In-line optical techniques to characterize the polymer extrusion.
Significant advances have been made in the field of polymer melt characterization in the last two decades. The use of the real time techniques, to probe both pure polymers and compounds during the processing, have attracted attention of many researches group due to the possibility of retrofitting in-process variables and gives the opportunity to tailor-made the material properties (1-4), (5-13), (23), (30). Hence, they render important information as regard to the variations induced in the process not sufficiently taken into account by the classical procedure of only controlling temperature and pressure (3). Various techniques have been proposed to in real-time characterize the melt behavior, spanning from spectroscopic methods (4) as well as those based on classical mechanical methods (5). Such developments focus on the understanding of in situ behavior of polymer systems undergoing a processing history, avoiding unwanted artifacts added by the so-called off-line characterization, such as sampling artifacts, specimen preparation, thermal cycles, mechanical relaxation, etc. On top of that off-line methods are time-consuming, and cannot be directly applied for retro feeding the process (1-4).
_ This article deals with the use of in-line optical detectors based on visible-light stimulus and procedures developed in our research group [6-14, 23, 301 to monitor the extrusion of multiphase polymeric systems. When studying morphology development of polymer mixtures the flow type plays a critical role in determining whether or not drop breakup is possible. For instance, in simple shear flow, it was found that, since the critical capillary number is reached, viscosity ratios in range of 0.1-1, promotes the most effective particle breakup (15-18). In contrast, viscosity ratios >4 lead drops of dispersed phase to keep a spherical shape and hence drop breakup is suppressed, no matter how high is the shear rate. On the other hand, the deformation of a disperse phase is retarded by a large interfacial tension, high dispersed phase viscosity, and a small droplet size. Besides, high elasticity of dispersed phase increases their resistance to deformation and, consequently, it hampers the drops breakup (18). Typically, tine particles distribution is required for fundamental improvement of material properties. Nevertheless, particles coalescence takes place along the mixing process, what can be avoided by the use of compatibilizers or low concentration of disperse phase, i.e., <1 w/w% (19), (20).
Yang et al. (21) studied the flow influence on particle evolution in dilute mixtures of poly(dimethylsiloxane) in poly(isobutylene) in a rheo-optical apparatus by the use of dichroism and light-scattering techniques. The latter one confirmed the morphology changes as found by dichroism along the experiment, i.e., droplet deformation, droplet retraction, fibril break-up and coalescence. Migler et al. (21) found coexistence between large aspect ratio strings and ellipsoidal droplets mildly deformed in the slit-die land for PE/PS mixture by visible-light microscopy visualizations; they also showed the depth dependence of the morphology. Recently, Zborowski and Canevarolo (23) proposed an in-line procedure to probe the deformation/recovery phenomena of disperse phase droplets in PP/PS dilute mixtures by turbidity in a slit-die attached to a twin-screw extruder.
As regard to the assessment of chain orientation and spatial distribution of stresses in polymer melts, flow-birefringence technique was investigated in the group led by Prof. Janeschitz-Kriegl and recognized it as a good complement to mechanical measurements (24-28). First the extinction angle was measured and, with the help of a compensator the optical path difference--OPD was determined and then the flow birefringence calculated. Although, good results were obtained by the use of a well-known optical peace with no need for calibration, experiments were time-consuming. Barone and Wang (29) adapted a slit-die into a capillary rheometer to study the occurrence of shark-skin in 1,4 polybutadiene measuring the flow-birefringence. They correlated the oscillating intensity at the exit of the slit-die to the sharkskin periodicity. They also recognized the sinusoidal dependence of the transmitted light intensity measured as a function of the degree of molecular orientation, set by the multiple order of the retardance phenomenon. Soares et al. (30) have lately developed and applied a rheo-polarimeter to in-line probe flow-birefringence in pressure-driven flow through the slit-die land (1-3 plane) fitted to a twin-screw extruder. Differently from literature, they quoted a significant effect of the optical glass strain birefringence in the total signal response, developing an in-line procedure to discount its contribution. A direct relation between the detector signal and polymer chain orientation was found, taken into account the different orders of retardance attained.
Following the increasing interest in real time measurements as shown above, here are presented some applications of in-line optical detectors and procedures based in our experience over the last 15 years, showing the potential of using turbidity, small-angle laser light scattering and rheo-polarimetry techniques to probe in real time the extrusion process.
Commercial extrusion resins were used for preparation of polymers mixtures and in-line monitoring. Table 1 shows the respective resins and its mixtures employed to demonstrate the devices application. In addition, the melt flow index for each resin and the temperature range applied for the assessment of melt viscosity ratios are depicted.
TABLE 1. Commercial resins employed, melt flow index and temperature range for rheological measurements for all mixtures. Resin/supplier MFI Mixtures Temperature (g/10 min) (a) ([degrees]C) High density polyethylene-- 0.33 HDPE/PA66 280 Braskem Polyamide 66--Polyform -- Polypropylene--Braskern 10 PP/PSI 220, 240, 260 Polystyrene--BASF 1.5 Polystyrene--Innova 4 PS/PMMA 240, 260, 280 Poly(methylmethacrylate) (b)-- 16.4 Unigel (a) ASTM 1238-95 / ISO-1133. (b) Free radical-based poly(methylmethacrylate).
The resins were characterized in a capillary Rheometer Instron model 4467 with a capillary of high length-to-diameter ratio LID = 34, hence there is no need for Bagley correction, in the range of 30-15,000 [s.sup.-1], at temperatures shown in Table 1. The melt shear viscosities are presented as a function of the corrected shear rate, applying the Rabinowitch correction, as shown in Figs. la, c and e, respectively. Calculation of consistency and power law indices allowed the estimation of viscosity ratios.
Set-up and Detection
System Figure 2 shows the schematic set-up used to carry out the in-line optical measurement during the extrusion. The optical detectors were coupled onto a slit-die fitted at the exit of a modular intermeshing corotating twin-screw extruder Werner-Pfleiderer ZSK30 with screw-diameter D = 30.70 mm and LID = 35. It was built with the following parts: slit-die with rectangular slit having dimensions of 41 x 15 X 1.5 [mm.sup.3] (sufficiently large width-to-height ratio) and a 20 mm diameter bore fitted with transparent glass windows to allow for visible light to pass orthogonally to the molten flow direction, water cooled jacket surrounding a stabilized visible light source (20-W dichroic bulb operating at 5 V, emitting in the 350-700 nm range), a second water cooled jacket surrounding the photodetector (LDR type CdS with 11 mm diameter from Hamamatsu Corporation, with maximum absorbance at 510 nm), USB data acquisition device NI-DAQ 6812 from National Instruments (AD signal converter) and a laptop computer with LabView 8.6 software for data collection, real-time calculation, screen presentation and data saving. The water-cooled jackets also possess slots in their bases for fitting optical pieces such as polarizers and monochromatic filters (Fig. 2b). This arrangement is fully explained in previous publications (6-14), (23), (30). Figure 2b shows schematically the design of the light-scattering device developed in our group. A water-cooled cone with a conical angle of ~22.5[degrees] is attached onto the slit-die and a laser beam, typically helium-neon ([lambda] = 632.8 nm), replaces the dichroic lamp. Again, radiation crosses the melt flow perpendicularly. A circular screen possessing a pinhole in its center is coupled to the large base of the cone. Here, an arrangement of 64 photocells well spatially distributed or a Webcam can be attached to screen and/or follow the transmitted light intensity changes and the scattering patterns formed. This light-tight assemble avoids external light to interfere in the experiments. As particles of the disperse phase crosses the incident laser light is scattered by the different phase domains in the melt, being the scattering cone captured by a diffusive screen. A detailed description can be found elsewhere (13).
The software designed allows the monitoring of screw rotation, screw torque, pressure at the slit-die entrance and transmitted light intensity, all measured in real time. In addition, the screw rotation control can be switched over between the original extruder control panel and the external computer running a real-time control software by simply enabling a hard key implemented on the data collector of the extruder. The entire data flow transfer is highlighted by the arrows shown in Fig. 2a. Finally, an on-off valve, operated by a lever with two states: opened/ closed, i.e., allowing flow/no flow of the extruder output was mounted ahead the slit-die during some experimental procedures with two main objectives: first it can be set closed to assess the optical influence of the glass windows on the base-line of the detector's signal for the rheo-polarimetry measurements (30); still having it set closed leads the extruder to "mimic" a torque-rheometer. Second, when it is set opened the optical monitoring of the processed polymeric system can be taken in real time during the discharge of the material (23). A detail of the on-off valve is shown in Fig. 2c.
In-line Turbidity. The turbidity of a multiphase polymeric system is dependent upon the type, volume fraction, particle size and shape of the disperse phase, i.e., increasing with increasing the number of particles. It also increases when, in a population of particles with different sizes, the concentration of particles whose sizes are close to the visible light wavelength range increases, as found in previous reports (6-12), (14). Because of the appearance of multiple scattering effects at large number of particles the detection is limited to low volume concentrations of disperse phase, up to 5% w/w, under which the linearity of the detector signal is obtained, as well studied in previous experimental bench validation. The validation method developed consisted of preparing solutions containing titanium dioxide particles dispersed in distilled water with known particle sizes (and narrow particle size distribution) and further continuous addition to distilled water circulating through a Teflon based slit-die 1mm thick, while data were collected in real time after the signal reached stability (6), (7), (14).
When light is normally incident upon a sample of thickness x (optical path), the transmittance or attenuation factor T, which is the ratio of transmitted intensity, 1, to incident intensity, [I.sub.0], is given by Eq. I,
I/[I.sub.0] = T = exp (-[[alpha].sub.s]x)
being a, the turbidity. Assuming the relation between the signal of the optical detector in the form of normalized voltage [V.sub.N] and the attenuation factor expressed by T = 1 [V.sub.N] (7-12), then one can obtain Eq. 2.
In (I/[I.sub.0]) = -In(1/1 - [V.sub.N]) = -In ([
[V.sub.s] - [V.sub.2]]/[[V.sub.s] - V]) = -[[alpha].sub.s]x (2)
being [V.sub.N] the normalized signal, Vs the signal for total extinction of light, 170 the baseline signal from the pure polymer matrix, and V the sample signal value. The optical path (x) of the detector is 0.15 cm (slit thickness), and if this value is inserted in Eq. 3 we can calculate the turbidity ([[alpha].sub.s]) by Eq. 4.
In log (I/[I.sub.0]) = -log(1/[1 - [V.sub.N]])/log(e) = - [[alpha].sub.s]X (3)
[[alpha].sub.s] = 15.38 log (1/[1-[V.sub.N]]) (4)
To start we do show the use of the in-line turbidity to monitor in real time the morphology of a polyethylene/ polyamide 66 (PE/PA66) mixture during processing in the transient state. The polymers were processed in a twin-screw extruder Werner and Pfleiderer ZSK-30 (see Fig. 2a). The process conditions were: screw rotation of 70 rpm, feeding rate of 2.0 kg [h.sup.1] and the temperature profile kept constant at 280[degrees]C in the whole barrel. Moreover, the valve was coupled ahead the slit-die and set opened. Nearly 0.5 w/w% of PA66 was added in HDPE and the mixture done in the steady-state extrusion processing having the melt turbidity being measured simultaneously. To better disperse the polyamide 66 second phase in the polyethylene matrix the mixture was extruded twice, as shown in Fig. 3, where the turbidity intensity levels off after the second run. Knowing that, the mixture from the second extrusion was used in the controlled die-pressure processing mode to assess the second phase deformation and recovery phenomena.
Initially, with the extruder empty, 100 g of the PS/ PA66 pre-extruded polymer mixture was manually added in the main hopper and pumped downstream up to the extruder die, set by the raise in the die-pressure gauge. At this point, the processing started while the valve remained closed for 9 min: in the first minute, the pressure was linearly increased from 0 to 2.07 MPa; in the second minute the pressure was linearly reduced back to 0; and during the third minute the die-pressure remained zero (see the continuous thin line in Fig. 4).
After cycling for nine minutes, the valve was opened and the discharge of the polymer mixture started, based on a methodology previously devised (23): the discharge die-head pressure is set by the software running a PID controller to vary cyclically in a ramp profile from zero to 2.07 MPa. While the molten polymer mixture is discharged, the response from the optical detector was recorded, corresponding to the turbidity of the molten flow, which is dependent on the size and shape of the second phase droplets. The die-head pressure profile was designed having a 1-min period of rest before ramping up, making a discharge pressure cycle of 3 min. Just one complete cycle was sufficient to have all molten polymer mixture out of the extruder and the turbidity response measured. Thus one has a controlled dwell time of material, what is of extreme importance to the process optimization.
Laser Light Scattering Another important technique widely used to understand the role of the flow field on the final morphology of immiscible polymer blends is low-angle laser light scattering. In turn, it can provide fast information concerning statistical evaluation of shape and size of scattering particles which do not have a well-defined geometry during the deformation processes in melt state. It also provides real-space images of objects down to sizes on the order of 100 nm, i.e., extremely useful at short-length scales (4), (13), (21), (33). The method lies on the fact that the scattering angle is inversely proportional to particle size, where for small particles the scattering is higher at high angles in contrast to the larger ones. In addition, the scattering effect is higher when heading toward small angles, indicating distance dependence (33). Some investigators (21), (22) have applied light scattering principles to study viscoelastic fluids following previous studies on micro-rheology of immiscible fluids (15-17).
Despite the features just mentioned, the low-angle laser light scattering technique was used here solely to qualitatively monitor the morphology evolution of a polypropylene/polystyrene mixture while extruded in the transient state. The cone was attached onto the slit-die and the lamp was replaced by the laser beam. By simply adapting a common WebCam camera into the screen center pinhole, it was possible to visualize the scattering effect in the transient state. Moreover, the turbidity was used separately but at the same processing conditions to follow residence time distribution (RTD) profile along the extrusion process. This can lessen the lack of understanding on the influence of the inherent geometric complexity of screw elements on the flow during extrusion. A detailed analysis of this curve can give information on the level of the axial mixture, initial, average and final residence times, presence of stagnation zones, etc (34).
Process conditions were: screw rotation of 75 rpm, feeding rate of 2 kg [h.sup.-1] and the temperature profile kept constant at 240[degrees]C in all barrel sections. Here, the experiments were carried out without the on-off valve depicted in Fig. 2a. During the extrusion the polypropylene was continuously fed through the main hopper (port 1) at a throughput of 2 kg [h.sup.-l]. With the polypropylene matrix flowing at steady-state pulses-like stimulus of 0.05 g ("tracer") of polystyrene--PS 1 were added through the degassing zone (port 3) synchronously with the monitoring start-up. This allows for the precise experimental time counting to locate morphological change images in the residence time distribution RTD curve (see Fig. 5).
Concerning to molten polymers, upon applying a pressure-driven flow, the imposed stress orients the macromolecules along the flow direction giving rise to optical anisotropy. This causes the ordinary and extraordinary rays of incident polarized light that transverses the birefringent medium to move with different velocities (double refraction) and so out of phase. Upon exiting the medium these rays merge together but with phase retardance, due to the optical path difference (OPD) encountered along the different courses traveled. The OPD can be measured by simply adapting a pair of crossed polarizers in which the polymeric flow, set in-between, bisects this right-angle (24), (28), (30), (32). In this arrangement the dependence of cross-polarized transmitted light intensity (in terms of normalized voltage, VN) (24), (29), (30), (32) to optical path difference follows a sine-squared relation, as shown by Eq. 5,
[V.sub.N] = [sin.sup.2] ([pi]/[lambda] OPD) (5)
For the sake of convenience Eq. 5 was rearranged into Eq. 6,
OPD = [lambda]/[pi] arcsin [([V.sub.N]).sup.1/2[ (6)
being [lambda] the wavelength of light used, i.e., 546 nm for green light. The optical path difference OPD is related to the birefringence following Eq. 7,
OPD = d[DELTA]n (7)
being d the thickness of the medium traversed by the light beam. Measuring the cross-polarized transmitted light intensity in real time one is able to quantify the flow behavior in terms of both OPD and birefringence levels.
To exemplify the use of the rheo-polarimeter recently developed, the influence of a second phase of polymethyl-methacrylate on both molecular orientation of a polystyrene matrix and the mixture turbidity was evaluated. The detector was previously validated at the bench using slabs from commercial extruded polycarbonates sheets. The slabs were heat-treated in silicon oil bath just below the glass transition temperature, within specifics treatment times, to partially relax the orientation and so obtain an array of optical path difference values. Their OPD values were quantified via the conventional method using cross-polarized visible-light microscopy measurements with the help of a Bereck-compensator (30). The slabs with known OPD were inserted into the slit-die and the expected sine-square transmitted light intensity dependence (Eq. 6) checked. The original polymethylmethacrylate sample being in powder form was initially melt extruded and palletized according to the process conditions shown in Table 2. Further, both resins PMMA and PS in pellets were manually tumbled and gravimetrically fed into the extruder using K-Tron gravimetric feeder for the preparation of diluted blends ranging from 0.15 to 1 w/w% (to avoid droplets collisions and coalescence (18), (19) of pol-ymethylmethacrylate in the polystyrene matrix. The screw speed, temperature and throughput profiles are shown in Table 2.
TABLE 2. Process parameters for the preparation of PS/PMMA polymer mixtures and their in-line monitoring. Procedure Temperature Screw Feeding profile rotation rate ([degrees]C) (kg [h. sup.-1]) Mixtures 200, 220, 220, 220, 220, 220 65 3 preparation (a) Monitoring (b) 200, 250, 250, 250, 250, 250 90 25 (c) (a) Carried out at flow steady-state. (b) With stop of feed. (c) Extruder working fully filled.
Initially the strain birefringence of the transparent glass windows need to be assessed to latterly be discounted from the total crosspolarized transmitted light intensity. This was carried out by adapting the valve ahead the slit-die, setting it closed and automatically pumping downstream a pre-weighted amount of polymer matrix (PS) up to the slit-die. Upon cycling the pressure up and down while recording the crosspolarized transmitted light intensity, the birefringence of the glass windows was quantified as a function of the melt pressure, done in the expected working range. A complete description of the experimental procedure can be found elsewhere (30). A sine-squared curve was fitted to the crosspolarized transmitted light intensity, according to Eq. 8,
I = [I.sub.0] + [Asin.sup.2] [[pi] ([P - [P.sub.0]/W)] (8)
being I the rheo-polarimeter crosspolarized transmitted light intensity, [I.sub.0] the crosspolarized transmitted intensity at zero base-line, A a constant, P the glass window pressure, [P.sub.0] the pressure shift, and W the wavelength of the sinusoidal curve, given in MPa. Knowing these coefficients the contribution of the strain birefringence of the glass windows can be in-line discounted latterly during the rheo-polarimetry measurements.
The "Stop Feeding" Processing Method
In-line monitoring was carried out while the "stop feeding" extrusion method is set for both pure polystyrene and PS/PMMA diluted blends. The valve was removed from the slit-die exit and the extrusion done as follows: with the extruder parameters set as shown in Table 2, the polymer mixture was fed gravimetrically at constant throughput of 25 kg [h.sup.-1] (extruder screw working fully filled), while the rheo-polarimeter is switched on. After 40 s running, the feed was ceased ("stop feeding") and the extruder, still running, emptied itself. The rheo-polarimeter data is recorded as a function of the pressure drop given by the pressure-gauge at the slit-die entrance. As time passes, a continuously reducing amount of molten polymer remains inside the extruder, dropping the pressure and so the shear rate at the slit-die land, conveniently granting optical measurements at a continuous dropping degree of molecular orientation in a single run. Steady-state experiments were avoided because is time-consuming and employs great amounts of material, once we are using a semi-industrial extruder. However, one would expect the same trend as shown in Fig. 6. In addition, in-line rheological measurements could be made and correlated to the in-line optical information, which will be the subject of a future paper.
RESULTS AND DISCUSSION
Figure 4 shows the pressure ramp cycling experiment and the turbidity readings carried out to follow the morphology evolution of polyamide 66 dispersed phase in polyethylene matrix while being discharged from the extruder.
It turned out that, during the first stage, in which there is no flow the turbidity varied but as the valve is set open (second stage), the turbidity followed the same pattern but with increased intensity. This optical effect reflects changes in the cross section size of the scattering particles due to their deformation. As the particles deformation is depth dependent in the slit-die land, i.e., it is affected by the local shear rate, with increasing die-pressure, some droplets in the vicinity to the die-wall deform into fibrils, while others retain a slender ellipsoidal shape. This happens provided the material characteristics for the blends are met, as for instance, the viscosity ratio value for Polyethylene/Polyamide 66 blends is in a very favorable range to produce fibrils (Fig. lb). Particles which are down to half of the slit-die height may only deform into ellipsoids or ultimately retain their spherical shape. With the interplay of fibril shape and slender ellipsoids formation by the influence of the imposed flow field, the dispersed phase particles reduce their scattering cross section considerably, which reduces the interaction with the visible-light and so reducing the turbidity. We reckon that the decay in the signal intensity in the interval of 9-10 min, while there is the discharge and the die-pressure is linearly increased, reflects this behavior. Upon the linear reduction of the die-pressure in the interval 10-11 min, the turbidity increases returning to its original level, indicating that the deformation of the droplets is reversible. recovering their original size, and so their cross-section, upon ceasing the shear.
In-line Low-Angle Laser Light Scattering
Working in the transient state, pulses of polystyrene were added to the main hopper into a steady state flow of the polypropylene matrix. After undergoing the whole extrusion process history, particles of second-phase upon reaching the optical detector scatter light and consequently increase the turbidity. This is well displayed in Fig. 5 in the form of a residence time distribution RTD curve, which gives the concentration frequency of pulse-like disperse phase ("tracer") in terms of optical response as function of time (6-9).
When repeating the entire procedure above, but now replacing the turbidity detector by the in-line low-angle laser light scattering detector, it was possible to monitor the morphological changes imposed by the flow-field as shown by the scattering pattern at some particular time intervals along the RTD curve (Fig. 5). At a throughput of 2 kg h I of pure matrix, the scattering only exhibits a circular pattern up to the initial time of the RTD curve. On the other hand, when the scatters (particles of polystyrene) reach the slit-die land and so the in-line LALLS detector, one can see a superposition of a streak (perpendicular to the flow direction I) and the circular pattern. This combination pattern emphasizes the coexistence of elongated ellipsoids/fibrils ("streak") and droplets domains ("circular pattern") in the melting bulk. As mentioned before, the flow profile in the slit-die land is depth-dependent and, accordingly, due to the fact that the laser beam is radiated perpendicular to the 1-3 plane, the light scattering pattern shows a superposition of morphologies resulting from different shear conditions. In the maximum peak of the RTD curve the streak is enhanced due to the greater number of scattering elongated particles of the second phase. The two-dimensional scattering profiles follow exactly the smooth decrease of the intensity in the tail of the RTD curve. In addition, as seen in Fig. Id, the melt viscosity ratio used is in the range that propitiates the formation of such geometries, visualized by the in-line optical devices.
Figure 6 shows the in-line cross-polarized transmitted light intensity as a function of the die-entrance pressure for a flow of pure PS as well in blends with PMMA up to 1w/w% obtained applying the "stop feeding" processing method. Also, a sine-squared fitting curve due to the glass windows strain birefringence contribution is seen. As quoted before, the shear rate at the slit-die land decreases once the gravimetric feeding of the extrusion is ceased. We drown your attention to the fact that the real experiment happens reading the plot from right to left, i.e., starting from a higher pressure, in which there is a higher degree of molecular orientation, dropping the zero-pressure when the extruder has emptied itself.
Starting from the higher pressure level of 1 MPa, the optical signal goes through a maximum and then reduces approaching the glass windows base line, following the expected sine-squared behavior. Both pure polystyrene and blends do show the same trend, starting over the half of the optical path difference first order (275 nm < OPD < 550 nm). For any particular polymer flowing system, different processing conditions that produce the same OPD value have uncoil the polymer chains stretching them to the same level of orientation and so having the same relaxation time. For convenience we chose the peak's maximum in Fig. 6, which defines precisely the half of the first order corresponding to an OPD of 275 nm. Thus, all processing conditions that have produced a maximum in the crosspolarized transmitted light intensity have taken the polymer coils to a level of deformation which has the same relaxation time and so denoting an iso-relaxation time curve. Consequently, following the peaks' maxima in Fig. 6 reveals a vector a named here as "Azimuthal influence of the concentration."
The vector a can be decomposed into the x axis [[alpha].sub.x] the component, which gives the shift along the pressure axis revealing that the increase of the poly(methyl methacrylate) content in the mixture reduces the ability of the polystyrene matrix to orient its molecular chains, i.e., a drop in the PS relaxation time is evidenced. As long as the relaxation time is defined as the ratio between the viscosity of the system by its elastic modulus a reduction in it mostly indicates a reduction in the viscosity of the molten polymer (assuming that its elastic modulus is kept constant because the temperature is also kept constant). Some works (35), (36) have suggested that the drop in viscosity found when mixing two incompatible polymers is likely due to an apparent interfacial slip between the phases. According to the literature (36), (37), systems containing high incompatible, i.e., chemically dissimilar components, possess a low degree of entanglements among polymeric chains at the phases' interface. This fact might explain the shifts in the curves of Fig. 6 as the disperse phase amount increases, increasing the number of interfaces and so its influence.
The decomposing of the vector a into the y axis, the [[alpha.subu.y]] component reveals the influence of the concentration of the second phase on the crosspolarized transmitted light intensity. Figure 7 shows a linear decrease in the crosspo-larized transmitted light intensity in the concentration range analyzed, i.e., increasing poly(methyl methacrylate) content reduced considerably the optical response due to the turbidity generated by the increasing number of scattering PMMA particles. Moreover, as the viscosity ratio of the mixture components (Fig. If) was set to be in the range prone for particle break-up, this has produced particles with size close to the wavelength of light used ([lambda] = 546 nm for green light) accounting for the scattering effects.
Normalizing the crosspolarized transmitted light intensities in Fig. 6 using Eq. 2, dividing them by the corresponding maximum's peak value (i.e., normalizing by the concentration, at the same degree of orientation) of each curve and substituting the VN values in Eq. 6, one is able to plot the molecular orientation of the matrix polymer in terms of its flow optical path difference OPD as a function of the melt die-entrance pressure. This procedure also eliminates the effect of the turbidity. This is shown in Fig. 8 presenting the level of chain orientation of pure PS and its blend with 1 w/w % of PMMA, in which the glass windows strain birefringence was deducted.
It turned out that for a polystyrene melt flow pressurized to 1.1 MPa the presence of only 1 w/w % of PMMA dropped the orientation level in terms of OPD of the PS matrix in 20%. Applying Eq. 7 and knowing that the slit-die thickness is 0.15 cm we can get the flow birefringence values in the order of -2.8 X [10.sup.-4] for pure polystyrene and -2.2 X [10.sup.-4] for 1 w/w% PS/PMMA blends.
Development of real-time characterization techniques is very attractive for they provide in situ information upon the material. In turn, it accounts for fast decision-making no matter the time of running so that the operator is able to retrofit process parameters and/or compounding to better define processing conditions. In this report we discussed recent developments and application of in-line optical-based techniques probing the polymer extrusion. Visible-light based techniques has been of concern in our research team; as a result of matter and radiation interaction, changes in the latter provide information concerning to mixing efficiency, particle size and shape and deformation of second phase, such as evaluated by turbidity and low-angle light scattering techniques. In addition, the rheo-polarimetry technique provides a unique opportunity of monitoring polymeric chains orientation in both pure polymers as in their mixtures, as for instance probing the influence of a second phase (polymer, lubricant ...) on the rheological character of the matrix.
The authors acknowledge all colleges for their contribution in the past 15 years, listed in the references. To Innova, Unigel, Braskem and Polyform for providing the materials.
Correspondence to: S. V. Canevarolo; e-mail: firstname.lastname@example.org
Contract grant sponsor: CAPES, CNPQ, FAPESP, Brazilian Institutions.
Published online in Wiley Online Library (wileyonlinelibrary.com).
[c] 2013 Society of Plastics Engineers
(1.) P.D. Coates, Polymer Processing: In-Process Measurements. Encyclopedia of Materials: Science and Technology, Pergamon Press, New York. Chapter 46 (2001).
(2.) P.D. Coates, A.L. Kelly, R.M. Rose, and M. Woodhead, In-Process Entry Pressure Measurements in Extrusion and Injection Moulding, Bradford University, Bradford (1998). ISBN 1 85143 173 X.
(3.) S.E. Barnes, M.G. Sibley, H.G.M. Edwards, P.D. Coates, Trans. Inst. Meas. Control., 29, 465 (2007).
(4.) I. Alig, B. Steinhoff, and D. Lellinger, Meas. Sci. Technol., 21, 19 (2010).
(5.) J.A. Covas, J.M. Nobrega, and J.M. Maia, Polym. Test, 19, 176 (2000).
(6.) T.J.A. Melo and S.V. Canevarolo, Polym. Eng. Sci., 42, 170 (2002).
(7.) T.J.A. Melo and S.V. Canevarolo, Polym. Eng. Sci., 45, 11 (2005).
(8.) L.A. Pinheiro, C.S. Bitencourt, L.A. Pessan, and S.V. Canevarolo, Macromol. Symp., 245/246, 347 (2006).
(9.) L.A. Pinheiro, H.G. Hu, L.A. Pessan, S.V. Canevarolo, Polym. Eng. Sci., 48, 806 (2008).
(10.) S.V. Canevarolo, M.K. Bertolino, L.A. Pinheiro, V. Palermo, and S. Piccarolo, Macronwl. Symp., 279, 191 (2009).
(11.) M.K. Bertolino and S.V. Canevarolo, Polym. Eng. Sci., 50, 440 (2010).
(12.) L.A. Pinheiro, C.S. Bittencourt, and S.V. Canevarolo, Polym. Eng. Sci., 50, 826 (2010).
(13.) L.C. Costa, "Uso de LALLS in-line na extrusdo de sistemas polinzericos bifdsicos", Master of Science, Universidade Federal de Sao Carlos (2007).
(14.) K.C. Reis and S.V. Canevarolo, Polym. Eng. Sci., 52, 12 (2012).
(15.) G.I. Taylor, Proc. R. Soc. Lond. A, 146, 501 (1934).
(16.) S. Tomotika, Proc. R. Soc. Lond. A, 150, 337 (1935).
(17.) H.P. Grace, Chem. Eng. Commun., 14, 225 (1982).
(18.) J.J Elmendorp, "A Study on Polymer Blending Microrheol-ogy," PhD thesis, Technische Hogeschool Delft (1986).
(19.) U. Sundararaj and C.W. Macosko, Macromol. Symp., 28, 2647 (1995).
(20.) C.W. Macosko, Macromol. Symp., 149, 171 (2000).
(21.) H. Yang, H. Zhang, P. Moldenaers, and J. Mewis, Polymer, 39, 5737 (1998).
(22.) K.B. Migler, E.K. Hobble, and F. Qiao, Polym. Eng. Sci., 39, 2282 (1999).
(23.) L. Zborowski and S.V. Canevarolo, Polyrn. Test., 31, 260 (2012).
(24.) R.K.N. Janeschitz-Kriegl Hermann, Polymer Melt Rheology and Flow Birefringence, Springer-Verlag, Berlin (1983).
(25.) J.W.M. Noordermeera, Flow Birefringence Study of Polymer Conformation, PhD thesis. Technische Hogeschool Delft (1974).
(26.) F.H. Gortemaker, A Flow Birefringence Study of Stresses in Sheared Polymer Melts, PhD thesis. Technische Hogeschool Delft (1976).
(27.) J.A. Van Aken, Birefringence and Stress in Polymer Melts Under Shear and Extension, PhD thesis. Technische Hoge-school Delft (1981).
(28.) J.L.S. Wales, The Application of Flow Birefringence to Rheological Studies of Polymer Melts, Master of Science, Technische Hogeschool Delft (1976).
(29.) J.R. Barone and S.-Q. Wang, Rheol. Acta., 38, 404 (1999).
(30.) K. Soares, A.M. da Cunha Santos, and S.V. Canevarolo, Polynz. Test., 30, 848 (2011).
(31.) J. Gregory, Particles in Water: Properties and Processes, CRC Press, Boca Raton (2006).
(32.) G.H. Meeten, Optical Properties of Polymers, Elsevier Science Publication, London (1986).
(33.) K. Sondergaard and J. Lyngaae-Jorgensen, Rheo-physics of Multiphase Polymer Systems: Characterization by Rheo-Optical Techniques, Technomic Publishing Company, Pennsylvania (1995).
(34.) P. Cassagnau, C. Mijangos, and A. Michael, Polym. Eng. Sci., 31, 19 (1991).
(35.) C.C. Lin, Polymer, 11, 192 (1979).
(36.) R. Zhao and C. W. Macosko, J. Rheol., 46, 167 (2002).
(37.) L.A. Utracki, Polym. Eng. Sci., 23, 609 (1983).
A.M. da Cunha Santos, (1) C.A. Caceres, (1) L.S. Calixto, (2) L. Zborowski, (1) S.V. Canevarolo (2)
(1) Programa de Pos-Graduacao em Ciencia e Engenharia de Materials, PPG-CEM, Universidade Federal de Sao Carlos, UFSCar, Brazil
(2) Departamento de Engenharia de Materials, DEMa, Universidade Federal de Sao Carlos, UFSCar, Rod. Washington Luis, km 235, 13565-905, Brazil
|Printer friendly Cite/link Email Feedback|
|Author:||Santos, Cunha; Caceres, C.A.; Calixto, L.S.; Zborowski, L.; Canevarolo, S.V.|
|Publication:||Polymer Engineering and Science|
|Date:||Feb 1, 2014|
|Previous Article:||Toughening modification of Poly(vinyl chloride)/ [alpha]-methylstyrene-acrylonitrile-butadiene-styrene copolymer blends via adding chlorinated...|
|Next Article:||Effect of curing conditions on the properties of multivesiculated polyester particle dispersions.|