SCALE UP EFFECT OF INTERNAL MIXERS

In the development of a new polymer and a new formulae of rubber compound, it has been long expected to develop a small laboratory-scale internal mixer which reproduces the same physical properties as the ones produced by the large industrial-scale mixer. The rubber compound may be characterized in terms of their physical properties. The properties are affected by the mixer itself and the mixing conditions. We recently designed a laboratory-scale mixer, which was Banbury-type and whose batch volume was 250cc. Various rotors with different rotor shape parameters and mixing conditions were arranged to make an experimental design. In order to characterize the mixing behavior and the physical properties of rubber compounds, we found that most important were the following factors: (1) the unit-work which is an integrated torque value over mixing time (external work) per batch volume, (2) the Mooney viscosity of dumped rubber compound, (3) the bound rubber which is the amount of polymer unextracted from the compound by the solvent even after 24 hrs and (4) the weight average molecular weight of the extracted polymer. In other words, when these four factors are close enough to each other for the compounds made by the small and large scale mixers, it may be possible to judge two compounds are almost same in quality. Further, each of the four factors was obtained as a function of the mixing conditions and the rotor shape parameters by means of the multiple regression analysis. In the experiments, three kinds of rubbers, that is, styrene butadiene rubber (JSR SBR1778N), ethylene propylene rubber (JSR EP96) and butadiene rubber (JSR BR01) were used, and each recipe of them was commonly applied in industory. Among the rotor parameters, the rotor tip clearance, the rotor tip width, the total bulkiness of rotor and the overlap ratio of wings had considerable influences. Among the factors of mixing conditions, the mixing time, the rotor speed and the temperature had significant influences. The results of the experiment were analysed by the multiple regression analysis methods. The optimum parameters of a laboratory-scale mixer for the reproduction of the mixing of the industrial mixer were found to be (1) larger rotor tip clearance, (2) larger rotor tip width, (3) bigger overlap ratio of wings, (4) higher temperature and (5) higher rotor speed than those of propotionally reduced dimensions and comparable conditions.