Band 9, Heft 3

Attempts to find a general viscoelastic fluid model which would be applicable to all different classes of flow problems and could produce experimentally verifiable results, have so far failed. Instead efforts have been concentrated in finding a suitable viscoelastic model for the particular class of flow problem in hand. Bearing this in mind, a simple viscoelastic fluid model which has shown promising results for cases where there is a dominant flow direction (i.e. the flow is viscometric) is considered here. Such conditions may occur inside a rubber mixer where in cylindrical coordinate system (r,θ,z), the elastomer is flowing dominantly in θ direction. This paper deals with the Galerkin finite element modelling of viscometric flows under non-isothermal conditions. The novel aspect of this work is the extension of the application of a viscometric constitutive equation known as (CEF) model to the simulation of predominantly circumferential flow in a two-dimensional representation of an internal mixer using polar coordinates. For a given set of boundary conditions, the mathematical model can simulate the velocity, pressure and temperature fields inside a mixer. Comparison of the numerical results with the available experimental evidence confirm the general validity of the mathematical model used.

Previous work by Benkreira et al. [1] has shown that the critical stage in distribution of additives in single screw extruders is the zone where the polymer melts; the subsequent pumping of the melt makes little difference to the mixture quality. It is suggested that the ideal masterbatch melts before the host polymer and coats each granule, while remaining sufficiently viscous to promote the shear heating and melting of the polymer. In order to test this hypothesis and quantify the requirements in terms of relative melting points and melt viscosities, a set of model masterbatches was produced and used for injection moulding at 1 % addition in a variety of polypropylenes and polyethylenes, but always under the same moulding conditions. Image analysis and “experts” judgement was used to rate the performance of these master-batches. The conclusions drawn from this work are that the designer of a masterbatch must consider the melting ranges of the masterbatch and its host polymer, and their relative viscosities in the temperature range of the forming process, in order to optimise the ease of distribution of the masterbatch.

The dispersion of an elastomeric phase (blend of two miscible copolymers: ethylene-vinyl acetate and ethylene-methyl acrylate) in a polypropylene matrix along a corotating twin-screw extruder was studied experimentally. A sliding barrel allowed a direct access to the screws and thus permitted a rapid cooling and sampling. The dispersion of the elastomeric phase was controlled by scanning electron microscopy and the size distribution was obtained by image analysis. The particle size distribution was measured at different locations along the screws in order to quantify the evolution of the morphology. Complementary studies were carried out on a pre-shearing rheometer (Rheoplast) in order to study the influence of the mechanical treatment upon particle size. The dispersion of the elastomeric phase was observed at the capillary exit by the above mentioned techniques. The results of these different experiments enhance the importance of the melting mechanism and put in evidence that the final morphology of the blend is controlled by the opposite mechanisms of break up and coalescence.

An in-line rheometer designed for use as a sensor for closed-loop control of melt processing operations has been previously described. This instrument can measure the true viscosity of a melt stream over a range of shear rates. In order to evaluate its performance in processing applications the rheometer was mounted on a twin-screw extruder. A chemical degradation process was used to introduce changes in the molecular weight distribution, and thus the viscosity of a polypropylene resin, and these changes were monitored using the rheometer. For rapid replacement of melt in the zone of controlled shear in the rheometer, flow simulations indicated that it would be advantageous to operate the rheometer in a cyclic mode. A stochastic process identification technique based on a pseudorandom binary signal was used to determine the parameters of a first-order plus deadtime model of the process. In order to estimate the portion of the system response due to rheometer dynamics, residence time distributions were measured both with and without the rheometer present. It was found that the rheometer has a dead time of about 6's and a first order time constant of less than 15 s. The response of the new instrument is thus significantly faster than than that of melt process rheometers now commercially available.

The most widely used pieces of equipment in plastics compounding are continuous mixers and intermeshing corotating twin-screw extruders. Key design aspects and processing characteristics of these two machines are compared and examples of processing typical compounds in both types of equipment are presented and discussed.

An experimental study of the peroxide induced degradation of polypropylene is described. Experimental studies in a batch mixer, a single screw extruder and in a Leistritz modular counter-rotating twin screw extruder are presented. Three different screw configurations were used in the latter machine. When comparisons are made at the same temperature, peroxide level and residence time, the greatest extent of degradation is found in the twin screw extruder and the poorest in the single screw extruder. These differences seem associated with the rapid effective melting in the twin screw extruder and the slow melting and poor mixing in the single screw extruder. Of the three modular twin screw configurations, the configuration containing no intense shearing elements is most effective. Experiments based on removing the modular screw and following the reaction as a function of position suggests that Leistritz shearing elements cause rapid heating and reaction of the peroxide before it is mixed into the polypropylene matrix.

In this work structure and mechanical properties of blends of bisphenol-A polycarbonate with polyamide 6 with different molecular weights and different terminal groups are presented. The results show the important role played in the change of the morphology and in the improvements of the mechanical properties by the increasing content of NH 2 terminal groups. These results are attributed to the reactions during melt mixing of such groups with polycarbonate.

The Box-Wilson experimental design method is employed to study the free radical process between polypropylene and maleic anhydride in the presence of a peroxide like radical initiator. The batch processes in the solution and in the molten state have been considered in an attempt to know how the physical characteristics of the reaction medium are affecting the modification level of the polypropylene. The analytical titration method proposed by Gaylord is used in order to evaluate the maleic anhydride grafting onto the polypropylene. The main goal of this preliminary study is to know and to show the quantitative values reached in the grafted maleic anhydride level inside the experimental range scanned. From the results the influence of the reaction and several observations can be made in order to evaluate and finally optimize the continuous process by reactive extrusion.

Melt strength data are presented for polypropylenes, differing widely in molecular weight and molecular weight distribution. Melt strength was assessed from uniaxial tensile experiments with a Rheotens apparatus. A new measuring procedure was applied by performing the stretching in front of the take-up wheels of the Rheotens instead of in between the wheels. It is shown that the melt strength of polypropylene is solely determined by its weight average molecular weight. At a given melt index level, narrow molecular weight distribution (controlled rheology) polypropylenes have a lower melt strength than the wider distribution reactor grades. At high temperatures, the melt strength follows an Arrhenius type temperature dependence, while at the lower test temperatures the melt strength of polypropylenes having a wider distribution is increased by flow-induced crystallisation.

The modelling of multilayer reactive injection moulding is studied. After presenting the equations describing the process, two important aspects are dealt with in more detail. These are particle tracking and the modelling of the flow front. First results demonstrate the possibilities of the modelling and show satisfactory agreement between predictions of numerical simulations and experimental results, concerning particle tracking in the whole flow domain and the break-through of layers in sequential two component moulding.

As a part of continuous study to search for the high performance in situ composite, this study examines the miscibility of liquid crystalline poly(ester imide) (PEsI) with poly(ether imide) (PEI) and its effect on the mechanical properties and the morphology of the blends of these polymers. Liquid crystalline polymers (LCPs) were synthesized from trimellitic anhydride and α,ω-diaminoalkane with 8, 10, 12 methylene groups. Thermal condensation of these diacids with 4,4′-diacetoxybiphenyl yielded the corresponding imides of thermotropic homopolyesters. Optical polarized micrographs show that the synthesized poly(ester imides) have the smectic structure. Partial miscibility was observed from the shift of the glass transition temperature determined by differential scanning calorimetry. Increasing concentration of the liquid crystalline PEsI in the PEI/PEsI blends shifted the glass transition temperature of the PEI phase to a lower temperature. Mechanical properties of the PEI/PEsI blends were changed with the number of methylene units in PEsI. Morphologies of the partially miscible system were found to be quite different from those of an immiscible system. Strong adhesion between the matrix (PEI) and dispersed phase (LCP) was observed, but fine fibril formation was not possible due to the absence of smectic-nematic transition.

Finished goods manufactured by the extrusion blow moulding process are strongly dependent on the parison formation stage, since it is the primary step in the process. The magnitude of the die swell and the degree of sag exhibited by the parison, after being extruded from an annular die, are dependent on the material's molecular structure and strongly affect the performance of the final product. Repassing or recycling of polymeric materials results in degradation reactions that modify the molecular structure, and therefore of parison swell and sag. Polyethylene simultaneously undergoes crosslinking and chain scission. Polypropylene undergoes predominately chain scission. The extent of these degradation reactions can be minimized with the use of appropriate stabilizers. This work involves the study of the effects of repassing on the parison formation stage of extrusion blow moulding. The repassing of the same material gives a representation of what occurs when recycling is undertaken. The effect of stabilizer addition is also studied. Particular emphasis is placed on parison swell and sag.

Reaction injection molding (RIM) involves the injection of low viscosity polymers into mold cavities at relatively high injection speeds. The short cycle times, coupled with low injection pressures and clamping forces make RIM well suited for the rapid production of complex parts. High injection speeds, low viscosities and narrow cavities lead to high Reynolds' number flows during mold filling. The presence of high Reynolds' number flows is a notable difference between RIM and conventional thermoplastic injection molding (TIM). Unfortunately, this type of flow with significant inertia effects can lead to reduced part quality. The material properties and characteristics of RIM parts are influenced by flow patterns, weld lines, air entrapment as bubbles or pockets, especially as the material flows around inserts. These defects and characteristics, coupled with a growing demand for high quality RIM parts, is responsible for the increased interest in the flow during mold filling. This paper presents the development and test of a model to stimulate high Reynolds' number flows in thin cavities present during reaction injection mold filling.