Surface and Coatings Technology, 53 (1992) 243—255 243

Adhesion and corrosion properties of a new NiZn/ZnCo-coated steel tire

cord

J. Giridhar
Department of Chemistry and Geochemistry, Colorado School of Mines, Golden, CO 80401 (USA)

W. J. van Ooij*
Armco Research and Technology, Middletown, OH 45043 (USA)

(Received December 22, 1991; accepted in final form May 18, 1992)

Abstract
A new steel tire cord is described which is coated with an electrodeposited coating consisting of a base layer of 1.5 ~smZn—l%Co and
a top layer of 0.5 jim Ni—20%Zn. This coating is proposed to replace the conventional brass coating. This new cord has superior
corrosion resistance and aged adhesion retention properties which are described in this paper. The phase structure and composition
of the coating before, during and after wire drawing were studied by X-ray diffraction, scanning electron microscopy and energy-
dispersive X-ray analysis.

1. Introduction oxygen-rich regions near the external surface, the brass

coating itself corrodes through dezincification followed
The most commonly used method for obtaining high by dissolution of the brass. This dezincification in turn
levels of adhesion between steel tire cords and natural results in the deterioration of adhesion between the brass
rubber (NR) compound vulcanizates is to deposit a brass coating and the vulcanized NR compound. These corro-
coating onto steel wires of about 1.4 mm diameter prior sion processes effectively reduce the service life of a steel-
to final drawing. The drawing of the brass-coated steel belted radial tire.
wire produces a steel cord filament of a diameter of The initial (unaged) adhesion strength of brass to NR
0.25—0.38 mm with a highly deformed thin brass coating compounds is dependent on the compound formulation
of about 200 nm thickness. Several of these filaments are as well as on the bulk and surface compositions of the
then corded to form the final cord for tire reinforcement, brass coating [1]. Further, the highly deformed brass
In spite of several advantages such as the excellent coating on the final cord filaments is porous and does
drawability and high level of adhesion to NR corn- not offer barrier corrosion protection to the steel core.
pounds, two major deficiencies of the brass coating arise In order to minimize the sensitivity of the coated steel
from the electrochemical corrosion behavior of the cord to deleterious corrosion processes such as those
brass—steel couple under salt corrosion and humidity encountered with conventional brass coatings, a new
aging conditions [1, 2]. Cuts in the tire tread down to copper-free alloy coating system based on Zn—Ni and
the steel belt and cracks in the tire NR compound due Zn—Co binary alloys has recently been proposed [3, 5].
to flexural fatigue may be produced in service on the According to this design, an ideal coating would consist
road [3, 4]. Since salt is widely used on icy road surfaces of a zinc-rich first alloy layer on steel in order to provide
in winter, such defects can expose the steel cords to salt galvanic protection against corrosion. This first layer is
corrosion through the formation of oxygen concen- followed by a second layer of a nickel-rich alloy for
tration cells [2]. improved initial and aged adhesion to NR compounds.
The brass coating is cathodic to steel under conditions In order for the new dual-layer coating system to
of low pH existing in oxygen-depleted regions deeper in qualify as a viable candidate to replace the brass coating,
the cracks. Hence it accelerates the corrosion of the it must satisfy several criteria, among which are a high
underlying steel core, thus reducing the strength of the initial and aged adhesion to NR compounds and a
reinforcing steel cords. At high pH levels existing in capacity to offer galvanic protection against pitting
_____________ corrosion of steel cords observed in actual tire samples
*Author to whom correspondence should be addressed, after on-the-road testing [4].

0257—8972/92/85.00 © 1992 — Elsevier Sequoia. All rights reserved

244 .1 Giridhiar, U. .1. ran Ooij NiZn Zn(o—coated steel tire earS

Preliminary studies involving deposition of various of ZnCo layers within 1.0 ±0.3 wt.~ cobalt. The differ-
binary alloy phases of Zn--Ni and Zn--Co on flat steel ences between the various samples of each of the threc
strips using a laboratory-scale rotating cylindrical cath- basic coating types tested were in their layer thicknesses.
ode electroplating cell have been published recently [6, All coated wires were drawn at 12.3 m s drawing
7]. Deformation, adhesion and corrosion characteristics speed (speed of the drawn filament emerging from thc
were reported. These studies led to the selection of the last die). The diameter of the final filament emerging
dual-layer coating system consisting of 2.0 p.m Zn—I %Co from thc last die was 0.25 mm.
topcoated with 0.5 p.m Ni---20%Zn which had optimum A set of well defined samples of NiZn and ZnCo
deformation and adhesion properties. The electrodeposi- single-layer and NiZn/ZnCo dual-layer coatings on steel
don, drawing and cording characteristics of steel wires wires with different total coating thicknesses as well as
coated with the new coating system were studied in an different thickness ratios of the NiZn and ZnCo layers
industrial pilot line and recently described elsewhere [8]. was produced. This sample set is listed in Table I and
In summary, non-toxic sulfate-based high-speed plating was prepared for the systematic determination of the
baths were developed for depositing ZnCo and NiZn effects of varying the thicknesses of individual alloy
alloy coatings on running wires. The dual-layer system layers on the adhesion behavior of the coated wires.
proposed has excellent interlayer adhesion and adhesion In Table I, compositions and coating weights (grams
to the steel substrate. The wires coated with the dual- of the alloy coating per kilogram of the coated wire or
layer alloy system can be drawn from 1.4mm diameter cord) and thicknesses of the as-deposited dual-layer
to 0.25 mm diameter filaments utilizing the same stan- coatings (sample numbers S2 SlO) on wires prior to
dard lubricants and dies used for brass-plated steel wires. drawing are given in terms of the individual NiZn and
In this paper we present results of characterization ZnCo alloy layers, whereas those of the final cords
studies of the new coating before, during and after final produced from the respective wires are given as overall
drawing. These studies include phase structure, composi- compositions and total coating weights. Direct determi-
tion, electrochemical behavior, corrosion and dezincifi- nation of the exact coating thickness on a cord filament
cation behavior in salt solutions. Adhesion performance by examining the cross-section of the cord using scanning
before and after humidity aging to a standard commer- electron microscopy (SEM) was all hut impossible
cial NR compound is also described. In all tests, the because (i) the drawn coating is extremely thin (less than
performance of the new cord is compared with that of 100 nm, as determined by Auger electron spectroscopy
a standard brass-coated cord. )AES) depth profiling [7]), (ii) mechanical mixing and
Analysis of cord—rubber interfaces was also performed interdiffusion across the steel—ZnCo and ZnCo NiZn
and the exact mechanism of adhesion was studied in interfaces are highly probable during the drawing process
detail for both the new cord and the conventional brass- and lead to difficulties in locating the two interfaces.
coated cords. Results of these studies will be published and (iii) mechanical polishing of the resin-mounted
in the near future elsewhere, cords/filaments itself smears the coating layers, thus
obscuring the interfaces. Hence only overall elemental
compositions (determined by atomic absorption
2. Experimental procedure spectroscopy (AAS) and energy-dispersive X-ray analysis
(EDX)) and coating weights (determined by AAS only)
2.]. Materials of the coatings on cords could be determined with
2.1.1. Tire cords reasonable precision ( ±lO%). not the actual coating
The various NiZn and ZnCo alloy coatings deposited thicknesses.
onto running wires of 1.4 mm diameter can be classified
into three types: (i) zinc-rich it-phase ZnCo single-layer 2.1.2. NR compounds
coatings, (ii) nickel-rich a-phase NiZn single-layer coat- The test compounds used in this study were of the
ings, and (iii) NiZnZnCo dual-layer coatings with the conventional sulfenamide-accelerated medium sulfur
NiZn surface layer on the ZnCo underlayer. The electro- type. The variable was the amount of the cobalt-
deposition, drawing and cording of these coated wires containing adhesion promoter, which was cobalt neodec-
have been described [8]. Numerous samples of each anoate. The compounds are listed in Table 2. All the
coating type were produced and characterized in terms compounds were vulcanized at 1 51 ~C for 30 mm while
of their composition and phase structure. preparing specimens for adhesion (ASTM D 2229-73)
An initial set of drawing and cording experiments was tests.
conducted on several different samples of single- and
dual-layer coatings of the NiZn and ZnCo alloys on 2.2. Sample preparation
1.4 mm diameter wires. Compositions of the NiZn coat- Coating weight and chemical composition of the
ing layers were kept within 20 ±3 wt.% zinc and that CuZn, ZnCo and NiZn/ZnCo dual-layer coatings were
 J. Giridhar, W. J. van Ooij / NiZn/ZnCo-coated steel tire cord 245

TABLE I. NiZn-, ZnCo-, NiZn/ZnCo- and CuZn-coated steel wires and cords

Cord As-desposited coatings on Coated cords, 5 x 0.25 mm

identification 1.4 mm diameter steel wire (closed) cord construction
code
NiZn layer ZnCo layer Total Wt.%Ni Wt.%Zn Wt.%Co
coating
g/kg Wt.%Zn g/kg Wt.%Co weight,
(jim) (jim) g/kg

NZ 4.56 12 — 3.98 87 13
(1.5)
ZC ~- ~- 4.43 1.2 3.97 — 98.7 1.3
(1.7)
S2 1.61 15 3.98 1.0 4.84 25 74.4 0.6
(0.5) (1.5)
S3 1.61 15 3.98 1.0 4.55 24 75.4 0.6
(0.5) (1.5)
S4 2.38 16 3.23 1.1 4.44 35 64.5 0.5
(0.8) (1.2)
S5 1.15 14 4.06 0.9 4.47 19 80.6 0.4
(0.4) (1.6)
S6 0.59 17 1.98 0.9 2.05 21 78.5 0.5
(0.2) (0.8)
S7 0.75 20 1.91 0.8 1.98 25 74.3 0.7
(0.2) (0.7)
S8 2.51 18 5.76 0.8 6.16 24 75.4 0.6
(0.8) (2.2)
S9 0.49 17 4.66 0.7 — — —

(0.2) (1.8)
SlO 2.90 12 2.43 1.3 4.08 48 51.6 0.4
(0.9) (0.9)

T2 Cu—36%Zn, 4.5 g/kg Cu—35%, 4.08 g/kg

(1.35 jim)

The same proprietary lubricant was used in all the wire drawing experiments; sample S3 was drawn at 10.0 m s’~and all the others at
12.3 ms~.

TABLE 2. NR compounds used in adhesion tests not dissolve completely in concentrated nitric acid but
became passivated. Hence NiZn-coated wires (and cords)
NR compound Compound descriptiona were digested completely in a hot approximately 1:
mixture of concentrated nitric and perchloric acids and
NRC-i A standard compound formulation containing a
fixed amount of cobalt decanoate adhesion the solution was analyzed for nickel and zinc. The
promoter measured nickel and zinc contents were corrected for
NRC-2 Compound formulation containing only one-half the trace amounts of zinc and nickel present in the steel
the amount of cobalt decanoate adhesion promoter core.
as NRC-I
NRC-3 Compound formulation containing only one-fourth Coating weight and composition of the individual
the amount of cobalt decanoate adhesion promoter layers in as-deposited NiZn/ZnCo dual-layer coatings
as NRC-i were obtained by analyzing samples taken before and
NRC-4 Compound formulation containing no cobalt after the NiZn alloy plating step. These measurements
decanoate adhesion promoter
____________________________________________________ however are subject to the assumptions that (i) loss of
HAll four compound formulations contained carbon black, silica, zinc ZnCo layer in the acidic NiZn bath is negligible and
oxide, sulfur (medium level), N-oxydiethylenebenzothiazole sulfena- (ii) deposition behavior of the NiZn alloy is not altered
mide (OBTS) vulcanization accelerator, antioxidants, processing oil
and a resorcinol—hexamethoxymethyimelamine (HMMM) resin adhe- by the change in the substrate metal from steel to ZnCo
sion promoter. alloy. Layer thicknesses and compositions measured by
analyzing cross-sections of as-deposited coatings on
analyzed by dissolving the coatings alone (not the steel wires with SEM—EDX agreed with the results obtained
core) in concentrated nitric acid and analyzing the from AAS analysis to within ±5%. Such an agreement
solution for nickel, cobalt, copper and zinc with AAS. shows that in most cases the two assumptions mentioned
However, the nickel-rich NiZn single-layer coatings did above are reasonable.
246 .1. Giridhar, W. .1. ran Oaij NiZn, ZnCa—eoated steel tire cord

The composition of as-deposited single-alloy-layer coatings were examined by XRD using a Siemens
coatings of ZnCo, NiZn and CuZn on wires was analyzed diffractometer 11 with monochromatic Cu K~radia-
using SEM—EDX and AAS. Cross-sections of the above tion. The coated wire samples of 1.4 mm diameter were
single-layer coatings and of the NiZn/ZnCo dual-layer cut into 25 small pieces with each piece approximately
coatings were obtained by mounting the coated wires in 20 mm long and mounted parallel to each other on a
a room-temperature curing epoxy resin and polishing glass slide using paraffin wax backing mount. The wires
them perpendicular to the wire axis. Thicknesses of the were mounted such that the entrance slit was oriented
coating layers were measured directly from secondary perpendicular to the wire axis. Mounting the wires in
electron micrographs of the cross-sections. this manner reduced beam scattering and defocusing
Possible changes in the phase structures of NiZn effects due to the curved surfaces of the wires and thus
ZnCo coatings on 1.4mm steel wires which might occur increased the signal intensity. The changes occurring in
during drawing were examined using X-ray diffraction the phase structure of the coated wires during the
(XRD). Samples of NiZn/ZnCo dual-layer coated wire drawing process were studied using a Scintag PAD V
S2 (see Table I) at different stages of drawing. ~‘~“ automated diffractometer with Cu K~radiation. The
(i) before the first drawing die, (ii) between the fifth and samples were mounted just as described above.
sixth dies, (iii) between the 10th and 11th dies.
(iv) between the 15th and 16th dies, and (v) after the last 2.4. Te.s’tii~g
(19th) die and cording, were taken from the drawing 2.4.]. Corrosion
machine while the wire was being drawn. As-deposited wires and cords coated with NiZn, ZnCo,
CuZn and NiZn/ZnCo dual-layer coatings were tested
2.3. Instrumental analysis for their resistance to aqueous corrosion by immersing
2.3.1. Atomic absorption spectroscopy (AAS) the wires and the cords in an aerated 4 wt.% sodium
All analyses were performed by air- acetylene flame chloride solution at 22 C and monitoring the time taken
AAS using Perkin—Elmer model 460 and Philips Pye to form white rust (zinc and zinc alloy corrosion pro-
Unicam model PU9000 spectrophotometer, after appro- ducts) and the time to form red rust (iron oxide--
priate dilutions of the acid solution containing the hydroxide).
dissolved coating.
2.4.2. DezinciJication
2.3.2. Scanning electron microscopy and X-ray The sensitivity of the cords to dezincification was
analysis (SEM—EDX evaluated by immersing ten cord pieces each of 50 mm
The samples were analyzed by SEM—EDX for surface
morphology and composition using a JEOL JX 840 length in a 4 wt.% NaCI solution of pH 7.0 at 90 C. In
instrument with a Traeor Northern EDX system TN- order to prevent attack of the exposed steel at the cut
ends of the pieces, the ends were sealed by a room-
5500. Typical conditions of analysis were 15 kV beam
voltage, 0.1 nA beam current and a working distance of temperature setting epoxy paint. The solution was stirred
at 500 ±50 rev mm The NaCI solutions were ana-
~.

37 mm. Analysis of sample areas for composition was

performed always at a magnification of 1000 x which lyzed by AAS for the concentrations of copper. zinc.
corresponds to an analysis area of approximately nickel and iron after 1, 5, 30, 60 and 120 mm. Some
100 p.m x 120 p.m. Identical conditions were used in all hydroxides precipitated during this test. They were sepa-
analyses involving a given set of samples. rated by centrifugation and filtration, redissolved in
In most cases EDX analyses were carried out at three concentrated nitric acid and then analyzed for metal
different areas on the sample surface. In cases where the content.
percentage variation in composition over the three
different areas exceeded 5%, four additional areas were 2.4.3. Iron dissolution
analyzed and the results were averaged over the seven The sensitivity to iron dissolution effects was deter-
measurements. The choice of the analysis areas was mined in the Uniroyal test. Ten pieces of the cords each
random except for the occasional eases where a macro- of 76.2 ±0.2 mm length with sealed ends less than or
scopic flaw in the coating was detected. Such areas with equal to 1 mm were immersed in 0.10 ±0.01 N HNO3
macroscopic flaws were excluded from analysis. The solution at 22.5 ±0.5 C. The solution was stirred at
EDX spectra of the cord samples were quantified using 500 ±50 rev mm The immersion time was exactly
- ~.

a proprietary software for semi-quantitative analysis one minute. The solution was then analyzed for dissolved
available with the EDX system. iron by AAS.

2.3.3. X-Ray diffraction (XRD) 2.4.4. Electrochemical polarization

The phase structures of NiZn and ZnCo single-layer Potentiodynamic polarization scans of the different
coatings as well as that of the NiZn/ZnCo dual-layer types of coatings in as-deposited condition on wires and
 J. Giridhar, W. J. tan Ooij / NiZn/ZnCo-coated steel tire cord 247

in drawn condition on cords were carried out using an Figure 1 presents the XRD patterns obtained from
EG&G PARC 342-I corrosion measurement system. A each of the five wire samples taken at different stages in
saturated calomel electrode (SCE) served as the reference the drawing process (see Section 2.2). The diffraction
electrode. The counter-electrode was a graphite rod. In pattern of the as-deposited NiZn/ZnCo dual-layer coat-
all polarization experiments, an initial delay time of 90 s ing S2 (see Table I) prior to drawing (indicated as S2
was utilized to allow for sample equilibration. The Wire in Fig. 1) clearly shows peaks corresponding to the
surface area of the wire or cord sample exposed to the f.c.c. ct-phase NiZn surface layer, the h.c.p. 9-phase ZnCo
solution was 1 cm2. The scans were made by varying base layer and the b.c.c. at-phase of the low carbon steel
the potential applied to the working electrode (wire or wire substrate.
cord sample) from a value approximately 100 mV lower It can be expected that as the NiZn/ZnCo wire is
(cathodic) than the equilibrium rest potential of the being drawn, the individual coating layers of NiZn and
sample to a value 250 mY higher (anodic) than the rest ZnCo are mixed together mechanically to some extent
potential. Tafel plots of applied potential vs. logarithm owing to the severity of deformation and thinning of the
to the base 10 of the absolute value current density (E coating. The gradual disappearance of peaks correspond-
vs. log [j~)were obtained from these scans. Corrosion ing to the coating phases, except for the 9-phase (101)
potential, corrosion current density, and polarization peak of ZnCo, is indicative of progressive thinning of
resistance were calculated from the Tafel plots using the the coating layers during drawing.
proprietary software included in the corrosion measure- The presence of the (101) ZnCo peak in the XRD
ment system. pattern of the final cord without a significant change in
peak position indicates that small amounts of crystalline
2.4.5. Adhesion ZnCo grains remain in the coating on the final cord.
The adhesion of alloy-coated steel wires and final This result is important because it shows that galvanic
cords to specific proprietary NR compounds of Table 2 protection of the underlying steel by zinc in the coating
was evaluated using the standard ASTM D 2229-73 cord can still be expected. The XRD patterns also show that
pull-out test. The wire or cord embedment length was no hard intermetallic compounds, which might adversely
12.5 mm in all cases. The aged adhesion performances affect the drawing process, form in detectable quantities
of some of the cords was evaluated by subjecting the during drawing.
vulcanized test pieces to an aging treatment in a con- The diffraction peaks in Fig. I are not seen to shift in
trolled climate chamber maintained at 90% relative position significantly, but to broaden owing to a reduc-
humidity and 65 CC for eight days prior to the cord pull- tion in subgrain size and possibly an increase in non-
out test, uniform strain in the coating. Hence it can be concluded
that interdiffusion between the ZnCo layer and the NiZn
layer does not occur to a significant extent under the
3. Results and discussion temperature and pressure conditions existing during fine
drawing of the coated wire.
3.1. Characterization ofalloy coatings Peaks corresponding to the NiZn tx-phase in Fig. I
3.1.1. X-Ray diffraction
The X-ray diffraction patterns of as-deposited coatings .

on wires obtained in the manner described in Section 2.3 ~ 2 98 2 25 2.22

d - pacing
I 54 1 34 1 20 I 09”
were only used to identify the various alloy phases ~ ,~,,,,_j I I

present in the coatings. These diffraction patterns cannot 720.6 L____j~

be used either for extracting information on preferred 640.0 52 WIRE, CHRI43ES DURING 000IJII4S I
orientation of crystallites or for quantitative subgrain ~ C-0~ 80I’PLES I~UN1EDHORIZONTaL

size analysis because of the curved nature of the wire 480.1 60

surfaces. The diffraction patterns of as-deposited ZnCo
single-layer coating samples with less than or equal to
1.5% cobalt exhibited peaks corresponding to 9-phase
~°

w~,o
~
~
002

~
(200) ‘~C-0~ ~

46

(h.c.p.) ZnCo alloy and to the steel substrate only. ~

Only peaks corresponding to the ct-phase (f.c.c.) NiZn ~ ~ 16 Ii LOLL .,1~. 20

alloy and to the steel substrate were present in the XRD _...._..J- ~
patterns of as-deposited NiZn single-layer coatings with ~ CORD

lO%—25% zinc. Significantly larger peak widths were —~-~ , ‘~~‘ do ‘IO ‘ 0 ‘

observed for the NiZn peaks which indicate either a Diffraction Angte, 20 tdeg.l

finer average crystallite (subgrain) size or substantial Fig. I. XRD patterns of the dual-layer NiZn/ZnCo coating (sample
non-uniform strains in the coating, or both. S2 of Table I) after different stages of wire drawing.
248 .1. (iiridhar. 14’. .J. tan Ooij NOLn/ZnCo—L’oated steel tire oarS

have disappeared almost completely at the end of draw- as revealed by our recent studies utilizing sputter depth
ing and cording operations. It is speculated that the at- profiling techniques [9]. It is thus very difficult to obtain
phase NiZn layer, which was already microcrystalline in direct analytical evidence for the layered structure of
the as-deposited condition [8], has become extremely drawn coating on the final steel cord filaments.
thin, highly strained (with non-uniform strains) and
possibly amorphous during the drawing process for the 3.2. Corrosion
following reasons. Table 3 summarizes the results of the salt corrosion
Substantial amounts of nickel (25 wt.%, as per AAS) tests. The data on time to red rust formation demonstrate
present in the coating after drawing confirm that the the excellent corrosion protection offered by the NiZn’
NiZn top layer is not worn away by the drawing ZnCo coating to the steel core in comparison with
operation. Segregation of nickel in surface layers of the conventional brass coatings. Another important aspect
drawn coating, evident from the EDX analysis results of the data in Table 3 concerns the behavior of the NiZn
presented in Section 3.1.2, indicates that the NiZn top single-layer coating. In the as-deposited condition, NiZn
layer remains at or near the coating surface during offers good barrier protection, as can be concluded from
drawing. Since the positions of the i~-ZnCopeaks do the long time to form red rust, and it corrodes under
not shift significantly and the formation of new phases formation of white rust.
is not detected, diffusion of nickel from the NiZn surface In contrast, the deformed NiZn single-layer coating
layer into the ZnCo base layer does not seem to occur quickly forms red rust, indicating incomplete coverage
during drawing. of the steel surface by the coating. White rust is no
Further, after drawing, the NiZn single layer coating longer formed in this case. This behavior is similar to
NZ (see Table I) did not show any detectable diffraction the corrosion behavior of the brass-coated cord sample.
peaks corresponding to the at-phase coating; only peaks In addition, the surfaces of all nickel-containing samples
corresponding to the steel substrate were detected. This (both wires and cords) darkened after I 0---20 mm of
observation suggests that in the case of NiZn/ZnCo immersion in the salt solution, probably owing to the
dual-layer coating S2, thinning of the coating layers formation of nickel oxides and hydroxides at the coating
during drawing may not be the only cause of the surface.
disappearance of NiZn peaks. Hence it seems highly The ZnCo coating is seen to form white rust quickly.
probable that the NiZn layer has become very thin and before and after drawing. However, this corrosion pro-
practically amorphous and that it is still present in the duct formation is delayed by a factor of about 3 when
surface layers of the drawn coating (see EDX results compared with a coating of pure zinc. The protection
below), against red rust formation is marginally better than for
pure zinc.
3.1.2. Coating analysis by energy-dispersive X-ray The dual-layer NiZn/ZnCo system is superior to all
analysis other coatings in terms of preventing both red and white
Crude elemental concentration depth information can rust formation, both before and after drawing. It appears
be obtained by collecting EDX spectra of the coated that the effects of the two layers are synergistic. Excellent
cords at different electron beam accelerating voltages, galvanic protection is offered by the ZnCo coating,
e.g. 13, 20 and 30kV. The variation observed in the which in itself is better than pure zinc in delaying rust
ratio of weight percentages of zinc to nickel in the formation. The NiZn top coating acts asa barrier
coating of cord S2 (Table I) measured at different volt- coating, even after drawing, if it is supported by the
ages was as follows: ductile ZnCo substrate layer. In this case it does not
Voltage (kV) Zinc’ to nickel ratio crack as badly as when deposited onto steel directly but
13 9.1 probably becomes amorphous, as indicated by the XRD
20 12.0 analysis (see Section 3.1). The presence of this surface
30 1 5.7 layer, even after drawing, was also demonstrated by the
The increase in the zinc to nickel ratio with increasing EDX analysis (see Section 3.1). It is speculated that the
accelerating voltage presents evidence for nickel enrich- amorphous surface layer of NiZn has an improved
ment near the surface and enrichment of zinc near the barrier protection capability, which reduces the rate of
steel—coating interface. This, in turn, is indirect evidence corrosion of the crystalline ZnCo layer. The composite
that the dual-layer structure of the coating is still present coating thus exhibits excellent resistance to both red
to a certain extent (i.e. in so far as the existence of a and white rust formation.
nickel-rich region at the cord filament surface and a
zinc-rich region at the coating—steel interface) after the 3.3. Electrochemical polarization
drawing operation. However, after drawing, the two The purpose of the electrochemical polarization
coating layers no longer exhibit well defined boundaries studies was to obtain more quantitative data on the
 J. Giridhar, W. J. van Ooij / NiZn/ZnCo-coated steel tire cord 249

TABLE 3. Corrosion resistances of alloy coatings on steel wires and cords

Coating type Coating layer Coating weight Time to form Time to form
composition of each layer white rusta red rust
(wt.%Zn) (g/kg) (mm) (mm)

As-deposited coatings on 1.4mm diameter steel wires

Pure zinc 100.0 1.90 20 1815

ZnCo 98.8 2.11 45 4320
NiZn 17.2 2.08 1380 1815
NiZn on ZnCo 19.0 0.88 60 3630
(dual layer) on 98.8 on 2.10
CuZn 35.2 4.18 — 20
Uncoated steel wire — — — 90

Coatings on 2 + 2 x 0.25 mm steel cords (after drawing and cording)

Pure zinc 100 1.44 10 240
ZnCo 98.7 1.91 20 1152
NiZn 20 1.83 — 5
NiZnjZnCo 69.6” 2.8” 1140 1720
CuZn 34.7 3.93 — 30

~Zinc corrosion products (zinc oxide, hydroxide and hydroxychlorides).

hOverall zinc content and coating weight of the coating on the cord.

actual corrosion rate and corrosion resistance of the —250

various coating systems. C)
Table 4 presents the electrochemical corrosion param- ‘~ -

eters measured from the as-deposited coatings. A typical ~

potentiodynamic polarization curve obtained from an
Ni—22%Zn alloy-coated wire sample (see Table 4) is —350
shown in Fig. 2. Table 5 summarizes the results of similar ~
experiments conducted with deformed coatings on cords. ~
The potentiodynamic polarization curve obtained from ~
0
400 -

an NiZn/ZnCo-coated cord sample (Table 5) is presented n-

in Fig. 3. The anodic part of the curve in Fig. 3 shows ~ —~o

a slight passivation behavior similar to that of NiZn/ J
ZnCo alloy-coated steel strips described in detail in an
earlier publication [7].
1o3
—500
The rest potential E° values listed in Tables 4 and 5 10~ 100 10’ 102
—2
were obtained by monitoring separately E°as a function CuvTont Denslty.j (pA cm )
of time. Data presented in both tables were the averages Fig. 2. Potentiodynamic polarization curve for an Ni—22%Zn alloy
of at least three separate scans. The scans were repeated coating on 1.4mm diameter steel wire (see Table 4).

TABLE 4. Electrochemical corrosion parameters of as-deposited coatings on 1.4mm diameter steel wires

Coating Coating Rest Corrosion Corrosion Polarization

type layer potential potential current density resistance
thickness E° E(corr) I)corr) R
2) (k~
5 cm2)
(jim) (mV vs. SCE) (mY vs. SCE) (jiA cm
Pure zinc 1.9 —987 —962 103 0.03
Zn--I.5%Co 1.8 —973 —940 84 0.05
Ni—22%Zn 1.0 —363 —349 27 0.64
NiZn 0.5
on on —918 —882 61 0.36
ZnCo 1.5
Cu—36%Zn 1.2 —419 —398 65 0.59

Potentiodynamic scans were made at 0.1 mY s~ scan rate in an aerated stirred 0.35 wt.%NaCI solution of p1-I 4.6 at 25 ±1 -‘C.
250 .1. Giridhar , tI’. .1. ran Ooij NiZitLn(o—coated steel tire o-ord

in order to verify the repeatability of polarization experi- Important conclusions can be drawn from the results
ments. The wire samples yielded more consistent data presented in Tables 4 and 5.
and repeatable results than the cord samples, indicating (I) There is a significant increase in corrosion current
that the drawn cords were more variable in terms of density and a decrease in polarization resistance of NiZn
surface defects, coatings after drawing; this observation is indicative of
The rest and corrosion potentials in Tables 4 and 5 a reduced barrier corrosion protection offered to the
are indicative of the phase or component of the coating underlying steel by the deformed coating, as was also
that is corroding at the highest rate. The corrosion concluded from the salt exposure experiments.
current density is a measure of the corrosion rate and (2) A considerably lower corrosion current density
the polarization resistance is a measure of the resistance and higher polarization resistance are observed for the
to corrosion offered by the corroding coating compo- drawn NiZn/ZnCo coating as compared with the drawn
nent. The rest potential of bare steel is close to —527 mV ZnCo coating. This observation underscores the impor-
t’s. SCE, hence both CuZn and NiZn single-layer coat- tant effect of the NiZn top layer in reducing the apparent
ings are cathodic with respect to steel. The results corrosion rate of the dual-layer coated cord.
presented in Tables 4 and 5 are very similar to those Finally, it should be noted that it appears from the
presented and discussed in greater detail in an earlier data in Table 5 that CuZn coatings exhibit almost the
publication for the same types of coatings on flat steel same corrosion current density as the NiZn/ZnCo coat-
strips [7]. ing, but with a higher polarization resistance than the
latter. However, by comparing the corrosion potential
of the CuZn and of the NiZn/ZnCo coatings with that
525 ~ of hare steel, it can be recognized that the corroding
phase is steel in the former and the ZnCo underlayer in
~ 575 . the latter system.

—625 . 3.4. Dezincific’ation and iron dissolution

The sensitivity to dezincification effects was deter-
mined for the alloy coatings NiZn, NiZn/ZnCo and
~ —675 - CuZn (NZ, S2 and T2 of Table I). These experiments
were conducted because it is widely believed that dezinci-
—725 - , fication sensitivity of CuZn coatings is directly related
to their humidity aged adhesion behavior [I. 2]. The
results are presented in Tables 6 and 7. They are
l~~ll~lI 2 ‘‘‘~ “‘~O~ expressed in Table 6 as milligrams per square meter of
the
time,element dissolved
the ratefrom the cord asof aeach
function of
2) indicating of dissolution element.
Current Density. j ~& cm
l’ig. 3. Poientiodynamic polarization curve ‘or a drawn NiZn,ZnCo In Table 7 the dissolved amounts of copper. nickel and
dual-layer coating on a S x (1.25 mm closed steel cord (see Table 5>. zinc are normalized with respect to the total amount of

‘FABLE 5. Electrochemical corrosion parameters of drawn coatings on steel cords

Coat ittg Total Rest Corrosioto (‘orrosion Polarizatioit

type coating potential potential current densit~ resistance
weight L”‘ L’(corr( j(corr) R,,
(g/kg( mV i-s. SCEI mV is. SCE)
0A cm ~) kQ cm-)

Zn -2°.,Co’ 3.6 —940 938

--- ISO 0.1)66
Ni—22°4Zn 4.0 —406 --417 45 0.37
NiZto
on 5,>) —718 —7)0 31 1.13
ZnCO
Cu—36%Zn 4.0 ----393 --413 33 ((.65
Bare steel”‘ --- —527 —505 38 ()~57

“Cord of 2 + 2 x 0.25 mm construction: all others were 5 x 0.25 mm (closed) cords.

“Brassed steel cord after complete removal of the brass coating in an approximately 1:1 mixture ol’ H,O, and concentrated NH4OI-1.
Potentiodynamic scans were made at 0.17 mV s~ scan rate in an aerated stirred solution of p1-1 5.65 at 25 ±I C’ containing 0.35 wi.’, Na~SO4
and 0.05 wt.°+ NaCI.
 J. Giridhar, W. J. van Ooij / NiZn/ZnCo-coated steel tire cord 251

the same element on the untreated cord. This normaliza- The dual-layer cord S2 did not form any white or red
tion enables qualitative comparisons of the relative rust at the surface. Some white rust precipitated in the
elemental dissolution rates among the three types of vessel. The amount of nickel dissolved from the coating
alloy-coated cords considered, is rather high here, too. This effect may be related to
When comparing NiZn with CuZn, the observed the undercutting effect of the dissolving ZnCo base layer,
behavior of the two alloys seems to be different, although as has been demonstrated and published before [7]. The
the normalized amounts of dissolved zinc are approxi- amount of dissolved zinc is considerably higher than in
mately the same (see Table 7). In the case of CuZn, zinc the case of CuZn or NiZn. The underlying mechanism
dissolves only partly and forms a visible surface film of for this difference is that in the case of the dual-layer
white zinc corrosion products. Zinc seems to dissolve system, zinc is dissolving in the process of cathodic
from the NiZn alloy, leaving a surface layer enriched in protection of the steel core. In the case of brass, which
nickel. It is remarkable that the initial rate of nickel is cathodic to steel, the zinc dissolution is a pure
dissolution is higher than the rate of copper dissolution, dezincification effect.
but the former levels off after about 120 mm, whereas The cathodic protection of the steel core by the dual
the copper dissolution increases sharply after about layer is clearly demonstrated by the very low amount
60 mm. It appears that the nickel-enriched surface of dissolved iron here. In the case of NiZn the low iron
becomes passive, forming a dark layer which probably dissolution rate is due to the barrier effect, whereas in
contains nickel oxide. In the case of CuZn, the surface the case of the NiZn/ZnCo coating the iron dissolution
remains active and after a certain degree of zinc deple- is reduced by a combination of cathodic and barrier
tion, copper begins to dissolve rapidly. protection.
It was further observed that after a few minutes of The results of the dezinciIIcation studies discussed
immersion, the CuZn-coated cord began to form large here are consistent with the electrochemical and corro-
quantities of red rust all over the surface. Also, a red sion data presented earlier.
rust precipitate formed in the solution after about 30 mm. The results of the iron dissolution test are presented
In contrast, the NiZn-coated cord formed only local in Table 8 for NiZn and CuZn coatings of equal thick-
patches of spotty red rust, presumably related to local ness. This test was carried out because it is a common
cracks in the coating, which developed during drawing, test for evaluating the porosity of brass coatings on steel
as discussed in the section on corrosion of the cords, tire cords. The behavior of the cords was compared with

TABLE 6. Dezincification of CuZn, NiZn and NiZn/ZnCo coatings on steel cords—elemental weight loss vs. treatment time

2 area of the cord surface (mg m2)

Dezincification Actual amount of element (mg) dissolved from the cord per I m
time (mm)
CuZn-coated cord (T2~) NiZn-coated cord (NZ)~ NiZn/ZnCo-coated cord (S2f’

Cu Zn Fe Ni Zn Fe Ni Zn Fe

16.76 22.12 97.21 77.10 12.06 134.09 67.04 166.27 0.00

5 20.11 29.50 442.50 87.16 13.07 184.37 83.80 327.85 83.80
30 26.81 57.98 1119.58 90.51 38.20 486.08 93.86 427.75 207.84
60 73.74 138.95 1237.00 161.20 43.24 703.97 117.32 533.68 278.24
120 231.30 149.17 1793.49 207.84 45.58 754.27 140.78 800.86 465.97

‘See Table I.

TABLE 7. Dezincification of CuZn, NiZn and NiZn/ZnCo coatings on steel cords—percentage of elemental weight loss vs. treatment time

Dezincification Wt.% of element dissolved from the coating per I m2 area of the coating surface (wt.% m2)
time (mm)
CuZn-coated cord (T2)~ NiZn-coated cord (NZ)~ NiZn/ZnCo-coated cord (S2)a

Cu Zn Ni Zn Ni Zn

1 1.3 3.1 4.5 4.7 11.3 9.4

5 1.5 4.2 5.1 5.1 14.3 18.5
30 2.1 8.3 5.3 15.0 15.8 24.2
60 5.7 19.8 9.5 16.9 19.7 30.2
120 17.8 21.2 12.2 17.8 23.7 45.3
aSee Table I.
252 .1. Girid/iar. U-’. .1. ran 0o5 NiZn.Zn(’o—o’oated steel tire o-,ii’tl

TABLE 8. Iron dissolution in 0.1 N nitric acid: C’uZn-coated cord rs. NiZn-coated cord

Nature of coating Iron dissolved per I ~ cord surface (g no

CuZn-coated cord (T2)’ NiZn—coatc’d cord (NZ)”

As—deposited coating on .78

1.4 mm diameter wire
Drawn coating on thc 29.24 I5.iil
final 2-i-2 x 0.25 mm cord

“See Table I.

that of the coated wires before drawing. The dual-layer deposited condition. Sample SlO has the lowest ratio of
NiZn/ZnCo was not included in this test, because this ZnCo layer thickness to NiZn layer thickness (1) and
coating dissolved completely in the acid that was used sample S9 has the highest ratio (9). The same NR corn-
here (0.1 N HNO3). The results for NiZn and CuZn pound NRC-I (Table 2) was used here. The cord samples
show that the barrier effect of both coatings decreased were tested in unaged condition only.
significantly upon drawing. However, both before and The following important observations can be made
after drawing, the NiZn coating performs better than when comparing the adhesion data presented in Tables
CuZn. This can probably be attributed to a lower 9 and 10. Considering sample NZ (NiZn single-layer
porosity of the former coating, as was also concluded coating) before drawing (Table 9) and after drawing and
from the electrochemical data presented in Section 3.3 cording (Table 10), it is seen that the adhesion of this
earlier, sample drops from a high value to an almost complete
loss of adhesion upon drawing. In contrast to the NiZn
3.5. Adhesion sample, the ZnCo coating ZC exhibits a significant gain
Table 9 gives the typical wire pull-out forces measured in adhesion upon drawing. Comparing the pull-out force
for as-deposited wires coated with the four types of alloy and rubber coverage data of cord samples S5, S6, and S8
coatings of Table I tested with NR compound NRC-I (Table 10) reveals that variation of total coating weight
of Table 2. The pull-out adhesion data presented show has very little effect on cord performance in the pull-out
that the as-deposited NiZn single-layer coating exhibits tests. The insensitivity of the cord pull-out force to coat-
the highest level of adhesion. The ZnCo. CuZn. and ing thickness exhibited by NiZn/ZnCo dual-layer coat-
NiZn/ZnCo coatings exhibit approximately equal levels ings indicates that the bulk of the coating is not involved
of adhesion to the NR compound NRC-I (Table 2). It can in processes that lead to the build-up of high adhesion
also be seen from Table 9 that the performance of all alloys levels. This characteristic of NiZn/ZnCo coatings is in
in the pull-out test is significantly better than that ofa pure direct contrast to that of brass coatings. Studies reported
zinc coating or of an uncoated steel surface. in the literature have clearly demonstrated that the brass
Table 10 presents the cord pull-out test data of the set coating thickness has a significant influence on NR corn-
of 5 x 0.25 open cords (NZ, ZC and 52-510) listed in pound --cord adhesion [I]. The nature of this influence
Table 1. Samples S2 -S5, S9, and SlO of Table I have the depends strongly on the compound parameters.
same total coating thickness of 2 p.m in the as- Another important observation which can be made
from Table 10 is the influence of the nickel content of the
TABLE 9. Adhesion to N R of as-deposited coatings on .4 mm NiZn/ZnCo coating. Comparison of the pull-out forces
diameter steel wires of cords S2, S4, S5, and S It) shows that an increase in
overall nickel content of the coating beyond approxi-
Coating Thickness of Wire pull-out Rubber rnately 25 wt.% results in a gradual decrease in pull-out
type each alloy laser force coverauc
(jim) ‘ (N) (°) force and in rubber coverage of the pulled-out cord sur-
face, as depicted in Fig. 4. The latter indicates that the
Zto l.5°’/,C’o 1.7 830 5O--75 drop in pull-out force is most probably due to a decrease
Ni-- l8°.’1-Zn 0.7 035 0 I)))) in actual NR ---coating adhesion strength rather than due
Ni—20°/oZn on , - - to a decrease in the cohesive strength of the vulcanized
Zn ----1 .2°~,,Co 0.8 on 1.4 84~ 50 75
dual layer( NR compound. At concentrations lower than 25 wt.9-’,.
(‘si -3591-Zn 1.2 897 25 50 the nickel content of the coating seems to have very little
Pure zinc )L7 390 0- 2S effect on both cord pull-out force and rubber coverage.
Ilncoated steel wire - The visually observed rubber coverage was always nearly
1.2mm diameter) 100% if the nickel content of the NiZn’ZnCo coatings
The wires were bonded to NR compound NRC-I (Table 2). was between 0 and 25 wt.1-~,.
 2. Giridhar, W. J. van Ooij / NiZn/ZnCo-coated steel tire cord 253

TABLE 10. Adhesion to NR of drawn coatings on steel cord samples listed in Table I

Cord ZnCo-to-NiZn Total Wt.%Ni Average Rubber

code layer thickness coating in the cord coverage
ratio in the weight on coating on pull-out (%)
as-deposited the cord the cords forcea
coating on wires (g/kg) (N)

NZ — 3.98 87 132 0
ZC — 3.97 — 608 75—100
S2 3.0 4.84 25 618 100
S3 3.0 4.55 24 605 75—100
S4 1.5 4.44 35 570 50—75
S5 4.0 4.47 19 600 75—100
S6 4.0 2.05 21 610 100
S7 3.5 1.98 25 605 75—100
S8 2.8 6.16 24 601 75—100
SlO 1.0 4.08 48 401 25—50

T2 Cu—65%Zn 4.08 — 643 100

coating
aAverage of eight tests.
Cords were bonded to NR compound NRC-I (Table 2).

650 I this type of compound cannot be achieved, The results

in Table 11 suggest that the amount of cobalt adhesion
- promoter has no significant effect on the pull-out force
and rubber coverage. It appears that the critical amount
2 ‘~“q,, of promoter required for obtaining high adhesion levels
550 ‘ can be considerably lower than the amount commonly
used in commercial NR compounds for brass-coated
500 CO~~
1e12 presents the results of experiments in which
the humidity aged adhesion of the NiZn/ZnCo-coated
-~ 45o - ___________________ - cord, drawn with two different lubricants, was compared
corLS2-S10(cf.Tab!s 1) with that of the CuZn-coated cord T2. The compound
- NRC-i (of.~Tcbls2) - was the standard commercial compound NRC-I of
Table 2. Table 12 clearly shows that the lubricant has a
small effect on both initial and aged adhesion of the
15 20 25 30 50 dual-layer-coated cord, Although the initial adhesion of
the two types of coatings is approximately the same, the
Wt.—% NI In Coating . , ,

aged adhesion of the NiZn/ZnCo-coated cord is signifi-

Fig. 4. Variation of cord pull-out force with nickel content of NiZn/ cantly higher. Since also the rubber coverage of the
ZnCo dual-layer coatings. pulled cords is better than for brass-coated cords, it can
be concluded that there is less degradation of the inter-
In another experiment the NiZn/ZnCo-coated cord facial film during humidity aging. In the case of CuZn
S2 and the CuZn-coated cord T2 (see Table I) were this degradation is believed to be due to the hydration
tested for adhesion to three different compounds of the of ZnO in the NR—coating interfacial layer, leading to
same basic formulation but with different amounts of a weak Zn(OH)2 layer. Such processes do not occur in
cobalt adhesion promoter (see Table 2). The results are the case of NiZn/ZnCo coatings, as was evidenced by
presented in Table 11. It is observed that the pull-out the corrosion and dezincification data presented earlier
force and rubber coverage for the brass-coated cord in Sections 3.2 and 3.4 respectively.
decrease somewhat with a decrease in the cobalt level.
For the best adhesion performance, the NiZn/ZnCo- 4. Conclusions
coated cords require the use of a small amount of cobalt
adhesion promoter in the NR compound. Without cobalt The results of the experiments and analyses described
adhesion promoter, a significant adhesion strength to in the foregoing can be summarized as follows,
254 .1. Giridhtar. U’. .1. ran Ooij . NiZn/Zn Co—coated steel tire carol

TABLE II. Effect of cobalt-based adhesion promoter on NR-cord adhesion

NR NiZn,’ZnCo-cootted cord (S2( CuZn-coated cord )T2)

compound
Average cord Rubber Average cord Rubber
pull-out force (N) coverage pull-out force (N) cos-crage
and standard deviation I9-~( and standard deviation

NRC-I X( 16) = 653 75 100 .V)8) = 673 75 100

o=70
NRC-2 X( 16) = (o78 75 I))>) X)8) = 671 75 10))
no = 46
NRC-3 .V(16(=692 75-100 X(8l=653 50 -10))
ro=50 n=69
NRC’-4 .Y( 16) = 118 1) X(8) 596 50-- 100
no=68

X(n) is the average value of n cord pull-out test measurements.

Details of cords and compounds are given in Tables I and 2 respectively.

TABLE 12. Effect of humidity aging and of the wire drawing lubricant on the NR compound cord adhesion”

Cord type Unaged adhesion Adhesion after 4 days of aging Adhesion after 8 days of aging

Cord pull-out Rubber Cord pull-out Rubber (‘ord pull-out Rubber

force (N) coverage (9/) force (N( coverage (‘1’,) force (N) coverage I” of

NiZn,’ZnCo S17 75—100 457 75-- 100 383 5)) 100

(dual layer),
lubricant A
NiZn,’ZnCo 465 75—100 430 75-10)) 35)) +1 75
(dual layer),
lubricant B
(‘Li- 3591-Zn 495 75- 10)) 330 25 5)) 340 25 50
(standard brass),
lubricant A

‘Adhesion to NR compound NRC-I (Table 2( humidity aging conditions. 9O5’~relative humidity, 65 C temperature.
A and B were two different lubricants of proprietary formulations used for drawing brassed steel wires.

(I) A new electrodeposited dual-layer alloy coating conventional brass coatings shows that the former exhib-
system designed for use on steel tire cords and consisting its superior corrosion resistance and electrochemical cor-
of an a-phase Ni—20%Zn alloy surface layer deposited rosion characteristics. The presence of the NiZn top
onto an 9-phase Zn—I %Co alloy base layer has been layer offers effective barrier protection to the ZnCo base
developed. The composition, structure, adhesion and layer. The latter layer offers galvanic protection to the
corrosion performances are reported. steel core. Thus the complete system significantly
(2) Both the steel--coating interface and the NiZn - increases the resistance of the coated cord to both red
ZnCo interface are very diffuse after the wire drawing and white rust formation.
operation. However, experimental evidence obtained (5) Adhesion of the dual-layer-coated cords to corn-
from XRD and EDX analyses and from electrochemical mercial NR compounds used for bonding low-copper
potential measurements suggest a dual-layer coating brass-coated cords is either comparable or better than
structure consisting of a thin, possibly amorphous, NiZn that of conventional CuZn-coated cords. After humidity
surface layer on a highly strained crystalline ZnCo layer. aging, the dual-layer-coated cord’s adhesion retention is
(3) Analysis of the cords shows that the NiZn/ZnCo significantly better than that of the brass-coated cords.
coatings are less porous than the conventional CuZn (6) The NiZn/ZnCo-coated cords require small
coatings and thus reduce the risk of corrosion of the amounts of cobalt-containing adhesion promoter in the
reinforcing steel core. NR compound. Without cobalt, the observed adhesion
(4) Comparison of the corrosion and electrochemical is very low and rubber coverage is zero. However, the
characteristics of the dual-layer coating with those of adhesion results suggest that the minimum amount of
 J. Giridhar, W. J. van 009 / NiZn/ZnCo-coated steel tire cord 255

cobalt required for achieving high adhesion levels can References

be lowered considerably in comparison with the amounts
used in compounds for brass-coated cords. I W. J. van Ooij, Rubber Chem. Technol., 57 (1984) 421.
2 Y. Ishikawa and S. Kawakami, Rubber Chem. Technol., 59 (1986)
1.
Acknowledgments 3 W. J. van Ooij, J. Giridhar and J. H. Ahn, Kautsch. Gummi, Kunstst.,
44 (1991) 348.
The authors gratefully acknowledge that most of the 4 J. H. Ahn, MS Thesis, Department of Chemistry, Colorado School
research on the new tire cord coating described in this of Mines, Golden, CO, 1989.
paper was supported by Societá Pneumatici Pirelli SpA 5 W. J. van Ooij, to Pirelli Coordinamento Pneumatici SpA., Eur.
and was conducted at the pilot plant facilities of its steel Pat. AppI. EP 283,738 September 28, 1988.
6 J. Giridhar and W. J. van Ooij, Surf Coat. Technol., 52 (1992) 17.
cord factory in Figline Valdarno, Italy. The NR com- ~ J. Giridhar and W J van Ooij, Surf Coat Technol., 53 (1992) 35
pounds were kindly provided by Pirelli Coordinamento 8 J. Giridhar and W. J. van Ooij, Wire Expo ‘92, Oct. 18—22, 1992,
Pneumatici SpA in Milan, Italy. Louisville, KY, USA.