Analysis of oxide defect formation on GI exposed outer body

automotive panels during continuous hot dip galvanizing

L. Berry1, C. Phillips

2, D Penney

3

1 College of Engineering, Swansea University,

Singleton Campus, Swansea, Wales

Phone: 07885826063

Email: lewis.berry@tatasteel.com; 748386@swansea.ac.uk

2 ZODIAC, Tata Steel Llanwern,

Tata Steel Strip Products, Wales

Phone: 07743843097

Email: craig.h.phillips@tatasteel.com;

3 College of Engineering, Swansea University,

Singleton Campus, Swansea, Wales

Phone: 07872606706

Email: D.Penney@swansea.ac.uk

ABSTRACT

With increased demand for GI outer panel material for the automotive sector, continuous improvement of zinc coating quality is

required. Various surface defects arise from poorly understood and uncontrolled snout practices during coating on a CGL. One such

defect produced within the ZODIAC snout, the arrowhead defect, has been characterised with use of SEM/EDX and XRD and

compared with contamination collected from the ZODIAC snout. The presence of ZnO, originating from the injection of wet HNx into

the snout as a method to reduce the formation of zinc vapour, was observed in regions consistent with the head end of the arrowhead

defect. Discrete intermetallic dross particles were entrained within the tail of the defect. XRD patterns of both the defect & snout

contamination have been presented to discern the nature of the contamination entrained within the zinc coating.

Keywords: Surface analysis, GI Full Finish, Defect, Snout Management, Quality

INTRODUCTION

The increased usage of GI full finish galvanized coated steel sheet for exposed outer body automotive panels has led to the need for

improved surface quality of GI material. The coating section of a Continuous Galvanizing Line (CGL) is the source of numerous

coating defects and various processes have been introduced in order to improve surface quality . Different characterisation techniques

can be employed to identify root causes of zinc coating defect 1)

, and act as a first step in quality centred process improvement. The

production of zinc vapour under a reducing HNx atmosphere within the snout 2)

, along with the formation of intermetallic dross 3,4)

are

two of the primary origins for coating defects. Whilst the formation of intermetallic dross particles are still a major source of defective

coatings, the introduction of wet HNx snout atmospheres have all but eradicated defects that originate from zinc evaporation. Zinc

evaporation is an undesirable process since gaseous zinc will condense. in the snout area and contaminate the incoming steel strip. The

subsequent zinc particles alloy with the steel surface forming intermetallics that are not readily wetted by the liquid zinc within the

zinc pot. The high dew point within the snout resulting from the injection of wet HNx promotes the spontaneous formation of a

protective barrier of zinc oxide on the surface of the zinc bath within the snout, as indicated in Fig. 1a thereby minimizing zinc

evaporation. Kim et al 5)

showed that the introduction of wet HNx into the snout would decreased the formation of zinc dust, formed

through oxidation of zinc vapour, to essentially zero. They investigated the relationship between the injected dew point of wet HNx

and the formation of bare spots which form due to localized surface intermetallics formed through interactions between the moving

steel and zinc vapour. Theoretically, the vaporisation of molten zinc is still possible through the porous oxide barrier and needs to be

further investigated in order to determine optimum snout conditions for the production of GI full finish. Whilst the introduction of wet

HNx does act as a method of reducing zinc vapour formation, the uncontrolled formation of barrier oxides can themselves induce

coating quality issues.

If the zinc oxide layer within the snout

becomes too large, zinc oxide particles

break away and become entrained within

the coating through the action of the

rapidly moving strip. It is therefore

necessary to control the size and thickness

of the zinc oxide layer through the use of

pumps that remove the zinc and a portion

of the oxide barrier layer from within the

snout. The use of pumps in the continuous

removal of surface oxides formed

produced through the introduction of wet

HNx atmospheres was introduced by

Becherer 6)

and resulted in the elimination

of zinc dust related defects. The Push/Pull

type configuration was optimised by

Phillips et al 7)

utilizing water modelling

to discern optimum settings for the removal of contamination. However, the use of a centrifugal

snout pump Push/Pull type configuration does not provide the required control of the bath surface

inside of the snout to produce high quality Full Finish material.

The uncontrolled formation of surface oxides has resulted in the formation of the arrowhead defect

highlighted in Fig. 2. The concentration of arrowhead defects through a Full Finish campaign can be

linked to reduced removal of surface oxides from the snout surface. The mechanism of formation of

the arrowhead defect can be observed with use of the ZODIAC snout camera, whereby the attraction

of surface oxides to the moving strip causes the contamination to be entrained within the galvanized

coating mainly affecting the prime surface coating, as observed in Fig. 3.

EXPERIMENTAL

Cross Sectional Analysis - Samples were cross sectioned using a Beuhler IsoMet 5000 linear precision saw. The contaminated

cross section was then polished to a 0.3μm finish in a 0.3μm Beuhler diamond based polishing suspension and Kemet Chem H

polishing cloth.

Incremental etching - The region of coating surrounding the arrowhead defect was indented with a Vickers hardness indenter as a

method of coating weight determination and incrementally etched with 5% Nital (Nitric acid in Ethanol) at 2 second intervals.

Once etched, the sample was immediately quenched with ethanol and dried.

Scanning Electron Microscopy and Energy Dispersive X-ray Spectroscopy - Samples were analysed using a Hitachi S-4800

Field Emission Gun Scanning Electron Microscope (FEG SEM) with an Oxford instruments Silicon Drift X-Max EDX Detector

(SDD) and Inca EDX software

White Light interferometry - Samples were analysed using s Veeko Wyko NT 9300 White Light Interferometer and subjected to

a stitched scan with a scan length of 80μm and a back scan length of 40μm at 5x optical magnification.

Figure 2 - Arrowhead defect

Figure 1a & b - Cross section of CGL snout indicating relationship between dry/wet

HNx environments, zinc vaporisation and the entrainment of surface oxides into zinc

coating

a b

Figure 3 - Arrowhead defect formation through entrainment of surface oxides into the coating due to attraction to the moving steel

X-Ray Diffraction - Samples were analysed with a Bruker D8 Discover Diffractomer using Cu Kα radiation (λ = 0.15406nm).

Samples were subjected to a hour 11000 step D c cle with step parameters of 0.5 seconds per step and 0.00 06 0

increment per step.

RESULTS & DISCUSSION

Physical characteristics of the Arrowhead defect

The characteristic microscopic topography of the defect under SEM analysis can be shown to be a consequence of microfolding of the

zinc coating, Fig. 4, consistent with that observed for the entrainment of other forms of contamination present during gas wiping 5)

.

The folding of the surface suggests the contamination present within the coating provides an obstacle to ideal coating weight control

during wiping. White light Interfereomtery can be used to map the topography of the defective surface, Fig. 5, indicating an outer

region of low surface roughness encompassing a region of increased roughness throughout the body of the defect. The increase in

roughness suggests a thicker zinc coating within the body of the defect in comparison with the smoother exterior of the defect,

whereby the thicker coating is subject to greater transfer of roll roughness during tempering 1)

.

Metallographic cross sections

In order to discern root cause of formation for the arrowhead defect the defect was cross sectioned and polished in a 0.3μm diamond

polishing suspension to provide a scratch free surface for SEM/EDX analysis. The Fe2Al5/Zn coating interface for regions of non

defective, Fig. 6, and arrowhead defective coatings, Fig. 7, were compared and contamination was observed within the defect sample.

EDX analysis of the contamination indicated regions of high concentrations of zinc and oxygen, Fig. 8. The presence of zinc and

oxygen containing species at the Fe2Al5 interfacial layer within the zinc coating suggests the entrainment of surface oxides from inside

the snout. Previous instances of oxide containing defective coatings were investigated by Vlot 1)

but associated with the entrainment of

oxidized zinc vapour as opposed to the entrainment of oxidized molten zinc. No appreciable amount of aluminium was observed

suggesting the formation of a ZnO barrier layer rather than ZnAl2O4, both of which would be possible given the dew point of the

injected wet HNx at ZODIAC.

Figure 5 - White Light Interferogram indicating

exterior of decreased roughness surrounding a high

roughness interior

Figure 7 - Cross section of Arrowhead defective GI Full Finish

indicating contamination at the Fe2Al5/Zn coating interface

EDX Colour assignment - Green = Iron, Blue = Zinc, Red =

Oxygen

Figure 6 - Cross section of non defective ZODIAC GI Full

Finish indicating no contamination at the Fe2Al5/Zn coating

interface

EDX Colour assignment - Green = Iron, Blue = Zinc, Red =

Oxygen

Figure 4 - SEM imaging of the arrowhead defect highlighting

microscopic folding of the zinc coating around the extremity of

the defect relative to microscopic uncoated (bare) spots

Incremental etch analysis

In line with previous work undertaken by Tata Steel R&D which indicated the preferential dissolution of the Zn coating with retention

of ceramic type species, the arrowhead defect was incrementally etched in 5% nital solution at 4 second intervals and then quenched

with ethanol. The region of coating surrounding the arrowhead defect was indented with a Vickers hardness indenter as a method of

coating weight determination and to allow for imaging over a known region for comparison post etch

Benchmark SEM/EDX

Initial SEM/EDX was carried out over the arrowhead defect in order to compare the same region of the defect pre and post etch. Initial

benchmark indicated regions of increased concentration of iron due to decreased zinc coating thickness and reduced surface roughness 1)

, including regions of oxide relative to microfolding of the zinc coating, Fig. 9.

8 second Incremental SEM/EDX

Figure 8 - EDX spectra of contamination observed at the Fe2Al5/Zn coating interface indicating

regions of high concentration of Zn & O

Figure 9 - Initial benchmark SEM/EDX of the arrowhead defect

EDX Colour assignment - Green = Iron, Blue = Zinc, Red = Oxygen

Figure 10 - Arrowhead defect region after 8 second incremental etch

EDX Colour assignment - Green = Iron, Blue = Zinc, Red = Oxygen

After a total etch of 8 seconds, SEM/EDX indicated the presence of intermetallic top dross within the coating, Fig 10. Due to the size

of the individual top dross particles no microfolding of the zinc coating was observed suggesting that improper top dross control

within the zinc pot was not the origin of the arrowhead defect. Whilst the majority of the sample was still covered by a thick zinc

coating, at the head end of the defect the interfacial Fe2Al5 was exposed after a relatively short etching period in line with phenomena

described Vlot et al 1)

which resulted in regions of decreased surface roughness due to decreased coating thickness. Furthermore the

presence of oxide rich regions within the zinc coating was observed relative to the microfolding of the zinc coating previously

described, Fig 10 & 11. This suggest that the presence of the zinc oxide rich contamination is the root cause in the formation of the

irregular folding of the coating and therefore the arrowhead defect

16 second Incremental SEM/EDX

After a total etch of 16 seconds, SEM/EDX indicated the total dissolution of the zinc coating at the head end of the defect. The

presence of intermetallic top dross particles was again apparent with no irregular flow of the zinc coating during wiping being

observed around the top dross, Fig 12. However, an interface between regions of thick and thin coating is observed with a discrete

zinc oxide particles observed at the interface as indicated in Fig. 13.

Figure 11 - EDX spectra of contamination observed within the zinc coating relatively to microfolding

characteristic of the arrowhead defect

Figure 12 - Arrowhead defect region after 8 second incremental etch

EDX Colour assignment - Green = Iron, Blue = Zinc, Red = Oxygen

Figure 13 - EDX spectra of contamination observed within the zinc coating relatively to microfolding

characteristic of the arrowhead defect

XRD characterisation of the arrowhead defect

The nature of the contamination observed within the zinc coating can be conclusively characterised with the use of X-Ray Diffraction.

The irradiation of the defective sample results in the characteristic scattering patterns through the interaction of the characteristic

crystal planes present within the contamination according to Bragg's law. XRD analysis of the arrowhead defect containing Full Finish

sample produced the X-ray diffractogram highlighted in Fig. 14.

The majority of the peaks oserved within the arrowhead XRD spectra are associated with the zinc coating which dominate over the

peaks associated with the contamination.Small contamination peaks have been characterised through comparison with known XRD

scattering patterns of the types of contamination that would be observed around the snout of a galvanizing line 8-12)

. The spectra has

been compared against known XRD spectra for the Fe2Al5 interfacial layer, intermetallic Fe2Al5Znx top dross, zinc oxide (ZnO),

alumina (Al2O3) and zinc aluminate (ZnAl2O4). The small peaks were used to characterise the presence of the Fe2Al5 interfacial layer,

zinc oxide and intermetallic dross within the tail of the arrowhead defect. Zinc oxide can be formed through two different mechanisms

as indicated previously within the snout, through oxidation of the molten zinc surface with wet HNx or oxidation of zinc vapour, with

only physical characteristics used to differentiate the nature of the contamination.

Snout area contaminaton comparison

To determine the origin of the contaminaton present within the zinc coating, a number of samples of different contamination were

collected from around the ZODIAC snout, with different phyiscal characteristics but consistent chemical composition. Previous

charcteisation of snout contamiantion was undertaken by Kim et al 5)

on different zinc and oxygen containing compounds using

ICP/XRD to determine stochiometric composition. SEM/EDX images of the different forms of contamination collected are

highlighted in Fig. 15, with XRD spectra highlighting presence of zinc oxide within all the samples collected, Fig. 16:

Figure 14 - XRD diffractogram of the arrowhead defect

White Powder Brown Powder Cream Powder

Black Powder

Figure 15 - SEM images of different forms of contamination collected around the ZODIAC snout

Figure 16 - XRD diffractogram comparison of different forms of snout contamination collected from ZODIAC

White powder characterisation - Pure ZnO platelets with nanorod growth through oxidation of molten zinc via injection of wet

HNx into snout to suppress the evaporation of molten zinc

Brown powder characterisation - mixture of metallic zinc/zinc oxide fingers entrained within regions of oxidic dross prior to

maintenance periods

Cream powder characterisation - Dispersions of hexagonally distorted zinc globules, traces of snout dust and finger like zinc

structures with similar morphology with respect to the brown powder contamination

Black powder characterisation - mixture of metallic zinc/zinc oxide formed through oxidation of of zinc vapour known as

"snout dust"

The entrainment of the white powder type contamination was observed with use of the ZODIAC snout camera as indicated in Fig. 3

and with consideration of the XRD patterns produced by the arrowhead defect and the contamination collected from the snout, a

consistent and distinct ZnO peak is observed at θ = 3 . Whilst the intensit of the ZnO peaks associated with other crystal planes are

masked by the intensity of the zinc peaks, this distinct ZnO [-1,0,0] crystal plane is readily observed and confirms the presence of

particles from the ZnO barier layer within the arrowhead defect. This assigment is confirmed with analysis of the SEM/EDX cross

sections and incrementally etched samples whereby discrete ZnO particles are observed within the zinc coating. The presence of ZnO

contamination as a result of ingot charging was disregarded due to excellent incoming ingot quality prior to use.

CONCLUSION

Whilst the use of wet HNx reduces the formation of zinc dust related defects during continuous galvanizing, the formation of surface

barrier oxides an induce other quality concerns. The entrainment of the surface barrier oxide into the zinc coating has resulted in the

formation of the arrowhead defect at ZODIAC. SEM/EDX characterisation coupled with incremetal etching using Nital revealed the

presence of the surface contamination within the zinc coating relative to the microfolding which is characteristic of the arrowhead

defect. If optimum settings for wet HNx injection/oxide removal are not continually in operation, increased occurences of coating

defects will be observed and result in poor quality full finish material. Therefore if alternatives to wet HNx injection can be found as a

method of suppressing zinc vapour formation, coating defects which oiginate from the formation of zinc vapour and surface barrier

oxide can be totally eliminated.

REFERENCES

1. M. Vlot, C. Price, M. Zuijderwijk, B. Van Veldhuizen, "Full Finish GI Manufacturing for the automotive industry", Proc. of 6th

International Conference on Zinc and Zinc Alloy Coated Steel Sheet, (2004)

2. J.L Arnold, S.L Boston, F. Caudill, "Control of defects caused by zinc vaporisation in the snout", Proc. of 6th International

Conference on Zinc and Zinc Alloy Coated Steel Sheet, (1989)

3. F. Chen, R Patil, "An In depth Analysis of Various Subtle Coating Defects of the 2000's", Proc. of 6th International Conference on

Zinc and Zinc Alloy Coated Steel Sheet, (2004)

4. N.-Y. Tang, F.N. Coady, "Coating Defects: Origins and Remedies", Proc. of Galvanizer's association, (2001)

5. H. Y. Kim, J.S Kim, S.J. Lee, "On Decreasing Zinc Dust in Snout equipped CGL", Proc. of Galvanizer's association

6. G. C. Becherer, "Utilizing pumps to continuously remove dross from inside the inlet snout", Proc. of Galvanizer's association,

(2001)

7. C. Phillips, N. Staples, M. Bright, "Cleanliness Enhancement of the Bath Surface within a CGL snout", Proc. of Galvanizer's

association, (2005)

8. M. Zelechower, J. Klis, E. Augustyn, J. Grzonka, D. Stroz, T. Rzychon H. Woznica, "The mircostructure of annealed galfan coating

on steel substrate", Archives of Metallurgy and Materials, Vol.57, 517-523 (2012)

9. G. Vourlias, N. Pistofidis, G. Stergioudis, E. Polychroniadis, " An overview of the solid byproducts of the galvanizing kettle",

Optoelectronics and Advanced Materials,Vol.1, 77-85 (2007)

10. S. Alam, M. Poloju, S. Sahu, M. Kumar, A. Singh, "SEM, EDX & XRD of Zinc Oxide Nanostructures synthesized by Zinc

oxidation", Microscopy and Analysis, 11-14 (2012)

11. E. M. A. Jamal, D. S. Kumar, M. R. Anatharaman, "On structural, optical and dielectric properties of zinc aluminate

nanoparticles" Bulletin of Materials Science, Vol. 34, 251-259 (2007)

12. G. Ragul, S. Sumathi, "Synthesis, Characterisation and Photocatalytic study of Zinc Aluminate Nanopowders against Rhodamine -

B and Crystal Violet Dyes" International Journal of Applied Engineering Research, Vol.8, 2175 - 2178 (2013)