Development of polymer blend morphology during compounding in a twin-screw extruder. Part II: Theoretical derivations

In Part II of the work, the intermeshing twin-screw extruder is briefly described and the theoretical procedures used to model its operation are summarized. Based on the microrheological considerations discussed in Part I, a predictive procedure of the morphology evolution during compounding of two immiscible polymers is proposed. In this first generation model, only the shear flow effects are considered. Furthermore, to avoid complications due to coalescence a low concentration of the dispersed phase was assumed. In the procedure, two drop breakup mechanisms are discussed. The first assumes that the drops do not break under flow while the second postulates that breakup occurs under flow. Two dispersion mechanisms are considered, the first postulating continuously increasing polydispersity of drop size and the second postulating that drop polydispersity is inversely proportional to deformation strain. The influence of the screw configuration and operating conditions on blend morphology evolution is studied. It is expected that the computed drop size distribution provides limiting values for the experimental data. Dependency of predicted morphology on operating conditions is also investigated. Increasing screw rotating speed (resulting in increasing energy consumption) and decreasing throughput (resulting in decreasing productivity) lead to prediction of finer drop size. In practice, therefore, a compromise would be required. The proposed procedure is limited to melt flow (excluding the die region) within the region of large capillary parameter values, k > 4k_crit.     

Development of polymer blend morphology during compounding in a twin-screw extruder. Part IV: A new computational model with coalescence

In Part II of this series of publications, the first generation model of morphology evolution during polymer blending in a twin-screw extruder was presented. The model was based on a simplified flow analysis, and an assumption that dispersion occurs via drop fibrillation followed by disintegration. In the present Part IV, several modifications of the model are discussed. (i) The flow analysis was refined by computing the pressure profiles. (ii) The flow paths and strain history of the dispersed droplets within the screw elements are computed directly, which makes it possible to determine the drop susceptibility to deformation and break. (iii) Besides the fibrillation mechanism, a drop-splitting mechanism for low supercritical capillary numbers is incorporated. (iv) The choice of breakup mechanism is based on micro-rheological criteria. (v) The coalescence effects are taken into account. (vi) The theoretical model is self-consistent, without adjustable parameters. The validity of theoretical assumptions was evaluated by comparing the model predictions with the experimental droplet diameters at different positions in the twin-screw extruder.     

Development of polymer blend morphology during compounding in a twin-screw extruder. Part I: Droplet dispersion and coalescence—a review

The theoretical and experimental data on the breakup of droplets are reviewed. Several factors influence development of droplets: flow type and its intensity, viscosity ratio, elasticity of polymers, composition, thermodynamic interactions, time, etc. For Newtonian systems undergoing small, linear deformation, both the viscosity ratio and the capillary number control deformability of drops. On the other hand, the breakup process can be described by the dimensionless breakup time and the critical capillary number. Drops are more efficiently broken in elongational flow than in shear, especially when the viscosity ratio λ ? 3. The drop deformation and breakup seems to be more difficult in viscoelastic systems than in Newtonian ones. There is no theory able to describe the deformability of viscoelastic droplet suspended in a viscoelastic or even Newtonian medium. The effect of droplets coalescence on the final morphology ought to be considered, even at low concentration of the dispersed phase, ?_d ? 0.005. Several drop breakup and coalescence theories were briefly reviewed. However, they are of little direct use for quantitative prediction of the polymer blend morphology during compounding in a twin-screw extruder. Their value is limited to serving as general guides to the process modeling.     

Computation of the morphological changes of a polymer blend along a twin-screw extruder

The control of the morphology of an immiscible polymer melt is of vital importance for the mastering of the final properties of the product. As polymer blends are produced using corotating twin-screw extruders, understanding and modeling the changes experienced by the blend during this process is of great interest. In the present study, starting from Ludovic software, developed for computing flow parameters in the twin-screw extrusion process, we present a computation of the droplet morphology development, based on the basic mechanisms of break-up and coalescence. Depending on the value of a local capillary number and on local flow conditions, different changes may occur: affine deformation, drop splitting, break-up by capillary instability, and coalescence. It is thus possible to follow, all along the screws, the changes in morphology, either for a single particle or for a particle distribution. Examples of these different computations are presented and compared with experimental results. Generally speaking, orders of magnitude of droplet size and tendencies when modifying processing conditions are correctly described, but the model still suffers from the absence of descrption of the melting process.     

Morphology development of PC/PE blends during compounding in a twin-screw extruder

The morphological development of a polycarbonate/polyethylene (PC/PE) blend in a twin-screw extruder was studied using a scanning electron microscope (SEM). The effects of extrusion temperature, viscosity ratio (the ratio of the viscosity of the dispersed phase to that of the matrix), and the screw configuration on the morphology of the PC/PE blend during the extrusion were discussed in detail. It was found that the morphology of the dispersed particles and the interfacial adhesion between the dispersed phase and matrix were both influenced by the extrusion temperature. The dispersed phase had a spheroidal shape and a small size during the high temperature processing, and an irregular shape and a large size when it was processed at low temperature. The PC phase with a lower viscosity was easier to disperse and also to coalesce. Therefore, the deformation of the low-viscosity dispersed phase during the processing was more intense than that of the high-viscosity dispersed phase. By comparing the effects of the different screw configurations on the morphology development of the PC/PE blend, it was found that the melting and breaking up of the dispersed phase were mainly affected in the initial blending stages by the number of the kneading blocks. While a kneading block with a 90 degree staggering angle was used, the size of the dispersed particles decreased and the long fibers were shortened, the large particles were drawn by the additional kneading zone. Finally, all of these structures were completely changed to the short fibers. POLYM. ENG. SCI., 47:14–25, 2007. © 2006 Society of Plastics Engineers     

Polymer blend mixing and dispersion in the kneading section of a twin-screw extruder

This paper examines the mechanisms by which a polymer is dispersed in a co-rotating twin-screw extruder. An experimental investigation of the morphological evolution has been carried out on a 45-mm co-rotating twin-screw extruder. Polyethylene/polystyrene (PE/PS) blends in the low concentration range (i.e., 5–15 wt% of PE) were used as a model system. The following general trends were observed. First, the minor phase right after melting is predominantly in a fibrillar form. Secondly, droplet and fiber diameter at this early stage of compounding are already in the micron or sub-micron range. Even though a wide variety of mixing section configurations were used, the fibers created in the early compounding stages were relatively stable throughout extrusion. Morphological evolution after melting must therefore be discussed in terms of variation in the fiber fraction (i.e., fiber to droplet transition) rather than in a change in particle diameter. A control volume model for the flow in kneading blocks is used to interpret the morphological results and to predict the deformation and breakup of dispersed phase fibers under shear and in absence of coalescence. Theoretical results indicate that fiber breakup under shear is not likely in the kneading block under the normal processing conditions, which is confirmed by morphological observations made at the mixing section exit. The influence of several geometrical parameters on mixing and pumping in kneading blocks is also discussed with the use of flow model results.     

The modeling of continuous mixers. Part I: The corotating twin-screw extruder

In many operations in polymer processing, such as polymer blending, devolatilization, or incorporation of fillers in a polymeric matrix, continuous mixers are used; e.g., corotating twin-screw extruders (ZSK), Buss Cokneaders and Farrel Continuous Mixers. Theoretical analysis of these machines tends to emphasize the flow in complex geometries rather than generate results that can be directly used (1–5). In this paper, a simple model is developed for the hot melt closely intermeshing corotating twin-screw extruder, analogous to the analysis of the single-screw extruder carried out in 1922 and 1928 (6, 7). With this model, and more specifically with its extension to the complete nonisothermal, non-Newtonian situation, it is possible to understand the extrusion process and to calculate the energy, specific energy, and temperature rise during the process with respect not only to the viscosity of the melt, but also to the screw geometry (location and number of transport elements, kneading sections and blisters, pitch, positive or negative, screw clearance, and flight width) and screw speed. To support the theoretical analysis, model experiments with a Plexiglas-walled twin-screw extruder were performed, in addition to practical experiments with melts on small- and large-scale extruders, with very reasonable results, In Part 2, the Buss Cokneader will be analyzed analogously.     

Three-dimensional flow modeling of a self-wiping corotating twin-screw extruder. Part II: The kneading section

Three-dimensional flow simulations of kneading elements in an intermeshing corotating twin-screw extruder are performed by solving the Navier Stokes equations with a finite element package, Sepran. Instead of using the whole geometry of the 8-shaped barrel a simplified geometry is used, representing a large part of the geometry during the rotating action of the kneading paddle. The goal of these calculations is to study the dependence of several factors that influence mixing, such as shear rate, elongation rate, pressure, and the flow profile in the extruder on various extruder parameters, such as fluid viscosity, rotation speed, and throughput. The shear and elongation rate and the pressure drop are calculated for varying viscosities. The various stagger angles possible for disc configurations in the corotating twin-screw extruder are modeled. The axial backflow volume is calculated for varying values of rotation speed and throughput.     

Three-dimensional flow modeling of a self-wiping corotating twin-screw extruder. Part I: The transporting section

A three-dimensional modeling of the transporting elements in a self-wiping corotating twin-screw extruder has been carried out by using the finite element package Sepran (1). This simulation uses the 3D geometry of the channel rolled over the twin-screw, which consists of the intermeshing and normal areas. The flow profile, the backflow volume, the pressure buildup, the shear and elongation rates, and the adiabatic axial temperature gradient have been calculated by solving the Navier-Stokes equations and the continuity equation for a Newtonian fluid. These results are given for different extruder parameters such as the throughput of the extruder, the rotation speed of the screws and the helix angle of the screws to better understand the influence of different extruder configurations. This study belongs to a program of research on the self-wiping corotating twin-screw extruder that also includes the modeling of the kneading elements (Part II) and, in the future, the study of scale-up and heat transfer.     

Morphology development of in situ compatibilized semicrystalline polymer blends in a co-rotating twin-screw extruder

This paper concerns the morphology development of in situ compatibilized semicrystalline polymer blends in a co-rotating, intermeshing twin-screw extruder, using polypropylene (PP) and polyamide 6 (PA-6) blends as model systems. The morphology of in situ compatibilized blends develops much faster that of mechanical ones. The size of the dispersed phase (PA-6) undergoes a 10⁴ fold reduction from a few millimeters to sub-micron during its phase transition from solid pellets to a viscoelastic fluid. The final morphology is reached as soon as the phase transition is completed, which usually requires only a small fraction of the screw length in a co-rotating twin screw extruder. Screw profiles and processing conditions (screw speed, throughput and barrel temperature) control the PA-6's melting location and/or rate, but do not have significant impact on the ultimate morphology and mechanical properties of in situ compatibilized blends. The finding that morphology of PP/PA-6 reactive blend develops rapidly makes it possible to produce compatibilized PP/PA-6 blends by the so-called one-step reactive extrusion. It integrates the traditionally separated free radical grafting of maleic anhydride onto PP and the compatibilization of PP/PA-6 into a single extrusion step.     

Extension-dominated improved dispersive mixing in single-screw extrusion. Part 2: Comparative analysis with twin-screw extruder

In Part 1 of this work, the possibility of improving single-screw extruders (SSE) better dispersive mixer was explored by harnessing extensional flows provided by the hyperbolic contracting–diverging channels of extensional mixing elements (EME). Addition of the EME to the pin screw generated enhanced breakup for polymer blends and nanocomposite systems without significant penalty in flow rate. In Part 2, experiments are performed on immiscible polymer blends (low-viscosity ratio and high-viscosity ratio) and nanocomposites on both SSE and twin-screw extruder (TSE) with the same rotation speed and throughput. Morphological results show tremendous improvement in dispersive mixing capability of SSE when equipped with EME that are mainly comparable to conventional TSE that is, with kneading blocks as mixing sections, although not as good as TSEs equipped with EMEs. Mechanical results also show enhanced modulus when EME is used in SSE operations.     

Reactive processing of LLDPEs in counterrotating nonintermeshing twin-screw extruder. III. Methods of peroxide addition

An ethylene–octene linear low-density polyethylene (LLDPE) was treated with peroxide in a reactive extrusion system. A counterrotating nonintermeshing twin-screw extruder (System 2) was contrasted with a corotating intermeshing twin-screw machine (System 1). In System 2, the peroxide solution was pumped into the melted polymer, while it entered with the polymer pellets in the feed section of System 1. Molecular structure changes and the rheological behavior of peroxide-modified resins are similar in both operations but System 2 is much more effective. Much lower peroxide levels were needed in System 2. However, reactions in this setup were also more difficult to control. The presence of microgel was clearly evident in System 2 products but not in those made in System 1. The results of such reactive extrusion processes depend critically on the method of the peroxide feed and mixing conditions. Reaction conditions that favor optimum economy and peroxide efficiency are those which may compromise product homogeneity. © 1996 John Wiley & Sons, Inc.     

Modeling melt conveying and power consumption of co-rotating twin-screw extruder kneading blocks: Part B. Prediction models

Intermeshing co-rotating twin-screw extruders are very versatile because their screw configurations can be tailored both to the application and to the properties of the materials used. Finding the best screw configuration is one of the main purposes of twin-screw extrusion modeling, and requires models that accurately predict conveying and power consumption. The better the process can be predicted, the better the requirements of the final product can be met. We present novel prediction models of the conveying and power-consumption behaviors of intermeshing co-rotating twin-screw extruder kneading blocks for Newtonian fluids. These are based on numerical simulations and therefore consider the complex three dimensional (3D) geometry of this element type without the need for common simplifications. Our models are thus capable of including all leakage flows and gap influences, which are usually ignored, for example, by the flat-plate model. Since our models are derived by symbolic regression based on genetic programming, they consist of algebraic functions and are low-threshold. They can be used to calculate various process parameters for individual kneading blocks or entire screw configurations, as illustrated by a use case.     

A study on polymer blending microrheology: Part IV. The influence of coalescence on blend morphology origination

The influence of shear induced coalescence on the origination of morphologies in polymer blending processes is investigated both theoretically and experimentally. In the theoretical part a route is proposed to estimate the fraction of collisions between disperse phase domains in simple shear flow that result in an actual coalescence. It was shown that under polymer blending conditions this “coalescence probability” is only substantial if the polymer/polymer interfaces exhibit a high degree of mobility. In the experimental part, the phenomenon of gravity induced droplet/planar interface coalescence is utilized to show the high degree of mobility of molten polymer interfaces. Seoul experiments on the relation between domain size and disperse phase concentration in polymer blends prepared on a single screw extruder were carried out. For extremely low concentration (<½ %) the domain size could be predicted satisfactorily by means of Taylor's classical criterion for Newtonian liquids, while at higher concentration coalescence increased the average domain size manifold.     

Reactive processing of LLDPEs in corotating intermeshing twin-screw extruder. I. Effect of peroxide treatment on polymer molecular structure

Commercial ethylene-octene linear low-density polyethylene (LLDPE) polymers were reactively extruded with low levels of 2,5-dimethyl-2,5 di(t-butylperoxy)hexane to modify their molecular structure and processing properties. Peroxide levels were kept low to avoid crosslinking. This article examines the effects of reactive extrusion in a corotating intermeshing extruder. Gel content analyses and examination of extruded thin tapes indicated that the products were gel-free, but line-broadening in high-resolution ¹³C-NMR spectra suggested that some crosslinking did occur. Molecular weight distributions were broadened toward higher molecular weights, as expected. SEC estimates of long-chain branching in reacted polyethylenes were consistent with the results of ¹³C-NMR analyses. Under our extrusion conditions, the products contained about one long branch per number-average molecule. This result and data on changes in carbon-carbon unsaturation indicate that the major chain extension mechanism is an end-linking reaction between terminal vinyls or allylic radicals formed at chain ends and secondary radicals. Both types are produced by hydrogen abstraction on the LLDPE. All long branches originated at tertiary branch points. Changes in thermal behavior, as measured by DSC analyses, paralleled those observed by temperature-rising elution fractionation (TREF). SEC molecular weight measurements and long-branch determinations by SEC and ¹³C-NMR can be used to quantify the effects of peroxide treatment on the molecular structure of polyethylenes. DSC and TREF techniques, however, appear to be more sensitive than are SEC or NMR. Relatively minor variations in the degree of mixing and temperature control during reactive extrusion have noticeable effects on the molecular structures of the peroxide-treated LLDPEs. © 1995 John Wiley & Sons, Inc.     

Phase separation kinetics and morphology in a polymer blend with diblock copolymer additive

The phase diagram for a low molecular weight blend of deuterated polystyrene (PSD) and polybutadiene (PB) was determined by temperature jump light scattering (TJLS) measurements and phase contrast optical microscopy (PCOM). The PSD/PB blend exhibited upper critical solution temperature behavior, and the critical temperatures measured by these two techniques were consistent. Upon addition of 0 to 0.12 mass fraction of a comparable molecular weight PSD-PB symmetric diblock copolymer, a linear decrease in the phase transition temperature was observed with increasing diblock copolymer content. At a constant, shallow quench depth, the kinetics of phase separation via spinodal decomposition as measured by TJLS were greatly retarded by the presence of the copolymer. Additionally, the time dependence of the concentration fluctuation growth did not seem to follow a universal functional form anywhere in the accessible q range when the diblock was present. The results of morphology study of the blends in the late stage of phase separation by PCOM also indicated that the phase separated domain sizes did not grow to the same size for a given annealing time as diblock content increased.     

Temperature dependence of surface composition and morphology in polymer blend film

Thin films of poly(methyl methacrylate) (PMMA) and poly(styrene-ran-acrylonitrile) (SAN) blend can phase separate upon heating to above its critical temperature. Temperature dependence of the surface composition and morphology in the blend thin film upon thermal treatment was studied using in situ X-ray photoelectron spectroscopy (XPS) and in situ atomic force microscopy (AFM). It was found that in addition to phase separation, the blend component preferentially diffused to and aggregated at the surface of the blend film, leading to the variation of surface composition with temperature. At 185 °C, above the critical temperature, the amounts of PMMA and SAN phases were comparable. At lower temperatures PMMA migrated to the surface, leading to a much higher PMMA surface content than in the bulk. The migration and preferential segregation of a blend component in thin films demonstrated here are responsible for the great difference between in situ and ex situ experimental (not real quenching or annealing) results of polymer blend films, and help explain the slow kinetics of surface phase separation at early stage for blend thin films reported in literature. This is significant for the control of surface properties of polymer materials.     

Polymer blend de‐mixing and morphology development of immiscible polymer blends during tube flow

This work is an investigation of morphology and de‐mixing of polymer blends during melt flow through a tube. Morphology is the relative size, shape and location of each distinguishable phase present in a polymer blend. De‐mixing is the shear‐induced migration of different types of polymers away from each other during the flow. The ability to tailor de‐mixing during extrusion can potentially result in a new family of plastics waste recycling processes with mixed waste entering an extruder and separate streams of different polymer types leaving it. Also, control of morphology development can lead to the formation of layered structures without the need for two or more extruders and co‐extrusion. This work is directed at elucidating morphology development and de‐mixing of polymer blends in the most simple process design: melt flow through a tube. Shear‐induced migration was quantitatively shown in various polyethylene‐polypropylene, polypropylene‐nylon 6 and polyethylene‐nylon 6 blends. The migration observed was in accord with the hypothesis that the system tends to minimize its rate of energy dissipation for a fixed flow rate. The ratio of the viscosity of the dispersed phased to that of the continuous phase greatly influenced the morphology of polypropylene‐nylon 6 and polyethylenenylon 6 blends: a droplet‐dispersed phase structure occurred at a high viscosity ratio, whereas a multi‐layer structure resulted at viscosity ratios near unity. Shear‐induced deformation and coalescence contributed to formation of the multi‐layer structure.