Continuous extrusion of microcellular polycarbonate.
Microcellular foams are defined as cellular structures having cell sizes on the order of 10 [micro]m or below, and cell densities greater than [10.sup.9] cells per cubic centimeter of unfoamed polymer. The main interest in microcellular foaming is based on the claim that mechanical properties would remain those of the unfoamed polymer, along with a significant density reduction. Many studies deal with the batch production of microcellular structures, based on the solid-state microcellular process (1, 2). This process is a two-stage process, where the polymer specimen is first charged in a pressure vessel with the physical foaming agent (PFA), usually nitrogen or carbon dioxide. This is followed by the foaming step, where the gas-charged polymer specimen is heated under atmospheric pressure to induce the gas release into microcells. This second step is traditionally accomplished in a temperature-controlled bath. The foaming step can also be accomplished through depressurization in a heated pressure vessel. While polystyrene is usually the preferred polymer choice, special interest has been also devoted to polycarbonate, a widely used engineering thermoplastic (1, 2).
Continuous production of microcellular plastics in extrusion has also been successfully attempted (3, 4). The suggested approach is based on the following sequence: formation of a polymer/gas solution, nucleation of large number of bubbles via a rapid pressure drop, suppression of cell coalescence, and expansion to the desired foam density. Although these requirements are common to a large extent to the classical low-density loam extrusion process, the key aspect in the microcellular process is the achievement of a high cell density number.
The present work will be focused on the extrusion of polycarbonate foams with either carbon dioxide or n-pentane as the foaming agent, and special attention will be given to the key processing parameters required to yield high nucleation numbers. Foaming agent concentration, processing temperatures and pressures. and influence of solubility and diffusivity on the nucleation mechanism are some of the key topics explored in this study.
Two polycarbonate (PC) resins were used for these experiments, texan 101, general-purpose grade from GE Plastics, has a melt flow index of 7.0 dg/min. The second polycarbonate, PK-2870 also from GE Plastics, is a branched resin used for blow molding, and it has a melt flow index of 2.5 dg/min. This second resin has been selected because of its enhanced foaming characteristics observed during batch microcellular foaming, as reported in Ref. 5. The use of a branched resin for solid-state foaming at 120[degrees]C and 30 sec resulted in enhanced cell population densities, with cells roughly two times smaller than those obtained with the linear grade PC. Both PCs had a density of 1200 kg/[m.sup.3]. The glass transition temperatures for the two resins were nearly identical: 149.7[degrees]C for the linear grade, 148.5[degrees]C for the branched.
Extrusion Equipment and Procedure
The foaming was carried out on a Leistritz 50 mm counter-rotating twin-screw extruder. A gear pump was placed at the extruder exit, prior to the die. The experiments described in this work were carried out with a single strand die of diameter 2.0 mm and die land 2.0 mm. The nominal flow rate for the foaming experiments was 10 kg/h. The screw speed was set al 100 rpm. The pressure at the inlet of the Near pump was maintained at 5.52 MPa. Prior to extrusion, the polycarbonate resins were dried overnight at 110[degrees]C.
The n-pentane was injected into the extruder barrel with a Shimadzu LC-8A preparative chromatography pump. A C[O.sub.2] gas regulator was used to adjust the C[O.sub.2] concentration. The screw was designed in such way to create a melt seal in the early sections. The injection port used for the PFA was located at barrel section #4. Close to the injection port, a pressure transducer was installed that read the injection pressure prevailing within the extruder barrel, as shown in Fig. 1a. This injection pressure is dependent on the PFA concentration, as well as on the barrel temperature profile, the total flow rate, the degree of fill and the extruder screw speed. The PFA mass flow rate and, incidentally, its concentration were calculated from the monitoring of mass loss of each foaming agent from their respective storage containers.
[FIGURE 1 OMITTED]
The temperature profile on the extruder was maintained constant for nearly all the trials: 250[degrees]C lot the first section. 270[degrees]C for barrel sections #2 to 5, 250[degrees]C for section #6, and 220[degrees]C for the remaining four barrel sections. The final melt temperature was adjusted through the temperature set-points of the gear pump and that of the die. As illustrated in Fig. 1b, the melt temperature, as measured using an immersion thermocouple, fairly matched the set temperature of the die. over a temperature range spanning 190[degrees]C to 280[degrees]C.
Characterization Equipment and Procedure
The dynamic viscosity was determined in a Rheometrics Mechanical Spectrometer (RMS-605), with parallel plates having a diameter D = 25 mm, at a frequency [omega] = 0.1 to 100 rad/s. The steady-state viscosity was obtained using a Rosand capillary rheometer, RCR, model RH-7. One set of capillaries with diameter of 1.0 mm was used with different capillary lengths (L/d = 0.25. 4, 10 and 20). All the capillaries have 180[degrees] entry angles. The Rabinowitsch and Bagley corrections were applied to the measurements. Moreover, along with the Bagley correction, a second-order polynomial fit was used to take into account the effect of pressure on viscosity. All viscosity measurements were performed at 270[degrees]C.
The foam density was determined by water volume displacement of a known mass of foam. The average cell size of the loam was determined by fracturing the loam at cryogenic temperatures. Scanning electron microscopy of the samples provided photographs that were then used to trace cell contours. The mean cell diameter (number-average) was computed by image analysis of the contour plots of approximately 100 cells.
The on-line rheological measurements were carried out with a Process Control Rheometer (PCR) from Rheometric Scientific. A description of the apparatus and the procedure can be found elsewhere (6). The method used for the determination of the degassing conditions using an ultrasonic technique can be found in Ref. 7. These experiments were conducted at a Iced rate of 2 kg/hr, and the barrel temperature set-points. from the gas injection zone to the instrumented slit die, were set to match that of the degassing temperature. A Rubotherm high-pressure magnet-suspension-balance was used lot the off-line determination of the solubility and dillusivity parameters (8).
RESULTS AND DISCUSSION
The viscosity of the two resins measured at 270[degrees]C. obtained from dynamic (complex viscosity, [[eta].sup.*]) and capillary (steady-state, [eta]) measurements, is illustrated in Fig. 2. Although the two resins differ in the level of their Newtonian plateaus, their viscosity responses tend to superimpose at higher rates as the fluids undergo shear-thinning behavior. The Cox-Merz rule is obeyed for the branched PC resin, as the complex viscosity curve superimposes quite well on the steady-state viscosity results. However, the curves do not overlap for the linear resin. The same discrepancy has also been observed between linear (HDPE) and branched (LDPE) grades of polyethylene (9). It was proposed that the enhanced entanglement in the branched resin extends the validity of the Cox-Merz rule in the low-deformation rate region.
[FIGURE 2 OMITTED]
Foam Density and Nucleation Density
The densities of the foam obtained with various concentrations of both physical foaming agents are shown in Fig. 3. In Fig. 4 the corresponding cell diameters are plotted. Trends exhibited by both PFAs are similar. At low foaming agent concentrations (below 1.5 wt% for C[O.sub.2] and below 5.0 wt% for n-pentane), the loam density decreases with increasing PFA content. Lowest loam densities were obtained with approximately 5 wt% of n-pentane, which has given density in the order of 60 kg/[m.sup.3]. Within these concentration ranges, the cell diameters remained quite large, close to 1 mm.
[FIGURES 3-4 OMITTED]
These responses are typical of a classical extrusion foaming process. The total volume of the cells, and hence, the reciprocal of the foam density, are proportional to the molar quantity of foaming agent dissolved in the polymer. However, nucleation remains poor under these low concentration values, and the majority of PFA molecules contribute preferentially to the growth of existing cells rather than to the creation of new cells during the nucleation and bubble growth stages.
Surprisingly, as the PFA concentration is increased to higher values, with concentrations of C[O.sub.2] set between 2.5 and 3.0 wt% and those for n-pentane between 7.0 to 8.0 wt%, the cell sizes diminish toward microscale values (below 10 [micro]m), and microcellular foams are obtained. However, in this concentration range, a reverse trend in the foam density appears, with less expansion taking place. This is especially true for carbon dioxide, in which case the foam density is reduced only by a factor of 0.9 compared to the unfoamed PC. Difficulty in generating both small cells and low foam density has been previously reported for PS/C[O.sub.2] systems (10).
The nucleation density [N.sub.d] can be calculated using:
(1) [N.sub.d] = 6/[pi][D.sup.3] ([[rho].sub.p]/[[rho].sub.f] - 1)
with D the mean cell diameter, [[rho].sub.f] the foam density and [[rho].sub.p] the matrix density (polycarbonate). Nucleation densities as high as [10.sup.9] were obtained for carbon dioxide with concentrations close to 3 wt%. and higher nucleation values on the order of [10.sup.12] were achieved for 7-8 wt% of n-pentane, as shown in Fig. 5. The transition from the first to the second zone, which corresponds to the dramatic increase in foam density, is thus associated with an abrupt change in the nucleation rate. As the higher nucleation density increases exponentially with a linear increase of the blowing agent concentration, the number of gas molecules per cell is reduced significantly, which affects inevitably the cell expansion and thus the resulting foam density.
[FIGURE 5 OMITTED]
In the case of carbon dioxide, a third zone can be defined for concentrations above 3 wt%. Any additional increase of the foaming agent concentration did not modify any further the foam characteristics: foam density remained around 1000 kg/[m.sup.3], and the cell sizes fluctuated in the 10-50 [micro]m-range.
Also, from Fig. 3 to Fig. 5, we should conclude that the type of resin used, either linear (PC-101) or branched (PK-2870), has no effect on the resulting foam characteristics, contrarily to the observations made from batch microcellular foaming (5). One possible explanation is based on the results from time sweeps, as shown in Fig. 6, that indicated that PK-2870 was prone to undergo chain scission, unlike PC-101. which indicated good thermal stability. Degradation of the branched resin, under the more severe conditions experienced during extrusion, could then be suspected. The time scale involved in the two processes, batch versus extrusion, can also be part of the explanation for the difference in the foam characteristics observed in the batch mode between the linear and the branched resins. The batch process involves low rates of deformation that take place over a significantly long period of time (10-150 sec), while foaming occurs very rapidly during extrusion, resulting in high deformation rates. As seen in Fig. 2, the main difference in the viscosity behaviors of the two PCs lies in the low-rate region, with relaxation times of the same order of magnitude as those of the batch process.
[FIGURE 6 OMITTED]
Although the two types of foaming agents have generated similar trends and foams, it should be noted that the so-called critical PFA concentrations al which we have started producing microcellular loams are quite different: 2.7 wt% for C[O.sub.2] versus 6.5 wt % for n-pentane. Even if the molecular weights of the PFAs are taken into account, 44.01 g/mole for C[O.sub.2] versus 72.15 g/mole for n-pentane, more molecules of n-pentane are required to generate microcellular structures. So we must state at this point that not only the absolute value of the concentration, in weight fraction or on a molar basis, is relevant to the production of microcellular foams, but also the concentration relative to its solubility limit, under the pressure and temperature conditions prevailing during the process, should be explored.
It has been recognized that the foaming agent can act as a very efficient plasticizer in the foam process, particularly when the PFA is combined with an amorphous polymer (11). The decrease in viscosity associated with the lowering of the glass transition temperature as the content in PFA is increased enables the process to be run at temperatures much below those that would be required for the neat polymer only. Moreover, during the bubble nucleation and growth phase, the gradual depletion in plasticizer makes the polymer matrix experience a sudden viscosity rise, which is highly beneficial for the stabilization of the cellular morphology.
Evidence of the plasticization behavior can be given by the continuous decrease of the pressure at the die level, as the concentration of the foaming agent is increased at nearly constant temperature, as illustrated in Fig. 7a. This graph also indicates that the plasticization effect is of the same order of magnitude for either foaming agent. The continuity in the trend displayed in Fig. 7a may also validate the tact that the foaming agent has remained completely dissolved at the die level, for the concentrations tested. Even the data points for a concentration close to 8 wt% for n-pentane belong to the same trend with a pressure at the die that falls below 2 MPa, which will appear further on to be in contradiction with the solubility results. In this case the low pressure experienced at the die level could be imparted to the gas separation that was followed by the migration of the gas to the polymer-metal interface, with the gas finally acting as a lubricant.
[FIGURE 7 OMITTED]
However, it should be noted that the slope for the decrease of the pressure drop experienced at the die level as a function of the PFA concentration is in agreement with the results for the viscosity reduction, as measured for n-pentane using an on-line rheometer attached at the end of the extrusion line, and displayed in Fig. 7b. However, one should take into account that the pressure results (Fig. 7a) were obtained tinder constant throughput or shear rate, while the viscosity reduction numbers in Fig. 7b were computed using a constant-stress basis.
Results displayed in Fig. 7b also incorporate trends observed for polystyrene (PS) mixed with either C[O.sub.2] or n-pentane. Irrespective of the type of PFA, the same plasticization curve was obtained for PS. However, measurements made with PC using the same set-up bill different temperatures do not yield the same conclusions. PC/n-pentane mixture at 220[degrees]C (i.e. 70[degrees]C above the glass transition temperature [T.sub.g] of neat PC) exhibits the same viscosity reduction as that experienced by PS/n-pentane al 175[degrees]C (75[degrees]C above [T.sub.g] of PS). Surprisingly, PC/C[O.sub.2] does not follow the same dependency, and the viscosity reduction measured is relatively small compared to that obtained with the other foaming agent, which is in contradiction to the results displayed in Fig. 7a. This measured viscosity reduction for PC/C[O.sub.2] shown in Fig. 7b, is nevertheless continuous over the investigated range of concentration (up to 3.5 wt% of C[O.sub.2]), and the same trend is followed by either the branched or the linear polycarbonate. The suspected anomaly could be partially imparted to the higher testing temperature (250[degrees]C in that case) and lower solubility of C[O.sub.2] under these conditions, combined with the pressure decrease experienced as the melt sample is pumped toward the on-line rheometer measuring cell, which would initiate phase-separation as the PFA/polymer mixture enters the rheometer slit die. Viscosity increase would occur because of the lack of plasticization associated with the depletion of solvent out of the PC matrix. This behavior did not show up for n-pentane in PC since lower pressures were required to maintain dissolved a specified amount of PFA.
Solubility and Diffusivity
Solubility results for carbon dioxide in polycarbonate are displayed in Fig. 8a and Fig. 8b. State-of-the-art solubility values, as obtained from Rubotherm experiments, indicate that at 180[degrees]C only 2-3 wt% of C[O.sub.2] could be dissolved tinder the standard processing pressures met in the extruder and down to the die exit. The results shown in Fig. 8a also indicate that the structure of the PC resins, linear versus branched. does not influence the solubility of carbon dioxide. However, ultrasonic experiments, which detect the onset of bubble formation during a decreasing pressure ramp, give greater solubility values, at even higher temperature, as shown in Fig. 8b. It should be noted that the results obtained off-line using a static ultrasonic bench agree very well with those gathered in-line during the extrusion experiments. It has been previously reported (7) that the shear involved during flowing conditions could induce premature phase separation, owing to the disentanglement of the macromolecules. The very good agreement found here between static and flowing experiments could originate from the Newtonian rheological behavior that the PC resins exhibit over a wide range of shear rates,
[FIGURE 8 OMITTED]
A tentative explanation for the significant difference between the Rubotherm and the ultrasonic experiments could rely on the low values of the diffusivity of C[O.sub.2] in PC. Diffusivity measurements performed at 100 and 180[degrees]C, and shown in Fig. 8c, yield mean values of 2.0 x [10.sup.7] at 100[degrees]C and 3.5 x [10.sup.6] [cm.sup.2]/s at 180[degrees]C. The diffusion coefficients of the branched and the linear PC are very close at 180[degrees]C, and are practically constant, irrespective of the pressure applied during the test, thus fixed over a broad range of concentrations.
The literature is rather scarce on the diffusivity of gases in molten polymers. It has been reported that diffusion of C[O.sub.2] in PS is about 1.2 x [10.sup.7] [cm.sup.2]/s at 50[degrees]C and that HFC134a in PS is about 6.7 x [10.sup.10] [cm.sup.2]/s at 90[degrees]C and 1.2 x [10.sup.8] [cm.sup.2]/s at 120[degrees]C (12, 13). Our results for the PC/C[O.sub.2] system fall between these two examples, with the PS/C[O.sub.2] being usually reported as a fast system, and that of PS/HFC 134a a slow one with inherent enhanced nucleation properties (14).
Detection of the bubble formation by the ultrasonic sensor could then be delayed by the low mobility of the C[O.sub.2] molecules while they gather into clusters during the nucleation process and the very first time period of the bubble growth.
Solubility measurements for n-pentane in PC, as obtained from in-line ultrasonic experiments, are illustrated in Fig. 9. Under the same pressure and temperature conditions, and based on the in-line ultrasonic results only, n-pentane is twice as soluble than carbon dioxide on a weight fraction basis. For example, with P = 5 MPa and T = 225[degrees]C, solubility of C[O.sub.2] is 4 wt% while it is close to 8 wt% for n-pentane. It should be noted from Fig. 9 that any attempt to measure the degassing conditions at 10 wt% of n-pentane yielded results that deviates from the trends observed at lower concentrations, and this may correspond to some critical limit that should be close to 8 wt%. Although degassing pressures as measured by the ultrasounds are closely related to the solubility, it should be kept in mind that the dissolution kinetics of the foaming agent in the polymer during the extrusion process may be strongly affected by the degree of mixing that prevailed in the last barrel zones, as well as the residence time, especially for those species that exhibit low values of diffusion coefficients (14). Deviations observed at high concentrations of n-pentane in Fig. 9 may be related to conditions where a homogeneous dissolution of the blowing agent has not yet been reached.
[FIGURE 9 OMITTED]
As shown earlier in Fig. 1a, the pressure measured at the injection port might in many cases be a fair indication of the solubility conditions for a pair of polymer and blowing agent. The values reported in Fig. 1a for n-pentane are in close agreement with those originated from the ultrasonic experiments for the 225 255[degrees]C range, which is similar to the temperatures prevailing in the extruder barrel. For C[O.sub.2], a similar trend is observed, at least for concentrations below 1 wt%. At this concentration, the pressure experienced in the extruder barrel level off, even if the weight loss monitored still indicates increasing concentration of carbon dioxide. Some scattering is also present in the pressure data. This could be imparted to mode of injection of C[O.sub.2], i.e. through a gas regulator versus a volumetric pump for n-pentane, and to unsteady state conditions prevailing during start-ups.
Effect of Temperature During Extrusion Foaming
The next series of foaming experiments deals with the effect of the melt temperature, as controlled through the gear pump and die temperatures. Using the linear grade of PC and keeping the amount of carbon dioxide constant, between 2.7 and 3.1 wt%, the temperature was gradually decreased from 280 to 200[degrees]C, in five steps. The foam density of the samples remained almost constant over the five tests, close to 1025 kg/[m.sup.3]. The pressure at the die rose from 3.0 to 11.9 MPa, and this increase should be related to the change in the viscosity due to lower temperatures experienced at the die. Taking into account that the pressure drop in the capillary die is occurring at constant rate for any temperature, it is possible to calculate an energy of activation that would match the one determined at constant modulus (5), which is approximately 100 kJ/mol. According to this observation, we may conclude that the carbon dioxide was kept in solution at any temperature during these five steps.
The impact of the temperature reduction on the cell structure was highly significant, as illustrated from the micrographs of some samples shown in Fig. 10. Presence of large cells can be observed for the sample obtained at high temperature, resulting from poor nucleation and coalescence. At lower temperature, the cell size distribution is more homogeneous, and the larger pressure drop experienced at the die exit seems to have been beneficial to the nucleation process. Even if a low melt viscosity is probably desirable in the early stage of foaming, a greater melt stillness obtained at lower temperature may limit the growth rate of the more well-developed cells, and this would yield a more uniform cell structure. This higher nucleation density experienced at lower temperature is in contradiction with the classical homogeneous nucleation theory that predicts high nucleation with elevated temperature. This illustrates well the significant role that play diffusivity and viscosity during the nucleation and early stage of bubble growth.
[FIGURE 10 OMITTED]
Bimodal Cell Distribution
Some evidence of the critical conditions met during the processing are well illustrated by two experiments conducted with the linear grade of PC mixed with 6.5 wt% of n-pentane. The two experiments have been performed sequentially. In the first one, the temperature profile on the barrel has been modified to induce more cooling in the last barrel zones: zones #6 to #9 were lowered to 180[degrees]C. The same temperature was also set on the gear pump and the die. The resulting foam had a density of 482 kg/[m.sup.3] and a mean cell diameter of 3.1 [micro]m. The foam morphology of this sample is illustrated in Fig. 11a, with the cell size distribution given in Fig. 11c. A bimodal distribution has been obtained under these processing conditions. with the smaller cells having a diameter in the range of 1 to 3 [micro]m, and the larger cells in the 10 [micro]m-range.
[FIGURE 11 OMITTED]
The next experiment was performed attempting even more cooling by lowering the barrel zones #6 to #9, as well as the gear pump and the die, to 170[degrees]C. The resulting foam for this latter trial has a density practically unchanged of 414 kg/[m.sup.3], while the mean cell size has increased to 5.5 [micro]m, with no sign of bimodality in this case, as shown in Fig. 11b and Fig. 11c.
Although the melt temperature has remained practically unchanged, 186[degrees]C for both trials, the pressure at the die level has increased from 2.55 MPa for the previous experiment to 7.58 MPa. The very low melt temperatures achieved in the two experiments have necessitated higher power consumption for the gear pump compared to the usually low values, i.e. less than 10 amps obtained for most of the other trials. Values of 17.5 and 28.2 amps were, respectively, obtained for these two experiments, similar to the consumption observed for the PC free of any foaming agent and extruded at 250[degrees]C. The huge rise in the pressure at the die could not be imparted solely by the increased viscosity that would have accompanied a lower extrusion temperature. First, as indicated through the melt temperature, no significant improvement in the cooling of the melt could be detected. Moreover, the magnitude of the pressure increase would be more easily related to a change in the degree of plasticization: only part of the blowing agent should have remained dissolved.
Figure 12 compares the plasticization trend detailed in the previous Fig. 7a with the two die pressure results obtained from these last two trials. If we associate a concentration of n-pentane that could have remained dissolved under each pressure, then we must conclude that some "extra" foaming agent has already phase-separated. Premature nucleation could thus be expected for both experiments. Lack of plasticization is also validated by the abnormally high values of power consumption at the gear pump, with increasing values associated to decreasing contents of PFA dissolved. Also, more mechanical energy from the extruder went into the polymer during its compounding stage, the torque on the twin-screw extruder having also increased from 40% to 48% of its maximum rated value.
[FIGURE 12 OMITTED]
We can hypothesize that the bimodality observed for the first trial was obtained through two consecutive nucleation steps, one launched prior the gear pump entrance, and the second one, for the remaining foaming agent still dissolved in the polymer matrix, experienced at the die exit. It has been reported for batch processing of mixtures of C[O.sub.2] and PS that stepwise depressurization with an intermediate high pressure stage could yield a bimodal cellular structure (15). The relatively low diffusion rates as those reported previously for C[O.sub.2] in PC, may explain why the foaming agent that may have already phase-separated has little chance to get back dissolved in the polymer matrix, within the narrow time-frame imparted within the die channel. Also, the relatively high pressures maintained within the die prevent the existing cells from growing significantly therein.
Figure 9 indicates that at 200[degrees]C, a pressure as low as 2.0 MPa would be required to maintain 6.5 wt% of n-pentane dissolved in PC. This condition should have been easily met throughout the extrusion process, with no occurrence of a pressure drop to initiate the first nucleation step prior to the gear pump entrance. This premature partial nucleation should then be associated with unstable conditions where part of the blowing agent was not properly homogenized throughout the PC matrix, owing to the very low diffusion coefficient that should have prevailed at that low temperature of 180[degrees]C. PFA concentration gradients in the melt should have been more prone to induce nucleation under higher pressures.
[FIGURE 9 OMITTED]
Microcellular foams of polycarbonate have been produced continuously on a twin-screw extrusion line using either carbon dioxide or n-pentane. High concentrations of the PFA were required, close to what we suspected to be the solubility limits prevailing under the processing conditions. The density of the microcellular foams produced with carbon dioxide was close to 900 kg/[m.sup.3], while higher solubility of n-pentane in PC has enabled the production of foams with lower densities in the range of 500-600 kg/[m.sup.3] still with microcellular structures. Mild temperatures and pressures were required, with temperatures in the 200[degrees]C-range and maximum pressures observed at the end of the extruder below 16 MPa. Bimodal cell size distributions were also obtained under certain conditions. The relatively high pressures in the die combined with the low diffusion rate of the PFA have prevented the existing cells to either expand or return in solution. Use of a gear pump to control the levels of pressure is believed to be one of the key parameters for this process. Lower density foams, in the range of 60-200 kg/[m.sup.3], were also obtained when using smaller amounts of foaming agent, with cell sizes in the range of 1 mm.
The authors wish to thank Dr. Paul Handa and Dr. Zhiyi Zhang for the excellent work in characterizing the solubility and diffusivity of carbon dioxide in the polycarbonate resins. Thanks are also due to Mrs. Nicole Cote for the rheological and morphological characterizations, and to Mr. Louis-Paul Phaneuf for his outstanding assistance during the extrusion trials.
(1.) V. Kumar and J. Weller, J. Eng. Ind., 116, 413 (1994).
(2.) M. R. Holl, J. L. Garbini, W. R. Murray, and V. Kumar, J. Polym. Sci.: Part B: Polym. Phys., 39, 868 (2001).
(3.) C. B. Park, Chapter 11: "Continuous Production of High-Density and Low-Density Microcellular Plastics in Extrusion," pp. 263-305, in Foam Extrusion: Principles and Practice, S.-T. Lee, ed., Technomic Publishing Co., Lancaster, Pa (2000).
(4.) Q. Huang, B. Seibig, and D. Paul, J. Cell. Plast., 36, 112 (2000).
(5.) T. G. Gopakumar and L. A. Utracki, Intern. Plast. Eng. Technol., 4, 1 (2000).
(6.) R. Gendron, L. E. Daigneault, and L. M. Caron, J. Cell. Plast., 35, 221 (1999).
(7.) A. Sahnoune, J. Tatibouet, R. Gendron, A. Hamel, and L. Piche, J. Cell. Plast., 37, 429 (2001).
(8.) O. Pfannschmidt and W. Michaeli, Proceedings of ANTEC '98, 1918 (1998).
(9.) L. A. Utracki and R. Gendron, J. Rheol., 28, 601 (1984).
(10.) E. J. Beckman, Proceedings of Foamplas 2000 Conference, p. 85, Chicago (May 2000).
(11.) R. Gendron and L. E. Daigneault, Chapter 3: "Rheology of Thermoplastic Foam Extrusion Process," in Foam Extrusion: Principles and Practice, pp. 35-80, S.-T. Lee, ed., Technomic Publishing Co., Lancaster, Pa (2000).
(12.) Z. Zhang and Y. P. Handa. J. Poly. Sci.: Part B: Polym. Phys., 36, 977 (1998).
(13.) B. Wong, Z. Zhang, and Y. P. Handa, J. Poly. Sci.: Part B: Polym. Phys., 36, 2025 (1998).
(14.) R. Gendron, M. Huneault, J. Tatibouet, and C. Vachon, Proceedings of Blowing Agents and Foaming Processes 2002 Conference, p. 97. Heidelberg, Germany, (May 2002).
(15.) K. A. Arora, A. J. Lesser, and T. J. McCarthy, Macromolecules, 31, 4614 (1998).
RICHARD GENDRON ([dagger]) and LOUIS E. DAIGNEAULT ([double dagger])
Industrial Materials Institute, National Research Council of Canada
75 de Mortagne Boulevard, Boucherville, Quebec, Canada, J4B 6Y4
([dagger]) To whom correspondence should be addressed. E-mail: email@example.com
([double dagger]) Currently at IPEX inc., Research & Development, 2441 Royal Windsor Drive, Mississauga, Ontario, Callada L5J 4C7.
This paper was presented at "Emerging Technologies for the New Millennium," held in Montreal, Dec. 10-11, 2001.
|Printer friendly Cite/link Email Feedback|
|Author:||Gendron, Richard; Daigneault, Louis E.|
|Publication:||Polymer Engineering and Science|
|Date:||Jul 1, 2003|
|Previous Article:||Expansion mechanisms of plastic/wood-flour composite foams with moisture, dissolved gaseous volatiles, and undissolved gas bubbles.|
|Next Article:||Effects of die geometry on cell nucleation of PS foams blown with C[O.sub.2].|