
ication temperature of the aluminium alloy (approx. 577 C for
eutectic AlSi alloy according to Murray and McAlister (1984)), the
time between the peak temperature and this point is dened as a
characteristic time tSol indicating the solidication process.
To measure droplet mass md , the wetted specimen is weighted
before and after wetting. By varying irradiation time tir before the
wire is drawn back into the nozzle, md can be altered. During the
experimental series, three different irradiation times were used
(tir = 1.4 s/1.9 s/2.4 s), producing droplets with an average mass of
about 90 mg, 110 mg and 140 mg. Fig. 3 gives the impinging tem-
perature Td,im and masses md for the three different irradiation
times.
The standard deviation of the droplet mass is very small (3 mg),
guaranteeing a very reproducible droplet size. The standard devia-
tion of the impinging temperature amounts up to 80 C (5% of
the mean value), which is small enough to enable systematic inves-
tigations. The mean temperature of the droplet is not affected the
irradiation time. Hence, the droplet size can be altered indepen-
dently from it.
To enable wetting experiments on zinc coatings with dened
pre-heating temperature, an inductive device, positioned under-
neath the specimen, is used. It guarantees a homogeneous
preheating temperature TS throughout the area of impact and can
be used to approximate the conditions of real-world local brazing
processes, where there is heat input not only into the braze metal,
but also (and signicantly so) into the base metal.
 M. Gatzen et al. / Journal of Materials Processing Technology 214 (2014) 123131 125

1600

measured droplet temperature Td

Td,im t sol
C

1200 defocused pyrometer

AlSi12 droplet
1000 highspeed camera

800

600

400 observation window

0
0,0 0,5 1,0 s 2,0
BIAS ID 130641
time

Fig. 2. Methododolgy to determine the impinging temperature Td,im and characteristic solidication time tSol from pyrometric measurements.

2.2. Specimen characterization different zinc coatings (hot-dip galvanized and electro-galvanized)
and coating thicknesses dZn . They are given in Table 2.
Cross-sections of all deposited droplets are prepared by cutting Typical cross-sections for both coating types are given in Fig. 4.
the specimen through the middle of the droplet, grinding and pol- The hot-dip galvanized coatings show a comparatively at surface
ishing the surface and etching it with Nital (90% ethanol + 10% nitric while the electro-galvanized coating is more wavy. Comparatively
acid) etchant. After preparation, microscope pictures are taken with big grains can be observed in the hot-dip galvanized zinc layer,
a magnication of 500:1. The cross-sections are used to measure the while the structure is very ne in the electro-galvanized coating.
average wetting angle. It is dened as the angle between the spec- WDS analysis of the specimens reveals the distribution of alu-
imen surface and the connecting line from the triple point and the minium, iron and zinc. The scanning results are also given in Fig. 4.
surface of the cap of the deposited droplet at a horizontal distance In the case of hot-dip galvanized specimen, the typical small FeAl-
of 0.25 m from the triple point (see Fig. 11). layer between the steel substrate and the actual coating can be
Representative cross-sections are additionally analysed with detected. All coatings show local variations in their layer thick-
wavelength-dispersive X-Ray spectroscopy (WDS) using a micro- ness, especially the ZE75/75 electro-galvanized coating. Hence, an
probe to quantify the zinc, aluminium and iron distribution. The average value is also given in Table 2.
analysis is performed for a Z100 hot-dip and a ZE75/75 electro-
galvanized specimen. Two areas, at the centre of the deposited
3. Results
droplet and at the outer toe of the deposited droplet, with a size
of 200 m 200 m are scanned.
3.1. Spreading behavior

2.3. Materials Fig. 5 shows a series the transient spreading processes on the
coated specimens at room temperature. In case of the uncoated
The experiments were carried out with a standard eutectic substrate, no spreading process occurred, and the droplet bounced
AlSi12 (wt.%12Si) ller wire of diameter 1.0 mm to generate the of the substrate due to the action of surface tension.
droplets. The chemical composition is listed in Table 1. There is an obvious difference between the spreading behav-
Five different substrate specimens were used, all with a size ior on the hot-dip galvanized and the electro-galvanized surface.
of 40 mm 40 mm and of almost the same thickness, but with The spreading process on the electro-galvanized surface is non-
uniform. After hitting the surface (t = 0 ms), the droplet is initially
deformed, obviously without wetting the surface (5 ms), until
1800 the galvanized surface starts to evaporate, which is indicated by
the occurrence of fumes. The droplet then rapidly starts to spread
droplet impinging temperature Td,im

C over the surface leading to strong dynamics within the droplet

t ir=1.9 s (1020 ms). The process is then dominated by strong wave-like
1600 surface motions until the droplet solidies.
t ir=1.4 s t ir=2.4 s
On the hot-dip galvanized surface, the spreading process starts
1500 immediately after the droplet is impinging on the surface, obviously
without any prior deformation or non-wetting stage. Simulta-
1400
neously the surface is evaporated, which is also accompanied by
the occurrence of fumes. The spreading continues almost steadily
1300
with only small surface motions of the liquid drop until it reaches its
1200 nal width. Noticeable is also the time when fume at the interface
0
0 50 100 150 mg 250
Table 1
droplet mass md
Chemical composition of the aluminium wire in %.
BIAS ID 130642
Si Fe Cu Mn Mg Zn Ti Be
Fig. 3. Impinging temperature Td,im and droplet mass md for different irradiation 11.0013.00 <0.60 <0.30 <0.15 <0.1 <0.2 <0.15 <0.0003
times tir .
126 M. Gatzen et al. / Journal of Materials Processing Technology 214 (2014) 123131

Table 2
List of used coated and uncoated low-carbon steel substrates.

Base material Coating specication Type Specimen thickness [mm] Measured coating thickness [m] Ave. coating thickness dZn [m]

DC04 Uncoated 1.0

DC04 +ZE50/50 Electro-galvanized 0.7 5.07.0 6.0
DC04 +ZE75/75 Electro-galvanized 0.8 4.09.0 6.5
DX56 +Z100 Hot-dip galvanized 0.8 6.08.0 7.0
DX56 +Z140 Hot-dip galvanized 0.81 10.011.0 10.5

Fig. 4. WDS micrograph of specimen cross-section with different coating types: (a) hot-dip galvanized coating (Z100; dZn = 7.0 m); (b) electro-galvanized coating (ZE 75/75,
dZn = 6.5 m).

can be observed for the rst time. While in case of hot-dip gal-
vanized coating it occurs almost immediately after impingement
(12 ms), it can be observed comparatively late (57 ms) on the
electro-galvanized surface.
To characterize this spreading behavior quantitatively, the tran-
sient spreading width bd (t), being dened as the diameter of the
circular interfacial area between the droplet and the substrate, is
measured from the recorded high-speed videos. Fig. 6 shows the
transient spreading width exemplary for individual droplets on
different coatings.
The spreading width on hot-dip galvanized zinc coating evolves
more rapidly and reaches its nal value already after a short period
of approximately 510 ms, while on electro-galvanized coating
the spreading proceeds slower and reaches its nal value after
1520 ms. A plateau value of the spreading width is reached after
510 ms on hot-dip galvanized surfaces and after 2030 ms on
electro-galvanized. This value is almost the nal width, although
the droplet is not completely solidied after this period. The nal
width bd (t ), measured after solidication (also from the high-
speed videos), is given in Fig. 7 for the different coatings and droplet
sizes.
Taking into account the standard deviation, a slight increase of
the nal width with increasing coating thickness can be observed
for a droplet mass of md = 140 mg at room temperature. This trend
cannot be observed in the case of smaller droplets. Especially the
three mean values of the nal width on 6.5 m electro-galvanized
coating are noticeable disturbing any trend. Hence, no clear ten-
dency towards an inuence of the coating thickness on the nal
width can be determined at room temperature. However, at an ele-
vated temperature of the substrate of 400 C (close to the melting

Fig. 5. Wetting process of AlSi12 on electro-galvanized (ZE) and hot-dip galvanized (Z) Zn coating at room temperature of the substrate.
 M. Gatzen et al. / Journal of Materials Processing Technology 214 (2014) 123131 127

10 ZE 50/50
The bulk cooling time lies in the range of 0.5 s up to 1.5 s, depend-
ZE 75/75 ing on the coating. By trend, a slight decrease of the characteristic
Z100
solidication time tSol with increasing coating thickness can be
transient spreading width bd(t)

mm Z140
observed.

6
md ~ 110 mg 3.2. Wetting angle
TS = 23C
4 In Fig. 9 the average wetting angle measured from typically three
bd(t) droplets is given for different coating thicknesses, droplet sizes and
2 substrate temperatures.
The wetting angles resulting after wetting on RT substrate
(TS = 23 C) do not show a distinct trend neither with increasing
0
0 10 20 30 40 ms 60 droplet size nor for increasing coating thickness. The average wet-
ting angles are in the range of 5964 in this case. The effect of
time t
BIAS ID 130645 the pre-heating temperature on the wetting angle is more strik-
ing. With an initial substrate temperature of TS = 400 C the average
Fig. 6. Transient spreading width bd (t) for different coatings measured from indi-
vidual droplet propagations. wetting angle decreases to a value of 2838 . Only in this case a
slight increase of the wetting angle with increasing coating thick-
ness can be observed. However, it is obvious that the investigated
14
md = 90 mg (RT) coating thickness and droplet masses have only little effect on the
md = 110 mg (RT) resulting wetting angle compared to the substrate temperature.
mm
width of solidified droplets bd

md = 140 mg (RT) Moreover, note that without coating no bonding was achieved.
md = 110 mg (PH)
10
3.3. Visual inspection of the joint and joint cross-section

8
6 lic bonding. The droplet on coated steel sheets did not show any
4 interface. The joint between aluminium droplet and hot-dip galva-
0
0 4 6 8 m 12
coating thickness dZn
Fig. 7. Final droplet width bd (t ) after solidication for different coating thick-
ness dZn and droplet masses md on substrates at room temperature (RT) and
pre-heated (PH).
temperature of the zinc layer), an increased coating thickness
seems to have a detrimental effect on wetting and, consequently, on
the width bd of the solidied droplets. It is worth mentioning that
in case of the uncoated specimen at room temperature as well as at
elevated temperature, the droplet did not develop a metallic joint.
The high-speed videos reveal that the droplets are not com-
pletely solidied once the nal width is reached. The upper part
remains liquid for a comparatively long time. The characteristic
time tsol , indicating the bulk cooling of the droplets is shown in
Fig. 8.
1,6
s m = 140 mg
characteristic time tsol
Fig. 10 shows the substrate after droplet solidication. In case
of uncoated steel, the aluminium droplet did not develop a metal-
imperfections both on the spherical cap of the droplet and along the
nized steel revealed a modication of surface appearance around
BIAS ID 130646 the toe of the deposited droplet. This could not be observed on
electro-galvanized steel sheets.
Droplets falling on the substrate at RT show a noticeable poros-
ity in cross-sections after solidication (see Fig. 11). It seems that
the number of pores increases with increasing coating thickness.
However, pre-heating the substrate signicantly decreases the
occurrence of porosity inside the deposited droplets.
3.4. Metallurgical composition
The WDS mapping result for zinc in the centre and at the toe of
the deposited droplets is given in Fig. 12 for a hot-dip galvanized
and electro-galvanized surface.
The measurements reveal an accumulation of zinc near the outer
toe of the droplet. The content reaches a value between 30 wt.%
and 90 wt.% in an distinct triangular area, while in the adjacent
region (about 100 m distance from the triple point in the bulk
droplet) less than 6 wt.% can be observed. At the centre of the
droplet, the zinc is completely removed from the interface. Only
m = 90 mg
a small amount could be measured within the aluminium phase
m = 110 mg
obviously being dissolved in the liquid aluminium during wetting.
The overall amount of zinc seems to vary depending on the initial

coating.
The interface at the centre of the droplet exhibits the typical
0,8 T = 23 C
FeAl intermetallic phase. It consists of a small reaction layer and
some protruding spike structures that extend up to 100 m into
the aluminium phase. It is also worth noting that the small amount
0,4
of zinc in this region is partially accumulated around the Fe spikes
(compare Fig. 12 and Fig. 13).
0,0
-2 0 2 4 6 8 m 12 4. Discussion
coating thickness dZn
BIAS ID 130647
The brazing experiments performed using the developed exper-
Fig. 8. Characteristic time tSol on uncoated and coated steel substrate with different imental model system show that the presence of a zinc coating is a
coatings and at room temperature. prerequisite to achieve wetting of steel substrates with droplets of
128 M. Gatzen et al. / Journal of Materials Processing Technology 214 (2014) 123131

100
md=90 mg (RT)
md=110 mg (RT)

md=140 mg (RT)
md=110 mg (PH)

wetting angle
60

40

20

0
0 4 6 8 m 12
coating thickness dZn
BIAS ID 130648

Fig. 9. Average wetting angle vs. coating thickness for different droplet sizes and substrate temperature. The angle was measured according to the triangulation indicated
in the cross section.

Fig. 10. Droplets deposited on the different zinc-coated substrate surfaces (TS = 23 C). In case of hot-dip galvanized coating a modication of appearance can be observed.

a commercial eutectic aluminiumsilicon wire. This is in line with
observations made with various thermal joining processes for alu-
minium to steel, where typically the presence of a zinc layer on the
steel sheets is required to obtain adequate joints.
As long as a zinc coating is present on the substrate, the result
of the wetting process (as characterized by the width of the
solidied droplet and the wetting angle) at room temperature is
not substantially affected by droplet mass as well as by the type of
the coating and its thickness.
At an elevated temperature, the resulting width of the solidied
droplets is increased for decreasing coating thickness and is larger
for the electro-galvanized than for the hot-dip galvanized speci-
mens (Fig. 7). This should primarily be related to the removal of heat
from the droplet, which is governed by both layer thickness and

Fig. 11. Pore formation in the solidied droplet. Droplets that wetted substrates that were initially at room temperature, revealed a signicant amount of porosity. In the
case of preheated substrate porosity was found to be explicitly lower.
 M. Gatzen et al. / Journal of Materials Processing Technology 214 (2014) 123131 129

Fig. 12. WDS scanning result of zinc distribution at the centre of the droplet and at the toe of the deposited droplet for different coating types after wetting on RT substrate
(TS = 23 C). Zinc is accumulated at the weld toe. At the centre of the droplet the zinc layer was removed from the steel surface and was dissolved in the aluminium.

layer surface structure. As the thermal diffusivity of zinc is approxi-
mately three times the value for steel (aZn = 39.64 mm2 /s (measured
by Frazier (1933)) compared to approx. aSt = 13 mm2 /s (derived
from calculations from Brandes and Brook (1998)), a thicker coating
should contribute to a faster cooling of the droplet on the sub-
strate, and, consequently, a shorter width of the solidied droplet.
This is substantiated by the fact that, for a higher coating thick-
ness, the characteristic solidication time tsol is slightly decreased,
indicating that heat is abstracted faster from the droplet (Fig. 6).
Moreover, in contrast to the hot-dip galvanized specimens, the
electro-galvanized specimens are characterized by a signicant
unevenness (or roughness) of the coating, which should impede
heat transfer from the droplet (at least in the initial stage of con-
tact between coated substrate and droplet), further contributing to
an increase in tsol . This could explain the non-wetting stage during
the initial few milliseconds of the interfacial process (see Fig. 4) and
also the slower propagation of the droplet on the electro-galvanized
coating. However, the experimental results so far do not permit to
distinguish between the effects of layer thickness and layer surface
structure.
In spite of the probable differences in local cooling rates, the
principle evolution of the microstructure in the bonding zone does

Fig. 13. Iron distribution at the interface around the centre of the droplet for hot-dip and electro-galvanized substrate. Spiky FeAl intermetallic phases occurred at the
steel-aluminium interface.
130 M. Gatzen et al. / Journal of Materials Processing Technology 214 (2014) 123131
not differ signicantly for the investigated coating types and thick-
nesses. Rather, two areas can be distinguished for all investigated
zinc layers: a large central area and a small area close to the toe
of the deposited braze metal. Whereas the latter is character-
ized by the presence of signicant amounts of zinc in the braze
metal, the former is nearly free of zinc (and shows slightly dif-
fering intermetallic FeAl structures). This suggests that the zinc
layer is liqueed during wetting of the coated substrate due to
heat transfer from the droplet, a suggestion further substantiated
by the presence of probably zinc-related porosity in the braze metal
deposited at room temperature (Fig. 11), where cooling time is too
short to allow complete evaporation of the zinc from the molten
braze metal.
The fact that the zinc is distributed as a at layer on the steel sur-
face is expected to be the key to the good wettability since it enables
the propagation of the aluminium on an almost liquid interface (the
zinc should be molten almost instantaneously after contact with
the overheated aluminium). The zinc layer, to a certain degree and
for the investigated case, has some effects comparable to commer-
cial uxes which are, e.g., lowering the surface tension of the liquid,
enabling a chemical reaction and reducing the effect of oxide layers
(no oxide layers on the steel, since the coating covers it from ambi-
ent gas). The afnity and solubility between aluminium and zinc in
liquid state enables the physical bonding after the zinc is molten.
The interaction between aluminium and zinc rather potentially is
governed by solubility effects, since the coating is molten almost
immediately over the entire interface. That would affect the free
energy of mixing inside the propagating droplet front and hence
supporting the wetting due to released energy.
In the context of wetting phenomena the observed wetting
dynamics can be identied as reactive wetting processes. How-
ever, the percentage of the reaction process to the overall spreading
kinetics, especially the enthalpy of mixing, e.g., on the result-
ing wetting angle, can hardly be evaluated since the spreading is
stopped by solidication of the droplet.
Since the coating thickness, and hence the local wt.% of zinc
in the propagating aluminium front was found to have no signif-
icant effect on the resulting wetting angle, the local composition
of the mixture on the spreading kinetics seems to play a minor
role in the given case. Moreover, the zinc-rich area close to the
toe of the deposited braze metal constitutes a local AlZn-base
alloy (with up to 90 wt.% zinc) with increased wettability, which
resembles the alloys used in hot-dip galvanizing (so-called gal-
valume with 55 wt.% Al). Ebrill et al. (2000) measured a typical
wetting angle of 25 in isothermal wetting experiments of liquid
alloy 55Al43.4Zn1.6S on steel at 600 C. This is in good correla-
tion with the wetting angles observed for non-isothermal wetting
on pre-heated substrates in our experimental model system and
together with the negligible effect of droplet mass further sub-
stantiates the signicant role of zinc and local alloying with zinc for
the wetting process.
Zhou and de Hosson (1995) have shown that surface roughness
has a signicant impact on the resulting wetting angle, e.g. for alu-
minium spreading on ceramics. Despite the fact that the substrates
have different surface topologies no such inuences were observed
in our case. This is linked to the fact that, in contrast to wetting on
ceramics, in the investigated case no real triple line in a formal sense
is developed (liquidsolidgas-interface), since the outer bound of
the droplet is propagating on a liquid rather than on a solid sur-
face. On the other hand, surface roughness may have an impact on
the heat transfer from the droplet to the coating in the very inital
stage of contact, as already stated earlier. In terms of wetting angle
no difference was found between hot-dip and electro-galvanized
coatings that are found to have different surface topologies.
All in all, the above ndings suggests that the wetting process is
inherently related to the zinc coating on the substrate a generalised
observation which is also made in brazing of joints between zinc-
coated components e.g. using MIG or laser brazing. Basically, and
the presented experimental ndings substantiate this, zinc has to
be liqueed to enable wetting. Wetting shall then start as soon
as a thin layer of zinc is molten along the initial contact zone of
the droplet with the substrate. Once this layer is liqueed, the alu-
minium droplet has a perfect thermal contact to the substrate, and
the heat is transferred through the entire interface, melting the
remaining coating. However, it is not yet fully understood how the
accumulation of zinc in the toe area of the deposited braze metal
can be explained.
In agreement with the line of argumentation of Agudo et al.
(2007), the origin of these accumulations seems to lie in uid
dynamic phenomena, meaning that the zinc is dissolved into the
liquid aluminium along the interface and being pushed by the prop-
agating aluminium melt to the outer bounds of the joint as local
aluminiumzinc alloy. Although comparable observations of zinc-
rich regions are found for a wide variety of different braze metals
(e.g. from Koltsov et al. (2010) for CuSi3 droplets, from Govekar
et al. (2009) for nickel droplets or from Tan et al. (2013) for magne-
sium ller wire), Agudo et al. (2007) attribute the spreading and the
accumulation of zinc to capillary forces caused by different afnities
of Zn to Al and to Fe. However, the velocity of the wetting process
and the wetting angle obtained for pre-heated substrates in our
study (which is comparable to the results from isothermal, reac-
tive wetting processes shown by Ebrill et al. (2000)) rather seem to
indicate that uid dynamics inside the droplet may dominate over
the action of capillary forces.
Independent from the underlying wetting mechanism, the
results obtained with the presented experimental system suggest
that, for practical brazing applications using aluminium ller mate-
rial to join zinc-coated steels, the relatively best brazing should be
achievable under the following conditions:
1. A zinc coating must be present to obtain wetting.
2. The substrate has to be sufciently pre-heated to decrease zinc-
related porosity and obtain small wetting angles.
3. In pre-heated condition the zinc-coating should preferably be
electro-galvanized and as thin as possible to obtain increased
wetting.
5. Conclusions
The non-isothermal wetting process of liquid eutectic
aluminiumsilicon droplets (commercial AlSi12 ller mate-
rial) on uncoated and zinc-coated low-carbon steel substrates
was investigated using an experimental model system capable to
provide liquid droplets with reproducible thermal and dynamic
properties. The major conclusions can be summarised as follows:
1. A zinc coating on the steel substrate is required to obtain wetting
of the aluminiumsilicon braze metal.
2. The thickness or the type of the zinc coating have a slight effect on
the temporal evolution of wetting (in special the characteristic
solidication time), but (at room temperature) do not affect the
result of the wetting process (as characterized by the width of
the solidied droplets and the resulting wetting angle).
3. At elevated substrate temperature, the width of the solidied
droplets is improved for thinner, electro-galvanized zinc lay-
ers compared to thicker, hot-dip galvanized layers. The wetting
angle is not affected.
4. The differences in wetting behavior of aluminiumsilicon
droplets between the various types and thicknesses of zinc lay-
ers on the steel substrates are attributed to differences in the
 M. Gatzen et al. / Journal of Materials Processing Technology 214 (2014) 123131
conditions for heat transfer from the droplet due to zinc layer
thickness and surface structure.
5. Independent from the type and thickness of the zinc layer, an
AlZn alloy is accumulated close to the toe of the deposited braze
metal, whereas in the central area of the bonding zone, no zinc
is found.
6. It is suggested that uid-dynamic effects govern the accumula-
tion of zinc in the toe of the braze metal, whereas the spreading
of the aluminium on the steel is governed by interfacial reaction
between aluminium and steel.
7. With respect to wetting behavior in practical, local brazing pro-
cesses such as MIG or laser brazing using aluminiumsilicon
ller material, thinner, electro-galvanized should perform supe-
rior compared to hot-dip galvanized zinc layers.
Acknowledgements
This work was accomplished within the Centre of Compe-
tence for Welding of Aluminium Alloys CentrAl. Funding by the
Deutsche Forschungsgemeinschaft DFG (Th-1669/1-1) is gratefully
acknowledged. The authors would also like to thank the Insitut fuer
Werkstofftechnik (IWT) in Bremen, Germany, and Mrs P. Meier in
special, for performing the WDS analysis. Moreover, the partners
from industry shall be acknowledged for supplying the zinc-coated
base materials.
References
Agudo, L., Eyidi, D., Schmaranzer, C.H., Arenholz, E., Jank, N., Bruckner, J., Pyzalla, A.R.,
2007. Intermetallic Fex Aly -phase in steel/Al-alloy fusion zone. J. Mater. Sci. 42,
42054214.
Bouche, K., Barbier, F., Coulet, A., 1998. Intermetallic compound layer growth
between solid iron and molten aluminium. Mater. Sci. Eng. A 249,
167175.
Brandes, E.A., Brook, G.B., 1998. Smithells Metals Reference Book. Butterworth-
Heinemann Ltd, Oxford.
Ebrill, N., Durandet, Y., Strezov, L., 2000. Dynamic reactive wetting and its role in
hot dip coating of steel sheet with an AlZnSi alloy. Met. Mater. Trans. B 31B,
10691079.
131
Frazier, R.H., 1933. A precise determination of the thermal diffusivity of zinc. Phys.
Rev. 43, 135136.
Giorgi, M.-L., Guillot, J.-B., Nicolle, R., 2005. Theoretical model of the interfacial
reactions between solid iron and liquid zincaluminium alloy. J. Mater. Sci. 40,
22632268.
Govekar, E., Jeric, A., Weigl, M., Schmidt, M., 2009. Laser droplet generation: appli-
cation to droplet joining. CIRP Ann. Manuf. Technol. 58, 205208.
Koltsov, A., Bailly, N., Cretteur, L., 2010. Wetting and laser brazing of Zn-coated steel
products by CuSi ller metal. J. Mater. Sci. 45, 21182125.
Kreimeyer, M., 2007. Verfahrenstechnische Voraussetzungen zur Integration
von Aluminium-Stahl-Mischverbindungen in den Kraftfahrzeugbau, vol. 30.
Strahltechnik, BIAS Verlag, Bremen, Germany.
Marder, A.R., 2005. The metallurgy of zinc-coated steel. Prog. Mater. Sci. 45, 191271.
Moeller, F., Grden, M., Thomy, C., Vollertsen, F., 2011. Combined laser beam weld-
ing and brazing process of aluminiumtitanium hybrid structures. In: Schmidt,
M., Zaeh, F., Graf, T., Ostendorf, A. (Eds.), Lasers in Manufacturing (LIM2011).
Elsevier, Amsterdam, pp. 215223.
Murray, J.L., McAlister, A.J., 1984. The AlSi (aluminiumsilicon) system. Bull. Alloy
Phase Diagrams 6 (1), 7484.
Peyre, P., Sierra, G., Deschaux-Beaume, F., Stuart, D., Fras, G., 2007. Generation of
aluminiumsteel joints with laser-induced reactive wetting. Mater. Sci. Eng. A
444, 327338.
Popoola, A.P.I., Fayomi, O.S.I., 2011. Effect of some process variables on zinc coated
low carbon steel substrates. Sci. Res. Essay 6 (20), 42644272.
Schmidt, M., Otto, A., Kgeler, C., 2008. Analysis of YAG laser lap-welding of zinc
coated steel sheets. CIRP Ann. Manuf. Technol. 57, 213216.
Song, G.M., Vystavel, T., van der Pers, N., De Hosson, J.Th.M., Sloof, W.G., 2012. Rela-
tion between microstructure and adheision of hot dip galvanized zinc coatings
on dual phase steel. Acta Mater. 60, 29732981.
Tan, C.W., Li, L.Q., Chen, Y.B., Mei, C.X., Guo, W., 2013. Interfacial microstructure
and fracture behavior of laser welded-brazed Mg alloy to zinc coated steel. Int. J.
Adv. Manuf. Technol., http://dx.doi.org/10.1007/s00170-013-4910-4 (published
online).
Thomy, C., 2009. Hybrid laser-arc welding of dissimilar materials. In: Flemming,
O.O (Ed.), Hybrid Laser-Arc Welding. Woodhead Publishing Ltd., Cambridge, pp.
270295.
Thomy, C., Vollertsen, F., 2012. Laser-MIG hybrid welding of aluminium to steel
effect of process parameters on joint properties. Welding World 56 (5/6),
124132.
Vollertsen, F., Thomy, C., 2009. On the meaning of zinc coatings in hybrid welding
of aluminium to steel. Proc. 28th Cong. On Application of Laser & Electro-Optics
(ICALEO09), LIA Publication 612 (CD) paper 107, 89-97.
Zhang, H., Liu, J., 2011. Microstructure characteristics and mechanical property of
aluminium alloy/stainless steel lap joints fabricated by MIG weldingbrazing
process. Mater. Sci. Eng. A 528, 61796185.
Zhou, X.B., De Hosson, J.Th.M., 1995. Inuence of the surface roughness on the wet-
ting angle. J. Mater. Res. 10 (8), 19841992.