Fillers time effects

In spite of the often large contribution of secondary filler aggregation effects, measurements of the time-temperature dependence of the linear viscoelastic functions of carbon filled rubbers can be treated by conventional methods applying to unfilled amorphous polymers. Thus time or frequency vs. temperature reductions based on the Williams-Landel-Ferry (WLF) equation (162) are generally successful, although usually some additional scatter in the data is observed with filled rubbers. The constants C and C2 in the WLF equation... 

Another explanation was the following. The organomontmorillonite used was a natural montmorillonite that contained iron. Chemical analysis of the clay confirmed the presence of a low amount of iron. It was recalled that iron and, in more general terms, metals are likely to induce the photochemical degradation of polymers. Iron at low concentration had a prooxidant effect that was due to the metal ion of iron that can initiate the oxidation of the polymer by the well-known redox reactions with hydroperoxides [93]. It was concluded that the transition metal ions, such as Fe, displayed a strong catalytic effect by redox catalysis of hydroperoxide decomposition, which was probably the most usual mechanism of filler accelerating effect on polymer oxidation. A characteristic of such catalytic effect was that it did not influence the steady-state oxidation rate, but it shortened the induction time. 

The time independent Mullins Effect can account for the near equilibrium hysteresis observed in propellants at low strains, but cannot account for the nonlinear time effects [1,2]. There is considerable evidence however, that the Mullins Effect in propellants is a very strong function of time [1,2]. Time dependent chain failure can be readily demonstrated by simply examining the influence of filler on viscoelasticity and some of the routine tests run on solid propellants. 

Another example of a time dependent Mullins Effect is that of a cross-linked polymer, with little or no time dependency when unfilled, which becomes significantly time dependent when filled [42], as shown in Figure 3.2. The more filler incorporated into the system, the more marked the time effect. Many propellant polymers fall into this category and nearly all propellants show time dependence over such long times that true equilibrium data cannot be obtained. This time dependence in the composite material and no time dependence in the unfilled pol3rmer cannot be... 

In summary, then, design with polymers requires special attention to time-dependent effects, large elastic deformation and the effects of temperature, even close to room temperature. Room temperature data for the generic polymers are presented in Table 21.5. As emphasised already, they are approximate, suitable only for the first step of the design project. For the next step you should consult books (see Further reading), and when the choice has narrowed to one or a few candidates, data for them should be sought from manufacturers data sheets, or from your own tests. Many polymers contain additives - plasticisers, fillers, colourants - which change the mechanical properties. Manufacturers will identify the polymers they sell, but will rarely disclose their... 

The ebonite compound before cure is a rather soft plastic mass which may be extruded, calendered and moulded on the simple equipment of the type that has been in use in the rubber industry for the last century. In the case of extruded and calendered products vulcanisation is carried out in an air or steam pan. There has been a progressive reduction in the cure times for ebonite mixes over the years from 4-5 hours down to 7-8 minutes. This has been brought about by considerable dilution of the reactive rubber and sulphur by inert fillers, by use of accelerators and an increase in cure temperatures up to 170-180°C. The valuable effect of ebonite dust in reducing the exotherm is shown graphically in Figure 30.3. 

Obviously, the discrepancy between the experimental data [238-241] and predictions of the theory [236,237] can be attributed to the difference of the coefficients of thermal expansion. The polymer exerts pressure on the filler, thereby masking the effect of the strength of adhesion on the modulus. The pressure on the filler may be sufficiently high. In [243] it was found, for example, that in PP, quartz particles experienced a compression force of about 100 MPa after cold drawing of the composite the force reduces to 50 MPa in the direction of drawing but at the same time increases to 300 MPa in the perpendicular direction. 

Data of Figs 8-10 give a simple pattern of yield stress being independent of the viscosity of monodisperse polymers, indicating that yield stress is determined only by the structure of a filler. However, it turned out that if we go over from mono- to poly-disperse polymers of one row, yield stress estimated by a flow curve, changes by tens of times [7]. This result is quite unexpected and can be explained only presumably by some qualitative considerations. Since in case of both mono- and polydisperse polymers yield stress is independent of viscosity, probably, the decisive role is played by more fine effects. Here, possibly, the same qualitative differences of relaxation properties of mono- and polydisperse polymers, which are known as regards their viscosity properties [1]. 

According to the concepts, given in the paper [7], a significant difference between the values of yield stress of equiconcentrated dispersions of mono- and polydisperse polymers and the effect of molecular weight of monodisperse polymers on the value of yield stress is connected with the specific adsorption on the surface of filler particles of shorter molecules, so that for polydisperse polymers (irrespective of their average molecular weight) this is the layer of the same molecules. At the same time, upon a transition to a number of monodisperse polymers, properties of the adsorption layer become different. 

Though the accuracy of description of flow curves of real polymer melts, attained by means of Eq. (10), is not always sufficient, but doubtless the equation of such a structure based on the idea of relaxation mechanism of non-Newtonian polymer flow, correctly reflects the main peculiarities of viscous properties. Therefore while discussing the effect a filler has on the viscosity properties of polymer melts, besides the dependences Y(filler modifies the characteristic time of relaxation. According to [19], a possible form of the X versus

[Pg.86]

The fact that the appearance of a wall slip at sufficiently high shear rates is a property inwardly inherent in filled polymers or an external manifestation of these properties may be discussed, but obviously, the role of this effect during the flow of compositions with a disperse filler is great. The wall slip, beginning in the region of high shear rates, was marked many times as the effect that must be taken into account in the analysis of rheological properties of filled polymer melts [24, 25], and the appearance of a slip is initiated in the entry (transitional) zone of the channel [26]. It is quite possible that in reality not a true wall slip takes place, but the formation of a low-viscosity wall layer depleted of a filler. This is most characteristic for the systems with low-viscosity binders. From the point of view of hydrodynamics, an exact mechanism of motion of a material near the wall is immaterial, since in any case it appears as a wall slip. 

A possible approach to interpretation of a low-frequency region of the G ( ) dependence of filled polymers is to compare it with a specific relaxation mechanism, which appears due to the presence of a filler in the melt. We have already spoken about two possible mechanisms — the first, associated with adsorption phenomena on a filler s surface and the second, determined by the possibility of rotational diffusion of anisodiametrical particles with characteristic time D 1. But even if these effects are not taken into account, the presence of a filler can be related with the appearance of a new characteristic time, Xf, common for any systems. It is expressed in the following way... 

---