-
Metallographic Preparation Technique for Hot-Dip Galvanized and
Galvannealed Coatings on Steel C. E. Jordan, K. M. Goggins, A. O.
Benscoter, and A. R. Marder Lehigh University, Materials Science
and Engineering Department, Bethlehem, PA 18015
A new metallographic technique for hot-dip galvanized and
galvannealed coatings has been developed. The new polishing
procedure and etchant have shown excellent results on commercial
hot-dip galvanized and galvanneal coatings, as well as on
laboratory-simulated hot-dip galvanneal produced under a variety of
thermal processing parameters.
INTRODUCTION
Galvanized coatings have been used for many years, providing
sacrificial and anodic corrosion protection of steel. Zinc can be
de- posited onto steel by a number of different processes,
including hot-dipping, electro- deposition, and vapor deposition.
Galvan- neal is a galvanized coating that has under- gone an
annealing cycle to transform the almost all zinc coating to an
alloyed iron- zinc coating. Hot-dip galvanneal has had ex- panded
use in car body parts in the auto- motive industry because of its
improved spot weldability and perforation corrosion resis- tance
over that of hot-dip galvanized coat- ings [1]. Because of the
increased use of hot-dip galvanneal by automobile manufac- turers,
there has been new interest in the research and development of the
older and less costly hot-dip zinc coating process.
Metallographic inspection of these coat- ings provides a useful
tool in the character- ization of the iron-zinc phase layer growth
that occurs during the galvannealing pro- cess. Metallography alone
cannot determine the identity of the phases present, but it can
provide useful information when used in conjunction with other
characterization techniques.
Almost 45 years ago, Rowland [2] made a significant contribution
to the technique
107
of metallographic preparation and etching of hot-dip galvanized
and galvannealed coat- ings. In that work he discussed a number of
etchants to be used on coatings depending upon their immersion
time, chemical com- position, and thermal history. The etchants
developed by Rowland were color etchants, which could be used to
identify phase layers within the coating based on the color differ-
ence between adjacent phase layers. Row- land specified the use of
different concen- trations of picric acid, ethyl alcohol, and water
to etch short- and long-time immer- sion hot-dip galvanized
coatings, galvan- nealed coatings, as well as coatings con- taining
aluminum. He also developed two alternative solutions for the
etching of gal- vanized coatings containing aluminum. Rowland's
work in color etchants has since been developed further by
Kilpatrick [3].
The coatings discussed in this article are short-time immersion
coatings that are ap- proximately 10~m in thickness. Metallo-
graphic preparation of thin zinc coatings can be difficult because
the outer edges of the relatively soft coating can become rounded
during grinding and polishing, thus making examination difficult.
Etching of these coat- ings can also be a problem because of the
small anode to cathode reaction area ratio, which causes the zinc
coating (anode) to react rapidly in acidic solutions. Other in-
Elsevier Science Publishing Co., Inc., 1993 MATERIALS
CHARACTERIZATION 31:107-114 (1993) 655 Avenue of the Americas, New
York, NY 10010 1044-5803/93/$6.00
-
108 C. E. Jordan et al.
vestigators such as Giallourakis et al. [4] have made an attempt
to avoid the difficulties of metallographic preparation and etching
by developing a cryogenic fracture technique for the
characterization of zinc coatings. This technique avoids the need
for polishing and etching of the zinc coating altogether.
A new etchant based on Rowland's work has been developed that is
better suited for today's thinner, short-time immersion, hot- dip
coatings. The etchant was found to work well for hot-dip galvanized
coatings contain- ing 0.00-0.15 effective wt.% aluminum [5] that
were deposited on a number of differ- ent steel substrates. The
etchant also per- formed well for coatings that were annealed under
a variety of temperature/time condi- tions. The preparation
technique also uses, in part, the work of Drewein et al. [6].
Although Drewein's techniques were devel- oped for electrodeposited
coatings, modihca- tions have been made for improved struc- tural
analysis of hot-dip zinc coatings in the present investigation.
PROCEDURE
SECTIONING
For standard 31.75mm (1.25in.) mounts, the sheet samples are cut
to 25 x 13ram-size sec- tions. Sectioning can be performed using a
tabletop hand shear so that the sheet can be accurately cut to
size. The samples are then placed to form a stack (Fig. 1) with the
25mm-long freshly cut edges parallel to one
another. The stack is assembled so that the long edges are
aligned and flush with one another. The flush orientation of the
samples is important during rough grinding where at least 2mm of
material must be removed from each sample in the mount, and
alignment ensures this. If one side of each sheet sample is of
particular interest, it is necessary to form the stack so that the
sides of interest are all facing in the same direction. The reason
for this orientation will be addressed later.
Samples can be separated from one an- other by placing a small
piece of double-stick tape (spacer) at each short end of the
sample, away from the edge of interest, as shown in Fig. 2. Any
spacer that separates the sheet samples but keeps their distance
apart to a minimum is suitable. At least six to eight sheet samples
should be used in each stack. Two additional dummy samples are
needed, one on each side of the stack, to maintain coating flatness
of the end samples. Because the epoxy resin used for mounting is
soft, stabilizers, such as two cut pieces of steel welding rod
material, should be placed on either side of the stack to ensure
mount flat- ness during grinding and polishing. As a point of
reference, an indicator (scrap steel material) can also be included
in the mount (Fig. 1) prior to filling the mould. Epoxy resin and
hardener are then used as the mount- ing media.
GRINDING
When the mount has cured, excess epoxy is ground off the surface
of the mount until the metal surfaces of all the samples have
Edges of I n ~ ~ 0 ~ Indicator
v f \
pacs* i / dmm, \ * , = / samp,s
%ox; . ' / ,=ta, M = ~ " stabilizer
FIG. 1, Planar view of the mount showing the stack arrangement
of the coated sheet samples.
Spacers
Sheet .. sample
' ~ Edge of Interest
FIG. 2. Individual sheet sample in the process of be- ing
incorporated into the stack arrangement.
-
Preparation for Hot-Dip Coatings 109
been completely exposed. The thickness of the mount is then
measured using a mi- crometer. The mount is rough ground to re-
move at least 2mm of material, so that the deformation introduced
into the coating during sectioning has been removed. The amount of
material removed can be routinely checked during grinding if one
uses the mi- crometer to monitor the thickness of the mount. When a
belt grinder is used for rough grinding, the sheet samples must be
kept parallel to the direction of the belt dur- ing grinding to
prevent edge rounding of the samples. Edge flatness is critical for
ana- lyzing the zinc coating located at the outer- most edges of
the steel sheet samples.
If the paper used in rough grinding was 120 grit, then grinding
should continue on 240-, 320-, 400-, and 600-grit papers, in that
order. The last step in grinding should leave scratches parallel to
the long edges of the samples to minimize rounding. The mount
should be moved laterally back and forth in the middle region of
the spinning wheel, off center, with the long edges of the stack
either perpendicular or parallel to the direc- tion of spin of the
wheel [6]. The mount should then be rotated 90 and held in a sim-
ilar manner while one grinds on a new grade of grit paper; thus,
the new scratches are perpendicular to those of the previous step.
This method of grinding guarantees that all of the scratches from
the previous step have been removed. During grinding, the edge of
interest is either the leading edge (first edge to encounter the
motion of the paper) perpendicular to the spin of the wheel, or it
is parallel to the spin of the wheel, with the edge of interest
closest to the center of the wheel. The placement of the lead- ing
edges in this manner again minimizes rounding the edges of
interest. After grind- ing on each paper, the surface is flushed
with alcohol, and the mount is blown dry and inspected under a
light optical microscope to ensure that (1) all of the scratches
are uni- form in direction in all of the samples in the mount, and
(2) that no scratches remain from the previous grinding step. The
previously described procedure can be performed on an automatic
grinder (with an applied load
of 25psi) by grinding for 60-90 s on each grade of grit
paper.
Immediately after the sample is removed from the 600-grit paper,
the surface is swabbed with an alcohol saturated cotton ball, and
the mount is flushed with alcohol. Ethanol (190 or 200 proof) or
denatured al- cohol is suitable for cleaning purposes. The mount is
ultrasonically cleaned for 30-60 s while standing the mount on edge
in a beaker of alcohol. Precautions should be taken to be sure that
all of the samples in the mount are submerged in the alcohol dur-
ing ultrasonic cleaning. The mount is blown dry and inspected.
Immediate cleaning of the mount in alcohol is crucial in maintain-
ing a clean, corrosion-free sample. Ultrafine grinding continues on
an 8- and then 3~m SiC papers (it is the author's preference to use
8- and 3~tm papers, but 12- and 5~m papers are also appropriate for
fine grind- ing), and then the mount is cleaned with al- cohol (as
described earlier) after each paper.
POLISHING
Polishing can begin with a stationary nap- less cloth similar to
the Leco Pan W cloth, impregnated with 3~tm diamond paste. Engis
diamond extender solution is suitable as a lubricating media for
this and all sub- sequent polishing steps. A diamond slurry and
extender (pH = 9.6 + 0.2) combination has also proven to be a
successful polishing media [6]. If a paste is used instead of a
slurry, a dummy mount should be used to work the paste into the new
polishing cloth. Using a dummy mount prior to the actual mount will
prevent large scratches from be- ing introduced into the samples.
The mount should be rotated in a clockwise direction applying a
heavy, even pressure. Polishing continues for I minute, and then
the mount is cleaned and examined under the light mi- croscope. The
scratches should appear in all directions, with no parallel
scratches remain- ing from the last grinding step. This proce- dure
is repeated using a new Pan W cloth impregnated with l~tm diamond
paste. Upon examination after this step, the scratches present
should appear finer.
-
1 1 0
Polishing can then be continued on a sta- tionary Struers DP NAP
cloth charged with l~m paste, or Struers DP spray, HQ. This
polishing cloth need only be charged infre- quently and can be
covered and stored for future polishing of coatings. A heavy even
pressure must be applied for 30 s, and then the mount should be
cleaned and examined. The samples should be almost free of
scratches. If large, significant scratches re- main, polishing
should continue for an ad- ditional 20-30 s, followed by cleaning
and examination. The finish polishing step is per- formed on a
separate Struers DP NAP cloth charged with 0.25~m diamond paste, or
0.25~m DP-spray, HQ. Heavy pressure for 20 s is required followed
by cleaning and examination. The samples should now be ready for
etching.
ETCHING
The etchant to be used should be prepared prior to the start of
any polishing procedures so that the sample can be etched at room
temperature immediately after polishing has been completed. The
etchant found to give the best results was a mixture consisting of
1% picric acid in amyl alcohol and 1% nitric acid in amyl alcohol.
The solution is-pre- pared by mixing equal parts of 1% picric acid
in amyl alcohol and 1% nitric acid in amyl alcohol in a beaker.
Equal amounts of the mixed solution are poured into two crucibles.
Into one crucible, 3-4 drops of hydrofluoric acid (to approximately
50ml of solution) is added, and a beaker of ethanol is placed near
the two crucibles. It is critical that the etch- ing solution be
prepared with amyl alcohol and not ethanol. Amyl alcohol-based
etch- ants etch more slowly than ethanol-based mixtures, thus
allowing for more control dur- ing etching [6].
To etch the samples, the mount is held with tongs so that the
metal surfaces of the samples face upward. The mount is im- mersed
into the crucible containing no hydrofluoric acid, and it is
slightly agitated for approximately 20 s. The sample is re- moved
from the crucible and immediately
C. E. Jordan et al.
placed (metal surface side up) into the beaker containing
ethanol. The mount is then re- moved, the surface flushed with
ethanol, and then immersed into the second crucible (HF added) and
slightly agitated for 10 s. The surface is flushed again with
ethanol, blown dry, and examined. If the samples are underetched,
the previous procedure is re- peated using the ratio of 2:1 for the
etch- ing time of the first solution to that of the second
solution.
RESULTS
The etched coatings are shown in Figs. 3, 4, and 5(a), and are
approximately 8-10,m in thickness. Figure 3 is a hot-dip galvanized
coating, Fig. 4(a-e) shows simulated gal- vanneal coatings, and
Fig. 5(a) is a commer- cial galvanneal product. All three types of
coat ings-hot-dip galvanized, simulated galvanneal, and commercial
galvanneal- exhibited good relief of structure using the described
preparation technique and etch- ant. Nomarski differential
interference con- trast in the light microscope allowed the
topographical features of the coatings in cross section to be
viewed.
Figure 6 is an x-ray spectrum of intensity versus 20 values of
the hot-dip galvanized coating shown in Fig. 3. The major peaks at
36.4 , 39.10, and 77.1 correspond to d spacings of 24.6, 23.0, and
12.4nm that are
FIG. 3. Cross section of a hot-dip galvanized coating (0.10
effective wt.% A1-Zn) deposited o n t o a drawing quality special
killed (DQSK) steel.
-
Preparation for Hot-Dip Coatings 111
(a) (b)
(c) (d)
FIG. 4. Cross section of a hot-dip galvanized coating deposited
onto a titanium stabilized interstitial free steel that was
annealed for (a) 1 s at 450C; (b) 5 s at 450C; (c) 10 s at 450C;
(d) 20 s at 450C; and (e) 60 s at 450C.
(e)
consistent with those for the zinc-rich Fe-Zn eta phase.
Therefore, the x-ray data indicate that the as-galvanized coating
contained es- sentially all eta phase. Examination of the coating
using wavelength dispersive spec- troscopy (WDS) analysis confirmed
the pres- ence of blocky crystals of an i ron-aluminum- zinc
intermetallic c o m p o u n d located at the coating/steel
interface.
Figure 4(a-e) shows examples of the sim- ulated galvanneal
coatings generated during thermal processing of an as-galvanized
coat-
ing (like that shown in Fig. 3) deposi ted onto an interstitial
free (IF) steel. Because the gal- vanized coating has undergone a
diffusional transformation upon heating, these coatings have
developed a more complex structure of iron-zinc phases. The x-ray
spec t rum presented in Fig. 7 was obtained from the coating shown
in Fig. 4(d) and shows that the largest intensity peaks occur in
the 2(9 range of 40-45 . High-intensity peaks for the Fe-Zn gamma,
delta, and zeta phases all occur at d spacings that cor respond to
this
-
112 C. E. Jordan et al.
(a) (b)
FIG. 5. (a) Cross section of a commercial galvanneal coating.
(b) Scanning electron image of the surface structure of the
commercial galvanneal coating shown in (a). (See text for
explanation of arrows.)
region. Therefore, it is difficult to identify and quantify the
phases present in the gal- vannealed (alloyed) coating by
conventional x-ray analysis.
For the shorter hold time annealed coat- ings [Fig. 4(a-c)], an
eta phase layer remains at the outermost part of the coating. Below
this layer there is thought to be a layer of blocky zeta phase
crystals, and a layer of columnar grains of delta phase. The longer
hold time annealed coatings show no eta phase remaining in the
coating; instead, they show the presence of a gamma Fe-Zn phase
layer at the coating/steel interface. Also present in these
longer hold time coatings [Fig. 4(d, e)] are cracks in the delta
and gamma phases, running perpendicular to the coating/steel
interface. Similar metallographic results were found for coatings
deposited on DQSK (drawing quality special killed), DQSK
preannealed, ultra-low carbon, and rephosphorized steel substrates.
The com- mercial galvanneal in Fig. 5(a) is very simi- lar in
appearance to the longer hold time simulation coatings such as
shown in Fig. 4(d). In the commercial product, a gamma
2 .00 ~..62 :i. 28 0 , 9 8 0 , 7 2 0 . 5 0 0 . 3 2 0.~.8 0 . 0 8
0 . 0 2
2 5 . 0
2.00 l 1.62 i .2B 0 . 9 8 0 . 7 2
0 . 5 0 t 0 . 3 2
O'~B I 0.08 0 . 0 2 I
55 .0
30 .0 35 .0
i i 9 . ,
6 55 ,0 0 .0
40 .0 45 .0 5 0 , 0 55 .0
70'.0 75 .0 00 .0 85 .0
FIG. 6. X-ray diffraction spectrum of inten- sity (counts)
versus 20 values of the as- galvanized coating in Fig. 3.
-
Preparation for Hot-Dip Coatings
xiO 3 i
113
5.00 l
4.05 1
3.20 1 2.45 1.80 ! 1.25 O. 80
O. 45
0.20
0.05
25.0 30.0 35.0 40.0 45.0 50.0 55.0
5.00
4.05 3.20
I 2.45 I ,80 ' I . 25
0.80
00"45,20 I I I I I I I I I I I I I I I I I I I I I I I 0.05
55.0 55.0 70.0 75.0 80.0 85.0 50.0
FIG. 7. X-ray diffraction spectrum of intensity (counts) versus
2f) values of the simulated galvannealed coating in Fig. 4(d).
phase layer is present as well as cracks per- pendicular to the
coating/steel interface.
DISCUSSION
One question that has been raised about the metallographic
procedure is whether the cracks observed in the microstructure were
created by the technique itself, because of the brittle nature of
the Fe-Zn phases that form, or are the cracks truly a
characteristic of the coating. Because the gamma phase is known to
be one of the most brittle of the Fe-Zn phases that develop [7], it
has been proposed that the cracks may have been gen- erated during
polishing. It has been sug- gested that because of an excess of
applied pressure during polishing and the brittle nature of the
gamma phase, the gamma phase initially undergoes cracking. The
cracks can then propagate along columnar delta phase boundaries in
the coating to form
cracks that extend the width of the coating. However, scanning
electron microscopy of the surface of the coating, where no sample
preparation had been performed, revealed that cracks were present,
as shown in Fig. 5(b) (arrows). This hgure is a scanning elec- tron
image of the surface of the coating after heat treatment but prior
to sample prepara- tion. Thus, the cracks appear to be present
prior to metallographic sample preparation. It has been found,
however, that upon further polishing of the long hold time sim-
ulation samples [where a signihcant gamma layer is present, as in
Fig. 4(e)], the coating can become more cracked and difficult to
work with. Care must be taken not to over- polish the samples.
The most essential part of the metallo- graphic technique
presented here is to keep the sample surface extremely flat and
clean. During the last steps of grinding and all throughout
polishing, the samples should be kept free of water, which can
cause cor-
-
114 C. E. Jordan et al.
rosion of the coating. Precautions should also be taken to
maintain edge flatness. The coatings are approximately 8-10~m in
thick- ness and have a lower hardness than that of the substrate
steel; therefore, to have the entire cross section of the coating
in focus, the softer outer edge of the coating must be as fiat as
possible relative to the substrate steel. The stack orientation of
the samples helps to reduce this problem. Nomarski differential
interference contrast also proved helpful in revealing the
topography or tex- ture of the coatings in cross section.
SUMMARY
A new etchant for hot-dip galvanized coat- ings has been
developed. It has proven successful for hot-dip galvanized,
laboratory- simulated galvanneal, and commercial gal- vanneal
coatings. The maintenance of coat- ing sample flatness and
cleanliness was found to be critical in the metallography of
hot-dip galvanized and galvannealed coatings.
References
1. Y. Hisamatsu, Science and Technology of Zinc and Zinc Alloyed
Coated Steel Sheet, Proc. Galvatech '89, The Iron and Steel
Institute of Japan, Tokyo, Japan, p. 3 (1989).
2. D.H. Rowland, Metallography of hot-dipped gal- vanized
coatings, Trans. ASM 40:983 (1948).
3. J. R. Kilpatrick, A new etching technique for gal- vanneal
and hot-dipped galvanized coatings, Prac- tical Metallography
28:649 (1991).
4. N. M. Giallourakis, D. K. Matlock, and G. Krauss, A cryogenic
fracture technique for characterizing zinc-coated steels,
Metallography 23:209 (1989).
5. S. Belisle, V. Lezon, and M. Gagne, The Solubility of Iron in
Continuous Hot-Dip Galvanizing Baths, 21st Meeting of the
Galvanizers Association, Monterrey, Mexico (October 1989).
6. C. A. Drewein, A. O. Benscoter, and A. R. Marder,
Metallographic preparation technique for electro- deposited
iron-zinc alloy coatings on steel, Materi- als Characterization
26:45 (1991).
7. G. E Bastin, E van Loo, and G. D. Reick, A new compound in
the iron zinc system, Z. Metalkunde 65:656 (1974).
The authors thank National Steel, Armco Steel, LTV Steel,
Dofasco, Rouge Steel, Cockerill Sambre, and Noranda for their
sponsorship of this work. Received November 1992; accepted May
1993.