Filler network effect

Modeling Dynamic Stress Softening as a Filler Network" Effect... 

It should be noted that for polymerization-modified perlite the strength parameters of the composition algo go up with the increasing initial particle size. [164]. In some studies it has been shown that the filler modification effect on the mechanical properties of composites is maximum when only a portion of the filler surface is given the polymerophilic properties (cf., e.g. [166-168]). The reason lies in the specifics of the boundary layer formation in the polymer-filler systems and formation of a secondary filler network . In principle, the patchy polymerophilic behavior of the filler in relation to the matrix should also have place in the failing polymerization-modified perlite. 

In addition to increases in high-strain loss modulus, reductions in low-strain loss modulus are also observed. This may be attributed to the improvements in polymer-filler interactions which may reduce the amount of filler networking occurring in the compound. The low-strain losses are dominated by disruptions in the filler-filler network, the Payne effect. 

The interaction between two fillers particles can be investigated by measuring the Payne effect of a filled rubber compounds. In this measurement, dynamic properties are measured with strain sweep from a very small deformation to a high deformation. With the increased strain, the filler-filler network breaks and results in a lower storage modulus. This behavior is commonly known as the Payne effect... 

Accordingly, we expect a power law behavior G,0 (O/Op)3 5 of the small strain elastic modulus for 0>0. Thereby, the exponent (3+df [j)/(3—df)w3.5 reflects the characteristic structure of the fractal heterogeneity of the filler network, i.e., the CCA-clusters. The strong dependency of G 0 on the solid fraction Op of primary aggregates reflects the effect of structure on the storage modulus. 

This paper represents an overview of investigations carried out in carbon nanotube / elastomeric composites with an emphasis on the factors that control their properties. Carbon nanotubes have clearly demonstrated their capability as electrical conductive fillers in nanocomposites and this property has already been commercially exploited in the fabrication of electronic devices. The filler network provides electrical conduction pathways above the percolation threshold. The percolation threshold is reduced when a good dispersion is achieved. Significant increases in stiffness are observed. The enhancement of mechanical properties is much more significant than that imparted by spherical carbon black or silica particles present in the same matrix at a same filler loading, thus highlighting the effect of the high aspect ratio of the nanotubes. 

Even dynamic measurements have been made on mixtures of carbon black with decane and liquid paraffin [22], carbon black suspensions in ethylene vinylacetate copolymers [23], or on clay/water systems [24,25]. The corresponding results show that the storage modulus decreases with dynamic amplitude in a manner similar to that of conventional rubber (e.g., NR/carbon blacks). This demonstrates the existence and properties of physical carbon black structures in the absence of rubber. Further, these results indicate that structure effects of the filler determine the Payne-effect primarily. The elastomer seems to act merely as a dispersing medium that influences the magnitude of agglomeration and distribution of filler, but does not have visible influence on the overall characteristics of three-dimensional filler networks or filler clusters, respectively. The elastomer matrix allows the filler structure to reform after breakdown with increasing strain amplitude. 

Another important point is the question whether static offsets have an influence on strain amplitude sweeps. Shearing data show that this seems not to be the case as detailed studied in [26] where shear rates do not exceed 100 %.However, different tests with low dynamic amplitudes and for different carbon black filled rubbers show pronounced effects of tensile or compressive pre-strain [ 14,28,29]. Unfortunately, no analysis of the presence of harmonics has been performed. The tests indicate that the storage (low dynamic amplitude) modulus E of all filled vulcanizates decreases with increasing static deformation up to a certain value of stretch ratio A, say A, above which E increases rapidly with further increase of A. The amount of filler in the sample has a marked effect on the rate of initial decrease and on the steady increase in E at higher strain. The initial decrease in E with progressive increase in static strain can be attributed to the disruption of the filler network, whereas the steady increase in E at higher extensions (A 1.2. .. 2.0 depending on temperature, frequency, dynamic strain amplitude) has been explained from the limited extensibility of the elastomer chain [30]. 

Otherwise, the independence of static offsets at moderate shearing rates can be interpreted in such a way that under such conditions no or only slight disruption of the filler network appears. So far this difference between shearing and tension/compression remains more or less unclear and more experiments are needed to clarify this point. We speculate that under static shearing only slight disruption of the filler network appears and the main effect is only a shear deformation of the shape of filler clusters and its backbone (see Fig. 2). 

The filler network break-down with increasing deformation amplitude and the decrease of moduli level with increasing temperature at constant deformation amplitude are sometimes referred to as a thixotropic change of the material. In order to represent the thixotropic effects in a continuum mechanical formulation of the material behavior the viscosities are assumed to depend on temperature and the deformation history [31]. The history-dependence is implied by an internal variable which is a measure for the deformation amplitude and has a relaxation property as realized in the constitutive theory of Lion [31]. More qualitatively, this relaxation property is sometimes termed viscous coupling1 [26] which means that the filler structure is viscously coupled to the elastomeric matrix, instead of being elastically coupled. This phenomenological picture has... 

The discussion in the Introduction led to the convincing assumption that the strain-dependent behavior of filled rubbers is due to the break-down of filler networks within the rubber matrix. This conviction will be enhanced in the following sections. However, in contrast to this mechanism, sometimes alternative models have been proposed. Gui et al. theorized that the strain amplitude effect was due to deformation, flow and alignment of the rubber molecules attached to the filler particle [41 ]. Another concept has been developed by Smith [42]. He has indicated that a shell of hard rubber (bound rubber) of definite thickness surrounds the filler and the non-linearity in dynamic mechanical behavior is related to the desorption and reabsorption of the hard absorbed shell around the carbon black. In a similar way, recently Maier and Goritz suggested a Langmuir-type polymer chain adsorption on the filler surface to explain the Payne-effect [43]. 

---