Dynamic aspects of brass adhesion.

Dynamic aspects of brass adhesion

Enhanced adhesion to various substrates has long been a distinguishing characteristic of rubber compounds reinforced with precipitated silica. Several years ago this laboratory developed a dynamic test to determine the fatigue properties of brass coated steel tire cord and the influence of silica on these properties (refs. 1-3). Work has continued on this disc fatigue test on both silica and non-silica formulas. This article describes some of the most recent results from brass bond fatigue evaluations of two types of silica skim compounds.

One possible source of the strong bonding which results from replacing carbon black by silica is the hydrophylic nature of silica. Surface silanols provide increased compatibility with the oxygenated groups of textile fibers or the oxidized surface of metals. Another pertinent silica characteristic, particularly in respect to humid aging effects, is the affinity for free water (below 100 [degrees] C). It may also reasonably be inferred that silica's high surface activity may modify copper sulfide formation in the direction of increased stability. The absence of a yet to be undertaken analytical investigation precludes further speculation of possible mechanisms. The work described here, based on the usual compounding technique of varying concentration and ingredient, leads to formula optimization in respect to brass adhesion.

The disc fatigue adhesion test

The apparatus used in the disc fatigue procedure is the Cord Compression Fatigue tester introduced by M. W. Wilson to evaluate the fatigue properties of organic cords. To test for brass bond fatigue, textile cord has been replaced by brass coated steel wire and the rubber compound becomes the major variable.

There are two unique features of the disc fatigue procedure:

* Discrimination - bonding systems which appear to be equal in static tests are revealed to be distinctly unequal in terms of fatigue life. Disc fatigue life correlation with tire performance has been demonstrated.

* Interfacial separation - bond fatigue separations occur at the rubber-brass interface and display a bare metal surface. Unlike the rubber cohesion values typical of pull-out procedures, disc fatigue measures adhesion of brass to rubber or to the steel base.

Schematic drawings of the test specimen and apparatus appear in figure 1. Generally, wire adhesion data are expressed in terms of the number of bond separation failures which occur in six replicate specimens (12 failure sites) together with the average of the maximum and minimum pull-out values after flexing. Disc fatigue conditions of 16% compression and extension with flexing time of six hours at 70 [degrees] C were used for the work described below. In many cases the most informative results are those obtained on samples which have been aged at high humidity and temperature before flexing. Adequate strain during flexing requires the use of a relatively rigid wire construction, 6 x .35mm + 3 x .20mm.

Since flexing takes place at constant strain, the fatigue life can be influenced by low strain modulus. Thus, comparisons of fillers and other compound variables must be carried out at relatively comparable 20% modulus and hardness values. For example, an increase in 20% modulus of more than 0.3 MPa, could produce one or two additional separation failures in a set of six specimens (table 1).

Silica-resin skim compounds

The change in dynamic properties which occurs when black is partially replaced by silica has been described (refs. 4 and 5). In tire tread formulas, the reduction in low strain modulus, hardness and tan delta due to the presence of silica appears to be the result of a decrease in agglomerate structure. However, the improvement in dynamic adhesion properties effected by silica shows evidence of chemical as well as physical influences. For example, the wetting characteristics of silica skims, in terms of a reduction in contact angle measurements, are substantially improved in comparison to those of all-black controls.

The formula recommendation in table 2 is both the result of considerable laboratory development and a prototype of many commercial wire skim compounds. Its features include 15 phr silica, a resorcinol (R-F) and hexamethoxy methylmelamine (HMMM) resin system and a sulfur content of 4 phr. The use of 80% active insoluble sulfur is possible, but has shown evidence of reduced scorch safety and slightly reduced fatigue life. The absence of cobalt in this recommendation reflects the deteriorating influence of certain cobalt compounds on fatigue life after humid aging (ref. 3).

A comparison of this compound to two non-silica black controls provides an excellent illustration of the ability of the disc fatigue procedure to discriminate among compounds of equal static pull-out bond strengths (table 3). In this case, the apparently equal resistance to humid aging predicted by static pull-out contrasts sharply to the catastrophic bond fatigue failures of the black compounds. The outstanding tire performance record of silica containing wire skims offers considerable credibility to these disc fatigue predictions.

Silica non-resin skim compounds

Many of the more recent brass adhesion studies with silica have involved non-resin systems, that is, formulas in which the in situ resorcinol based resin is omitted. In these systems the bond is derived solely from the sulfur reaction with copper with modification by silica, cobalt and other materials.

One of the first dynamic adhesion studies of a non-resin system explored the effect of black replacement by 12 and 24 phr silica at three sulfur concentrations (table 4). Comparable 20% modulus was maintained by reduction in oil content at higher silica levels. Silica-sulfur contour curves of disc fatigue failure (figure 2) indicate that a 24 phr silica presence is required to provide complete freedom from wire separation - under test conditions of 16% deformation, 70 [degrees] C, and eight hours' flexing. A sulfur optimum appears at 5.5 phr. Tabulation of various static and dynamic properties for three silica levels (with sulfur constant at 5.5 phr) illustrates the influence of silica (table 5). Of most interest is the irrelevance of Monsanto fatigue-to-failure data which show that fatigue life remains essentially unchanged at all silica levels. This lack of correlation between vulcanizate fatigue and bond fatigue emphasizes that true metal adhesion - rather than rubber cohesion - is determined by the disc fatigue procedure.

Among other dynamic vulcanizate properties, no significant change is found in either rebound resilience or storage modulus (E[prime]). Tan delta shows a decrease typical of silica replacement of carbon black.

Any analysis of the effect of filler type on vulcanizate properties must recognize the influence of hardness and low strain modulus (i.e. 1 to 20% modulus). As previously discussed, disc fatigue flexing occurs at constant deformation. Thus, test specimens of higher hardness will be subject to greater stress at the wire interface, and suffer a higher incidence of fatigue separations. In the case under discussion, the bonding effect of the silica has reduced the fatigue separations to zero notwithstanding a slight increase in durometer and 20% modulus of the 24 phr compound.

The role of cobalt

Previous work with silica-resin brass bonding systems indicated that the bond fatigue life of humid aged specimens was degraded by the presence of certain cobalt compounds. Since the non-resin formulas described above all contained 3 phr cobalt borate neodecanoate (CP216), further work was required to investigate a range of concentrations, as well as the possibility of omitting cobalt entirely. Sulfur concentration was the second variable. Contour plots of disc fatigue data for specimens humid aged five days at 90 [degrees] C (figures 3 and 4) show little dependence of fatigue separation on cobalt (including zero) within a sulfur range of 4 to 6 phr. Above 6 phr of sulfur and 1 phr of cobalt, separation failures quickly reach the maximum of six. A similar relationship is seen in a graph of the pull-out values of the flexed, bisected specimens (figure 4). The sulfur optimum, as noted in the initial study, remains in the 5.0 to 5.5 area. At low or zero cobalt concentrations (figure 3B) the number of separations depends only on sulfur content. It thus appears that a brass bond resistant to conditions of high humidity and temperature can be attained solely from optimum concentrations of sulfur and silica, without the use of cobalt. Formula recommendations based on these conclusions together with vulcanizate and processing properties appear in table 6. Silane treated silica (Ciptane) has provided similar adhesion properties with a faster cure rate and lower heat build-up. Among processing variables, the effect of cure temperature has been investigated. Initial data show that an increase in temperature from 160 [degrees] C to 180 [degrees] C produced fewer disc fatigue failures in the 24 phr silica non-resin compound. The effect of test temperature on bond fatigue is one of several studies now in progress.

Summary

The disc fatigue procedure for dynamic evaluation of brass adhesion properties provides an opportunity to measure rubber-to-metal bond fatigue life unencumbered by rubber cohesive failure. As a result of disc fatigue studies, two brass skim compounds can be recommended: one which combines silica with bonding resins, and one in which silica and sulfur are the only bonding ingredients.

Table 1 - hardness influence on disc fatigue separations

PHR

Hi-Sil 210                       10                   20
N330                             55                   45
Naphthenic oil             0      6     12      0     6      12
Durometer A               72     71     67     69     68     61
20% Modulus A. PSI       180    175    145    170    150    110
            B. MPa      1.24   1.20   1.20   1.16   1.03   0.75
DF failures              2.5    1.8     0     2.6    0.3    0.3
DF pull-outs lbs./inch   195    190    180    180    195    185

Table 2 - Silica RF resin skim compound

Formula                  Properties:
Natural rubber     100   Processing
Hi-Sil 210          15   ORD 160 [degrees] C
N330                50      T2                      2.0
HPPD                 1      T90                     6.0
TM Q                 1   Mooney
Napthenic oil        5   scorch 130 [degrees] C
Stearic acid         2      T5                      9.5
Zinc oxide           6      ML 100                   52
R-F resin            3   Vulcanizate
Second stage             Durometer                   70
Sulfur               4   M300                     2,000
HMMM                 1   TB                       3,500
TBBS               1.2   EB                         500
                 189.2

Table 3 - disc fatigue evaluation of wire skim compounds

Compound features

                            Silica   Black     Black
                             resin   resin   non-resin
Hi-Sil 210                      15       -       -
N330                            50      60      70
Cobalt CP216                     -       2       3
R-F resin                        3       3       -
HMMM resin                       1       1       -
Sulfur                           4       4     5.5
TBBS                           1.2     0.7     1.2

Disc fatigue dynamic adhesion Conditions: 16%; 70 [degrees] C; 4 hours; 60 Hz Sample conditioning: 5 days 90 [degrees] C; 100% RH

Failures/6 specimens             0       6       4
Flexed pull-out lbs./inch      190     100     120

Static pull-out (ASTM D2229) Sample conditioning: 5 days; 90 [degrees] C; 100% RH

  Lbs./0.5 inch                 85      80      80
  % Cover                       60      30      70

Table 4 - non-resin compounding study

Formula: SMR - 100; Hi-Sil 210 - var.; N326 - var.; oil - var. HPPD - 2; stearic - 2; ZnO - 6; CP216 - 3 Sulfur - var; MBS - 1.2 Variables:

Hi-Sil 210     0     0     0    12    12    12    24    24    24
N326          70    70    70    58    58    58    46    46    46
Naph. oil      8     8     8     4     4     4     0     0     0
Sulfur       4.5   5.5   6.5   4.5   5.5   6.5   4.5   5.5   6.5

Table 5 - non-resin compounding study vulcanizate properties

Silica, PHR                              0      12      24
Disc fatigue separations                 5       3       0
Monsanto fatigue, Kc                    10      11      12
Durometer: original                     76      77      79
           aged 1D90 [degrees] C        82      81      83

Pendulum rebound, %

at 23 [degrees] C                     50.1    52.5    50.1
at 100 [degrees] C                    63.6    66.3    64.1

DMA: 60 [degrees] C; 1Hz

E [prime], MPa                        21.6    20.1    20.0
Tan delta                            0.157   0.134   0.122

Stress-strain; humid aged 1 day

20% Modulus, PSI                       320     315     380
Tensile                              2,880   2,700   2,600
Elongation, %                          300     330     350

[Tabular Data Omitted]

PHOTO : Figure 2 - disc fatigue dynamic wire adhesion

PHOTO : Figure 3 - cobalt and sulfur effects

PHOTO : Figure 4 - cobalt and sulfur effects

References

[1]M.P. Wagner and N.L. Hewitt, Journal of Elastomers & Plastics, April, 1978. [2]M.P. Wagner and N.L. Hewitt, Rubber Chem. & Tech., September, 1979. [3]M.P. Wagner and N.L. Hewitt, Kautschuk & Gummi, October, 1984. [4]N.L. Hewitt, Rubber World. November 1984. [5]N.L. Hewitt, Rubber World. May, 1990.