Band 9, Heft 4

The VIC mixer is an intermeshing internal mixer, whose unique feature is its ability to vary the clearance between the two rotors. The flow field in the chamber of an internal mixer is difficult to analyze due to the complex geometry and transient character of the flow. Here, a fluid dynamics analysis package – FIDAP, using the finite element method was employed to simulate the flow patterns in a VIC 1.9 mixer. A 3D flow analysis was carried out for the whole mixing chamber. The problem of time dependent flow boundaries was solved by selecting a number of sequential geometries to represent a complete mixing cycle. The flow field was characterized in terms of velocity profiles, pressure contours, shear stresses generated and a parameter λ quantifying the elongational flow components. The last two parameters are the most important ones in analyzing dispersive mixing efficiency. Distributive mixing was studied numerically by means of tracking the evolution of particles originally gathered as clusters. The pairwise correlation function and the particle concentration were used to quantify the degree of mixing. The dynamic average flow field characteristics based on the particle tracking were also calculated. The effect of changing gap size on dispersive and distributive mixing was investigated.

Data regarding the influence of screw geometry on homogeneity and fibre length reduction during single screw extrusion of short glass fibre reinforced PP are presented. Three screw geometries, two grades of PP with different melt flow indices, and four levels of filling (10 to 40 wt.%) were the main variables studied. The fibre breakage mechanism was evaluated with respect to solid and melt conveying and melting mechanisms in the screw channels. The screw with a long compression zone produced a material of poor homogeneity, a stable melting mechanism and a medium level of fibre length reduction. The very short compression zone screw produced unstable melting which considerably improved the homogenization process of the fibres in the polymer matrix. In this way the extrudates remained smooth up to the 30 wt.% filling level. The homogeneity was comparable to that obtained by a process including two steps of compounding: twin screw and single screw extrusion. However, the fibre degradation in the screw with a very short compression zone was substantially lower than in the process including two compounding steps.

Procedures are described for computation of pressure, fill factor and temperature profiles in a molular co-rotating twin screw extruder with multiple feed ports and vents. This is a modification of our earlier Akro Co-Twin Screw software. Sample calculations are given for non-isothermal flow simulation of compounding of thermoplastics with talc. It is shown that the effect of temperature on viscosity plays a significant role in pressure fill factor, and temperature profiles and in overall temperature rise.

The effect of reactive polymer blending on the rheological properties and the deformation of dispersed phase in extruded films is described. Three types of polymer blends are studied: (1) a non-reactive blend of nylon 6 (PA 6)/isobutylene-co-para-methylstyrene elastomer; (2) a compatibilized blend of PA 6/brominated (isobutylene-co-para-methylstyrene) elastomer (XP-50); (3) a compatibilized, dynamically vulcanized blend of PA 6/XP-50 plus curatives. Mixing is performed with a Brabender mixer and a twin screw extruder. Rheological measurements were performed at 240 °C which is above melting temperature of PA 6. The storage moduli and complex viscosities of type (3) are the highest, those of type (2) are the next and those of the mechanically mixed non-reactive components are the lowest. The dispersed phases in extruded films show that the deformation of the dispersed elastomer in type (3) is the smallest. The sizes of the dispersed-elastomer domains in type (2) and type (3) are much smaller than those of type (1). There are large and highly deformed dispersed phases in the film of type (1). Also the interfacial bonding is poor.

A mathematical model is set up for melt flow and heat transfer in a slit die, coupled with heat transfer in the die body. The model is reduced to two dimensions by neglecting variations across the width of the die. Conduction, convection and viscous dissipation in the non-Newtonian melt are included, together with conduction and electrical heating in the die body and heat losses by free convection and radiation from the die surfaces. The model is implemented using finite elements, and provides detailed temperature fields on the die cross-section, displaying, for the first time, the effects of thermal coupling between the melt and die body. The consequences for processing are discussed for several situations, including horizontal and vertical operation of the die, with a polished or corroded surface, and with or without insulation. The appropriate use of dimensionless groups for the interpretation of the numerical results is discussed. A criterion in terms of a Pearson number is developed, for assessing whether an isothermal melt flow analysis is applicable. A simple analytic result is proposed for calculating the die set temperature required to minimise melt temperature variations. This is tested against the numerical results and is broadly validated.

This paper describes first results obtained in our group on the study of the reactive extrusion process for maleation of polypropylene in the presence of a radical initiator (peroxide). In accordance with the conclusions obtained in a previous paper the viability of such reactive processing is confirmed. Furthermore the necessity of a careful study about the values of the experimental process conditions in order to know the set of conditions which maximize the grafted maleic anhydride into the polypropylene, can be clearly concluded. We observe that the same mechanism seems to be governing the grafting reaction in the continuous and in the batch process. This can clearly be concluded from the results shown in this paper and in previous work.

A viscoelastic model for PET cast film extrusion is developed on the basis of some important assumptions: constant width, vertical film, isothermal regime, Maxwell's constitutive equation. The general model is non-stationary, but it can be first used to cope with steady-state extrusion, which makes it possible to introduce the concept of “unattainable region”, as in fiber spinning. Then the dynamic responses to random or sinusoidal fluctuations of the chill-roll velocity are analyzed. The predictions of the thickness response are in good agreement with experimental observations.

The draw resonance instability in a concentric bicomponent coextrusion fiber spinning has been studied experimentally using a linear low density polyethylene (LLDPE) for the core and two low density polyethylenes (LDPE) for the skin-layer. These materials have been chosen to investigate a coextrusion flow in which each components have quite different extensional rheology. When they were spun individually, the LLDPE showed the draw resonance at a certain critical draw ratio whereas the LDPEs did not show the oscillatory instability. When the LLDPE and LDPE were spun together, the presence of the more viscoelastic LDPE layer delayed the onset of draw resonance of the LLDPE to a higher draw ratio. The degree of changes in the critical draw ratio was observed to be dependent on the rheological properties of the LDPEs. Although the experiments were conducted in a non-isothermal condition, various aspects of the experimental observations appeared to be in qualitative agreement with the predictions of isothermal linear theories. The present study is specifically for a coextrusion fiber spinning. Nevertheless, the observed trend of the instability should be also relevant to other coextrusion processes such as extrusion coating or film casting.

A novel theory which employs the K-BKZ viscoelastic integral constitutive model was developed in this study for simulating the non-isothermal injection molding filling process and the frozen stresses of a three-dimensional thin part. The simulation of viscoelastic model in such a problem has not yet been discussed in most previous works as a result of the complexity of this problem. A quasi-steady approximation concept was developed in the present investigation for solving the non-isothermal filling process via K-BKZ viscoelastic integral constitutive model. Additionally, the numerical method of flow field was based on the control volume finite element method. The flow field of generalized Newtonian fluid was used as the initial guess of flow kinematics. The quasi-steady state approximation was introduced for each element to calculate the flow kinematics and stress profile from the K-BKZ integral constitutive model. The finite difference method with streamline upwind characteristic was adopted here for calculating the temperature field of process. When the cavity is fully filled, the subsequent non-isothermal stresses would relax after cessation of flow. Thereby, the frozen stresses (or frozen birefringence) could be obtained.