An inside look at extrusion melt temperatures.

For years, many processors have been operating under the misconception that extruder temperature and melt temperature are always one and the same. Yet thermal differences between the equipment walls and the center of the melt stream occur more often than not. Astute processors may already recognize this phenomenon intuitively, but most have still not made the jump to the practical benefits of knowing the true, varied thermal picture across the whole melt stream.

Chief among these benefits is the heightened product quality available to those who can better control melt temperature. If a processor experiences seemingly unsolvable extrusion quality problems, we can usually explain it in terms of the oft-ignored thermal variations across the melt stream. After all, temperature affects many characteristics of the final product, including strength, dimensional uniformity, color, and surface quality. Fortunately, this thermal non-uniformity is a function of common process configuration choices and material selection. As such, it can be minimized through process adjustments.

Why do these radial temperature gradients exist? In part, they result from the unique physical characteristics of polymers. Because they exhibit high specific heat relative to other materials, polymers require large amounts of energy for melting. Polymers also tend not to conduct heat away from themselves, working instead to store heat. Furthermore, polymers can be thought of as reactions in progress--a combination of raw materials exposed to different heat histories and to different rates of reaction during the manufacturing process. And the raw materials themselves vary day in and day out, producing viscosity inconsistencies even within the same grade. Indeed, this variation has been tacitly acknowledged by some resin suppliers, which specify their products, melt indexes with a tolerance factor.

Another factor working against truly isothermal melt streams springs from the extrusion process itself. The action of the screw lends itself to temperature gradients through localized shear heating. Moreover, fluid mechanics and boundary-layer phenomena result in polymer flows with a number of distinct regions with regard to temperature, viscosity, and flow velocity.

Together, these polymer and flow characteristics result in nonuniform temperature. The melt on the outside of the channel may equal the barrel or die-wall temperature, but regions deeper into the melt stream probably would differ. To this day, however, flush-mounted thermocouples perform the vast majority of temperature-sensing duties in extrusion. Accordingly, the prevailing attitude has been: "If I can't measure something, why worry about it anyway?"

For this very reason, melt temperature remains a relatively unexplored variable. Advances in instrumentation, however, allow us to get a better handle on melt temperature by measuring temperature and sensor position simultaneously. This way, a single melt steam can be characterized in terms of a thermal profile rather than by a single, potentially unrepresentative measurement--as provided by flushmount thermocouples in the equipment wall.

Recently, we conducted a series of experiments that examine the effects of screw speed and design, machine temperature settings, and the material's rheological properties on the melt-temperature profiles of several common resins. The results cast quantifiable doubts on the old isothermal assumption. And though these experiments focused on just a few materials--LLDPE, polycarbonate, PP, virgin HDPE, and HDPE reclaim--work presented at last month's SPE ANTEC conference in New Orleans showed the influences of these factors on melt-temperature profiles in nylons and acetal as well.

TRACKING DOWN TEMPERATURE

The first set of trials, which were conducted on a Battenfeld-Fischer FBZ1000 extrusion-blow molding machine at the Canadian National Research Council in Boucherville, Que., revealed how changes in screw rpm, heater-band settings and material choice can alter melt-temperature gradients. The materials used were a virgin HDPE blow molding grade, a 50% mixture of the virgin and regrind HDPE, and a polycarbonate. The Dynisco Autoprobe I, a motor-actuated thermocouple, provided the temperature readings and an analog indication of the sensor's radial position in the melt stream.

Temperature profiles were collected in two places: in the head near the tip of the machine's 1.2-in., 25:1 L/D screw and in the die near the extruder exit. A total of five heater zones were present, three along the screw and two on the head and die. Although temperature settings varied from run to run, all five zones remained equal during a given run.

All the trials employed a straight annular die with an 8.0-mm gap and a 35-mm I.D. Annular extrusion was chosen because it applies to a variety of high-volume processes, such as tubing, pipe and blown film, as well as blow molding itself.

Throughout this first set of trials, variations in process parameters and material were indeed reflected by different profiles. All other factors being equal, raising temperature settings reduced the magnitude of the temperature gradient across the melt stream (see Fig. 1). RPM adjustments, by contrast, played a smaller, though still noticeable, role in increasing the gradient in HDPE. The influence of material choice is shown by temperature gradients that tend to be slightly larger for regrind blends in comparison with virgin material (see Fig 2). Polycarbonate, meanwhile, showed a potential for larger gradients than HDPE (see Figs. 2 and 3) In terms of positioning inside the equipment, temperature tends to be lower at the die than at the head (Fig 3) . This difference is due to the absence of screw-induced heating, allowing melt and machine temperature to approach equilibrium.

PROFILE EXTRUSION TRIALS

The second set of experiments, performed with two polypropylene homopolymers on a Welex 2-in. single-screw extruder at the University of Massachusetts at Lowell, similarly shows how material choice can produce melt streams with entirely distinct temperature profiles. This study employed a 30:1 screw with a compression ratio of 3.10:1 and a profile die producing a 0.250-in.-diam. rod. Even with process settings held constant, extrusion of PPs that differed only in their molecular-weight distribution again revealed unique profiles (see Fig. 4). The same behavior occurred when temperature settings for the die and extruder were reversed to raise the latter's temperature.

Perhaps more important to many processors were similar trials with LLDPE. Results here showed how changes in rpm can radically alter a temperature profile. Screw-speed changes alone produced pronounced changes in the melt-temperature profiles of an LLDPE, with the flattest profile coming from relatively high rpm (Fig. 5).

A BASIS FOR ACTION

The importance of knowing the radial melt-temperature profiles associated with given process choices and materials cannot be understated, owing to the strong influence of temperature on a melt's rheological properties and thus on the final product. On a more basic level, unknown factors can too easily mask thermally induced quality problems. For instance, processing conditions can easily drive the temperature in the center of a melt stream beyond the machine settings, potentially into the range of a material's thermal degradation.

Unfortunately, there appears to be no universal prescription for flattening temperature profiles. With HDPE, for example, as screw speed increases so too does the temperature gradient. By contrast, with LLDPE, a relatively high rpm produced the flattest profile (see Figs. 1 and 5). Nonetheless, the knowledge of individual temperature profiles is--or should be--an essential part of high-quality extrusion operations. By rigorously characterizing temperature behavior across the whole melt stream, processors can then work to maximize temperature uniformity and the resulting product quality. So the more temperature information you have, the better.

Key to the Graphs

a Measured at the head of a Battenfeld-Fisher extrusion-blow molding machine fitted with a 1.2-in., 25 L/D screw. HDPE is Petromont DMDY 6200.

b Ratio of probe's position in melt stream (r) and total radial distance of melt stream (R).

c "Regrind" is 50% reclaim HDPE/50% virgin HDPE mixture. PC is Miles Makrolon HMS blow molding grade. Screw speed is 28 RPM throughout.

d "Head" measurements taken in the Battenfeld-Fischer extrusion-blow molding machine. "Die" measurements taken in a straight annular die on the same machine. Screw speed set at 28 RPM for all runs.

e Measured in a 2.0-in. Welex single-screw extruder. The screw has a 30:1 L/D and a compression ratio of 3.10:1. The rod die used was held at a constant temperature of 390 F.

f Low-viscosity PP homopolymer from Himont (MI of 45 g/10 min); medium-viscosity PP homopolymer from Himont (MI of 7.5 g/10 min)

g Measured on the Welex and rod-die line with machine temperature set at 475 F throughout. LLDPE is Dowlex 2042.

Acknowledgements

The author wishes to thank the following individuals for their valuable contributions to the research and to the article:

A. Garcia-Rejon to the Canadian National Research Council, Industrial Materials Institute.

Robert Malloy of the University of Massachusetts at Lowell, Department of Plastics.