Using rheological data in the development of coextrusion processes.
The development of successful coextrusion systems demands careful control and matching of the constituent polymers, ideally at an early stage of the development process. Poor matching may result in a complete lack of adhesion between the different layers and/or unacceptable surface quality, caused in part by instabilities at the interface. Pilot and full-scale trials are potentially a costly part of the development cycle and there is a place for less expensive tools that yield relevant information. Here we look at one such approach--a coextrusion die designed specifically for use with Rosand capillary rheometers, which can be applied to small-scale coextrusion experimentation.
CHARACTERISING POLYMERS USING CAPILLARY RHEOMETRY
Capillary rheometry is a well-established technique for the study of extrusion. Molten polymer under controlled temperature in the barrel of a capillary rheometer is forced by a piston to extrude through a narrow die of given dimensions. The rheological properties of the polymer can be calculated from knowledge of the piston drive speed and pressure drop across the die.
Shear viscosity is determined by measuring pressure drop over the die length for a known volumetric flow rate of polymer. Shear stress is calculated from pressure drop and the area of die in contact with the polymer; shear rate from volumetric flow rate and die geometry. Shear viscosity, the ratio of the two, is an important parameter for defining how the polymer will flow,
Capillary rheometers permit the characterisation of polymer behaviour across a wide range of shear rates, providing insight into common processing problems such as melt fracture. Melt fracture results in poor surface quality in a way that is complex and arguably still not fully understood. However, its onset can be established using capillary rheometry, which consequently is very valuable for determining a workable processing window.
As polymers are viscoelastic materials, extensional viscosity (also known as elongational viscosity) as well as shear viscosity should be measured for a more complete understanding of a polymer's properties. Extensional viscosity is a measure of a material's resistance to stretching forces, which occur in many polymer processing applications. Extensional viscosity can be determined from parameters that are routinely measured during a twin bore capillary rheometry test, using the Cogswell convergent flow method (1) Extensional viscosity is strongly dependent on polymer architecture, specifically molecular weight distribution. As extensional viscosity is highly differentiating, it allows rationalisation of processing differences between seemingly identical polymers, as the following example demonstrates.
Data for two different high density polyethylene samples are shown in figure 1. The samples appear very similar in terms of their melt flow index, density and shear viscosity behaviour, but are actually quite different with respect to molecular weight and polydispersity (molecular weight distribution). In fact they were produced using different commercial catalyst systems: Phillips and Ziegler-Natta. Extensional viscosity data (see figure 2 ) differentiate clearly between the two samples, lower values being observed for the sample with a narrower molecular weight distribution (Ziegler-Natta).
This analysis highlights the value of studying both shear and extensional viscosity. The extrusion behaviour of these two samples will be different even though a superficial analysis might indicate them to be essentially the same. Extensional viscosity detects the difference in molecular architecture, providing important information for optimising extrusion.
Extensional properties of polymers are also critical in co-extrusion processes--recent research by Zatloukal et al (2) using two identical polymers in a co-extrusion process has shown that interface instabilities can be generated by exceeding critical extensional rates, and that such instabilities can be predicted using rheometry and simulation.
A capillary rheometer can also be used to study die swell (the expansion of a polymer post-extrusion through a capillary die), directly with a laser die swell system or by analysing the viscoelastic material properties. Convergent flow into the die causes polymer molecules to be stretched and oriented in the die, setting up molecular tensions which can relax as the polymer exits the die.
EXPERIMENTAL INVESTIGATION OF CO-EXTRUSION
As briefly described already, rheological analysis gives critical insight into extrusion behaviour. Practical experimentation, particularly on a small scale, is also extremely valuable, providing data for process development and model validation. Recognising this, FlemingPTC has developed (GB Patent GB2396429 Measuring the elasticity of a plastics material) a unique die for detailed co-extrusion analysis. Building on the capability of a Rosand RH7 capillary rheometer from Malvern Instruments, the die exploits the design features of the rheometer to provide an integrated solution for the study of co-extrusion behaviour.
Figure 3 shows a schematic of the die, the green and yellow areas representing two dissimilar polymer streams. It is attached directly to the rheometer barrel, taking its feed from the twin bores of the instrument. A Y-shaped transition body brings the two streams together, with hydraulic pressure driving the combined flow into a rectangular channel of known dimensions. Along the channel, a series of three transducers measures pressure drop, providing data for shear stress determination. The fourth transducer is included for simultaneous elasticity work based on the Lodge hole pressure method (3), (4), (5).
Polymer flow rate through the die is controlled by the rheometer cross-head, allowing control of the shear rate. Temperature is maintained using a band heater, controlled via the rheometer. The geometry of the die can be varied using an insert system. Combining pressure drop and volumetric flow data with information about the channel geometry, shear stress and shear rate enables calculation of shear viscosity. Pressure drop data are also used to check the stability of flow through the channel, according to Hagen (6) and Poiseuille (7).
The die therefore provides a sample of extruded material for further study, an opportunity for practical experimentation, and data for simulating co-extrusion processes.
SIMULATING CO-EX TRUSION OF HDPE AND PP
Simulating the co-extrusion process using measured rheological data is an excellent route to process optimisation. The Compuplast Virtual Extrusion Laboratory (VEL) is a proprietary software package specifically designed to simulate almost all extrusion processes; coextrusion constitutes one of the main specialities. Material properties and rheological data in combination with appropriate boundary conditions are the required inputs.
The first stage in developing an extrusion model is to define material characteristics in terms of both rheological behaviour and thermal properties. Viscosity data can be described using correlations such as the Carreau WLF model (8), (9), provided that measurements are made at several (minimum three) different temperatures for regression. The goodness-of-fit of the model is obviously important, underpinning the accuracy of the overall simulation. Data for HDPE and PP are shown on page 10 in figure 4; the viscosities of the two polymers are markedly different.
The next step is to recreate the die geometry within the software, and define boundary conditions for the system. In this case the model being developed is for co-extrusion from the experimental die described above, which has clearly defined geometry. Boundary conditions are specified in terms of polymer mass flow rate (for each polymer), temperature of the die and piston speed (of the rheometer). While volumetric flow rates for the two polymers are identical, mass flows may be dissimilar because of density differences.
Solution of the formulated problem generates a velocity profile for flow of the polymers through the die. The PP, with lower viscosity, travels more rapidly than the more viscous HDPE, a highly filled product. Figure 5 illustrates the impact of this behaviour on the position of the interface between the two polymers, the grey area representing HDPE. It is clear that rather than being in the centre the interface lies well over to one side, the HDPE layer being much thicker despite both polymers being extruded at the same volumetric flow rate.
Samples provide a practical illustration of the impact of this behaviour on extrudate quality (see figure 6). Even at relatively low extrusion rates (1 mm/min) there is wavy distortion of the PP which occurs due to the lower viscosity PP having a significantly higher velocity than that of the higher viscosity HDPE. At higher rates the extrudate becomes severely twisted as the velocity differences are exacerbated; the higher velocity PP stream must fold back on itself. A measurement of interface position can be made from the untwisted sample and the result compared with simulated data. In this case the calculated position of the interface is 6.5 mm from the left hand edge of the extrudate, that is, the HDPE layer has an estimated thickness of 6.5 mm; falling exactly between the measured thickness from the extrudate of between 6 and 7 mm, which validates the model.
USING EXPERIMENTAL AND SIMULATED DATA
Close agreement between experimental and simulated data increases confidence that the model accurately represents the system, allowing exploration beyond the original experimental envelope. Furthermore, following the development of an accurate model, stresses within the system can be determined for efficient co-extrusion processes, to develop an optimal system.
Information relating to conditions at the die wall and the interface are particularly important when assessing a co-extrusion system. Parameters such as shear and extensional stress and rate, temperature and/or relative velocity all give insight into the process, particularly when there are defined values that yield optimal performance.
For example, empirical rules gleaned from both academic study and observation which apply to co-extrusion are:
* The stress at the wall should not be too low, otherwise the 'self cleaning' effect of the flow is lost and stagnation can occur.
* Similarly, wall shear stresses above 140kPa are synonymous with the initiation of sharkskin in polyolefins and should thus be avoided.
* Elongation rates at the interface must not exceed 30 s (-1).
* Interface shear stresses in excess of 68 kPa should be avoided when using polyolefins due to their influence on the restoring force leading to interfacial instabilities.
Referring back to the extrudate in Figure 6, predictions of wall shear stress show that it is below the 140 kPa critical stress at 117 kPa, and indeed no sharkskin is evident. Stress at the interface is predicted to be 20 kPa, thus also below the 68 kPa critical suggested. The distortions in this case remain wholly attributable to the differential layer velocities created by the mismatched polymer viscosities.
By interrogating the model, process performance can be assessed against this, highlighting areas that require change. The model also allows investigation of the impact of different solutions--changes in flow conditions, geometry, processing temperature--allowing optimisation without the need for further experimental work. This speeds up the development process and reduces cost.
Good rheological data allow the highly effective simulation of co-extrusion processes, providing models for system optimisation. The unique die discussed here allows the study of co-extrusion directly on a small scale, the generation of data for model validation, and the measurement of practical effects such as flow instability. In combination, these tools deliver a powerful and cost-effective approach for polymer processors seeking insight into a co-extrusion process, particularly early in the development cycle.
(1) Cogswell, FN, Converging flow of polymer melts in extrusion dies, Polymer Engineering and Science, 12, 64 - 73 (1972).
(2) Martyn, M, Spares, R, Coates, P and Zatloukal, M. Analysis of interfacial instabilities in co-extrusion flows for LDPE melts, Journal of Non-Newtonian Fluid Mechanics 156,150 - 164 (2009).
(3) Lodge, AS. An attempt to measure the first normal stress difference N1 in shear flow for a polyisobutylene/decalin solution"d2b"at shear rates up to 10(6)s(1), Journal of Rheology. 33, 821 - 841 (1989).
(4) Lodge, AS. Elastic Liquids, Academic Press, London & New York (1964).
(5) Lodge, AS. Normal stress differences from hole pressure measurements, in Collyer, AA, and Clegg, DW, Rheological Measurements, Elsevier, New York (1988).
(6) Hagen, G. Uber die bewegung des wassers in egen cylindrischen ruhren, Annual Review Of Physical Chemistry, 46, 423 - 442 (1839).
(7) Poiseuille, JL. Recherches experimentales sur le mouvement des liquides dans les tobes de tres petites diametres, Comptes Rendus, 11, 961 - 1041 (1840); 12, 112 (1841).
(8) Carreau, PJ. Transaction of the Society of Rheology, 16, 99 - 127 (1972).
(9) Williams, ML, Landel, RF and Ferry, JD. Journal of the American Chemical Society, 77, 3701 - 3707 (1955).
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||feature: testing & inspection|
|Comment:||Using rheological data in the development of coextrusion processes.(feature: testing & inspection)|
|Publication:||British Plastics & Rubber|
|Date:||Feb 1, 2010|
|Previous Article:||Now MAPRA has its own branded gauges.|
|Next Article:||SATRA offers toy testing to new US requirements.|