Orientation and Crystallization in Poly(Ethylene Terephthalate) During Drawing at High Temperatures and Strain Rates.
Poly(ethylene terephthalate) film has numerous applications requiring a range of different properties, which are directly related to various details of the molecular organization imparted during film formation. Greater understanding of the influence of process parameters on the evolution of microstructure is key to the optimization of properties for specific applications and to the design and control of processing conditions for improved productivity.
The commercial process most commonly used in the manufacture of biaxially-drawn PET film involves three steps: 1) Amorphous, unoriented film is drawn longitudinally (in the machine direction) between sets of rolls, which is a constant force (CF) deformation; 2) The film is drawn in the transverse direction (orthogonal to the machine direction) at a constant extension rate (CER); 3) The film is heat set at temperatures in the region of 220[degrees]C to impart dimensional stability.
The draw temperature in the forward draw is generally in the range 80-95[degrees]C, and in the transverse draw it is usually 5-15[degrees]C higher. For some applications the required properties are more readily obtained by reversing steps 1 and 2, known as the "inverse" process. A third method is simultaneous biaxial drawing in which the amorphous, unoriented film is stretched in two orthogonal directions simultaneously at a constant extension rate. Again, drawing temperatures are normally in the range 80-95[degrees]C.
In the first part of this paper, we provide a brief review of structure formation in PET film during CER and CF drawing from the amorphous state (step 1, above) at temperatures above the glass transition, focusing on those aspects that help to interpret the results of the experimental study reported in the second part of the paper. In the experimental study, we monitor orientation and crystallinity development during CF drawing at temperatures that are 30-40[degrees]C higher than in conventional processes, and we assess the effects on film structure of two-stage drawing, in which the high temperature draw at 132[degrees]C is followed by drawing at 100[degrees]C along the same axis.
Constant Extension Rate Drawing
The drawing behavior of amorphous PET shows a dramatic change as the draw temperature is increased through the glass transition range. At temperatures below about 65[degrees]C, the deformation involves necking. and the development of molecular orientation with strain is essentially independent of draw temperature and strain rate . In this temperature region, the deformation does not involve cooperative molecular rearrangements, and the orientation development can be modeled by the pseudo-affine scheme in which molecular units orient independently [1, 2]. However, on drawing at temperatures higher than [T.sub.g] (about 75[degrees]C), the strain is essentially homogeneous throughout the specimen and orientation increases more gradually than in "cold drawing". Above [T.sub.g], the rate dependence of orientation development during deformation increases with temperature. In the region of 80[degrees]C, the rate dependence is relatively small and the deformation behavior is similar to that of a rubber-like mate rial, in which entanglements provide a network of crosslinks [3-6]. But as temperature is raised further, the acceleration of relaxation processes provides an increasing viscous flow component to the deformation behavior, attributable to the slippage of entanglements, especially at moderate to high draw ratios.
Studies of microstructure development during constant extension rate (CER) drawing of PET film, at strain rates [less than or equal to] [2S.sup.-1] and temperatures between 80 and 100[degrees]C, have revealed the strong influence of the time available for molecular relaxation [7-20]. At a given temperature, the time available for relaxation controls the level of molecular orientation attained at a given draw ratio which, in turn, determines the degree of crystallinity that can develop. This is because crystallization is fast relative to the timescale of the drawing operation and there is time to reach, or approach, the pseudo-equilibrium crystallinity level corresponding to a particular combination of orientation and temperature.
Figure 1 shows schematics of the effects of strain rate, temperature and molecular weight on the development of noncrystalline orientation with draw ratio [lambda] [8, 12, 17]. Noncrystalline orientation develops faster with increasing strain rate, with increasing molecular weight and with decreasing temperature. This is because the time available for molecular relaxation decreases with increasing strain rate, and the relaxation rate decreases with increasing molecular weight and decreasing temperature.
The time-dependence of orientation development is responsible for the strain-rate dependence of crystallization onset. At a given draw temperature, a critical molecular orientation must be reached to induce crystallization. (For the purposes of the present discussion, this critical orientation will be considered essentially independent of strain rate [8, 9], although recent data has revealed evidence of a significant strain-rate dependence ). Since orientation develops more slowly at lower strain rates, the critical orientation, and therefore the onset of crystallization, is shifted to higher draw ratios as strain rate decreases (Figure 2) [8, 11, 14]. The effect of increasing temperature is to increase the severity of the strain-rate dependent shift in crystallization onset, because the rate dependence of noncrystalline orientation development increases with temperature (Figure 1) [8, 14]. Decreasing molecular weight has a similar effect to increasing temperature .
There is evidence that the degree of crystallinity at a particular draw ratio is largely dependent on the level of noncrystalline orientation obtained and, to a much smaller degree, on the time taken to reach that orientation . It can then be argued that, at a given level of noncrystalline orientation, there is a pseudo-equilibrium crystallization region where the crystallization kinetics are much slower than in the initial crystallization stage and that the level of crystallinity at the onset of this region increases with the degree of noncrystalline orientation attained (Figure 3) . Within the strain rate range typical of CER drawing, there is always sufficient time to approach or enter the region of slower kinetics, so that the degree of crystallinity induced at each draw ratio represents a pseudo-equilibrium value. Moreover, it has been found that changing the strain rate of deformation, at a given draw temperature, simply shifts the crystallinity-time curves along the log-time axis without changing their shape, and that the shift factor and the strain rate are related by a power law [11, 13]. The value of the exponent n increases with temperature and decreases with increasing molecular weight (Figure 4) because n reflects the strength of the strain-rate dependent shift in the onset and kinetics of strain-induced crystallization, which is related to the rate of molecular relaxation [13, 19].
At low strain rates, increasing draw temperature causes a pronounced increase in the draw ratio for onset of crystallization ([[lambda].sub.c]), because higher temperatures enhance the rate of orientational relaxation [7, 13, 21]. However the temperature-dependent shift in [[lambda].sub.c] diminishes as strain rate increases, and at sufficiently high strain rates, increasing temperature shifts [[lambda].sub.c] to lower draw ratios (Figure 5) . This is because higher temperatures not only increase the rate of orientational relaxation, but also increase the rate of crystallization at a given level of noncrystalline orientation and reduce the critical orientation for onset of crystallization [8, 17]. Thus, when the time available for relaxation becomes very short, at high strain rates, the effect of enhanced crystallization rate dominates [13, 20].
Structure evolution during hot-drawing of PET can be divided into three regimes that are reflected in the stress-strain response (Figure 6) [10, 14]. The first regime involves the extension of an amorphous network of entangled chains, and the magnitude of the induced stress depends on the degree of molecular relaxation occurring in the course of the deformation. The second regime begins at the onset of crystallization, where the provision of crystallite junctions arrests or slows down entanglement slippage and gives rise to an inflection point in the true-stress vs. strain relationship. In this regime (crystallization regime 1) crystallinity and noncrystalline orientation develop rapidly while stress develops slowly. At a characteristic level of crystallinity, an abrupt decrease in the rate of development of both crystallinity and noncrystalline orientation marks the transition to the third regime (crystallization regime 2) and coincides with a sharp upturn in stress. There is now various evidence [10, 14, 21 -23], including a comparison of the stressstrain-orientation behavior of PET with that of uncrystallizable PEMT , that the increase in stress largely results from an increase in polymer viscosity arising from the interconnection of crystallites. The slowing of orientation in the high-stress region may reflect translational slippage between groups of crystallites held together by extended tie-chains [13, 14].
Constant Force Drawing
During constant force (CF) drawing, the structural evolution of the film is reflected in the evolution of strain rate rather than in the evolution of stress. In the temperature range 80-100[degrees]C, strain rate passes through a maximum in the course of CF deformation (unless the deformation is stopped before the maximum is reached) . This is because relaxation phenomena decrease the polymer modulus in the course of the deformation, reducing resistance to deformation, and then orientation-induced crystallization increases the modulus again, slowing the deformation kinetics [24-27]. Thus, constant force drawing at some average strain rate involves an excursion to strain rates that can be several times higher than the average value.
In the range of (average) strain rates between 5 and 17 [S.sup.-1], which are typical rates for commercial processes, studies have shown a negligible influence of strain rate on orientation development in CF drawing, whereas increasing temperature between 83 and 96[degrees]C decreases the rate at which the chains orient (Figure 7) . Since the high strain rates in CF drawing reduce the time available for relaxation of orientation, axial orientation of the noncrystalline chains evolves more rapidly than in CER drawing , as does orientation of the benzene rings in the film plane . However, there is strong evidence that in the region of the strain rate maximum, during which crystallization is taking place, there is insufficient time for crystallinity to reach pseudo-equilibrium levels associated with a given level of noncrystalline orientation (Figure 8) . Thus, at a given orientation, the degree of crystallinity is lower in CF drawing than in CER drawing.
It appears that laboratory studies of PET drawing have been limited to temperatures below 115[degrees]C, because at higher temperatures it is difficult to heat the film with sufficient speed to avoid thermally induced crystallization in the unoriented film prior to drawing. In order to examine structure evolution from the amorphous state during drawing at higher temperatures, we have built a constant-load drawing instrument that can heat the film to 150[degrees]C, and probably higher, with sufficient speed to avoid crystallization in the undrawn material. A description of the drawing instrument and some preliminary results from high temperature drawing are presented in the following sections.
Amorphous, unoriented PET film was supplied by Goodyear. It has a number average molecular weight of 21000 (i.v. of 0.66 [dlg.sup.-1]), a density of 1337 [kgm.sup.-3], and a thickness of 253 [micro]m. The film was of high clarity and did not contain [TiO.sub.2] or other additives.
High Temperature Deformation
The film specimen, measuring 8 mm in the draw direction and 75 mm across, was marked with horizontal and vertical ink lines in order to measure the deformation accurately (see Ref. 10). The specimen was clamped between the upper (fixed) grip and the lower (movable) grip of the constant-load drawing instrument (described below), heated to the desired temperature, and then immediately drawn to the total imposed draw ratio by applying a nominal stress of 7 MPa to the lower grip. During extension, the segments of the ink grid near the vertical edges of the film become narrow, whereas the center segments maintain their initial width. Microstructure characterization was carried out only on segments drawn at constant width (uniaxial-planar symmetry). Draw temperatures of 120[degrees]C and 132[degrees]C were used.
Since CF deformation is far from homogeneous at temperatures above [T.sub.g] (unlike CER drawing), accurate characterization of the deformation kinetics at each temperature would require a fast video recording of the deformation of individual segments. Our experimental setup does not include such a system at present, and we can therefore only provide a rough estimate of the kinetics involved. The average strain rates resulting from the applied load were between 20 [S.sup.-1] and 72 [D.sup.-1] depending on draw ratio and draw temperature. The maximum strain rate appears to be reached at draw ratios in the region of 3-3.5 (after a drawing time of approximately 0.04-0.06s) and seems to have a value in the range 75 [s.sup.-1]-150 [s.sup.-1]. However, individual monitoring of segment deformation is essential for a reliable determination of the maximum strain rate, and we expect to provide a more confident description of the deformation kinetics in subsequent publications.
In an attempt to determine the "natural draw ratio" (or equilibrium deformation) under a nominal stress of 7MPa at 120[degrees]C and 132[degrees]C, the film was allowed to deform without an imposed limit. We found that, at both temperatures, the film broke under the load when certain segments of the film reached a draw ratio in the region of 8-10. This indicates that the structural evolution taking place during drawing at these temperatures does not increase the strength of the film sufficiently to counteract the increase in true stress that results from the reduction in cross-sectional area.
Low Temperature Deformation (Second Draw)
After the first draw step at 132[degrees]C, some film segments were redrawn at 100[degrees]C along the same axis. Specimens were removed from the drawing instrument after the first draw and the size of each segment marked on the film was measured. The drawn specimen was cut back to the original length of 8 mm in such a way that the (measured) segments of interest were in the center. The specimen was then remounted in the instrument, redrawn at the second temperature and quenched. After removing the specimen from the apparatus, the segments were re-measured. The nominal stress applied was 19.5 MPa, and the draw ratio imposed in the second draw [[lambda].sub.2] were as follows (with the draw ratio in the first draw [[lambda].sub.1] given in parenthesis): 3.3 ([[lambda].sub.1] = 2.6), 3.1 ([[lambda].sub.1] = 2.9), 1.9 ([[lambda].sub.1] = 3.9), 1.6 ([[lambda].sub.1] = 4.1). This resulted in final draw ratios [[lambda].sub.f] of 8.8, 9.0, 7.4 and 6.6.
Constant Load Drawing Instrument
The constant-load drawing instrument was designed to permit constant load drawing at temperatures up to about 180[degrees]C. It includes software, written in LabVIEW(R), that controls all aspects of the experiment. Further details of instrument operation during a typical drawing sequence are given below.
1. An infrared heater, situated behind an hydraulic shutter that separates it from the specimen, is heated-up to a predetermined temperature. When the heater reaches the temperature, the shutter opens, exposing the undrawn film to infrared heat.
2. A pyrometer focused on the film measures the film temperature as it heats up. As soon as the film reaches the required draw temperature the load is applied by removal of a pneumatically activated pin from the lower grip, which is attached to a dead-weight.
3. The lower grip falls until it is stopped by contact with a collar, the position of which has been pre-adjusted to permit the desired grip displacement (draw ratio).
4. An optical sensor triggers the quenching operation such that cooling starts immediately at the end of the draw. The quenching operation involves three simultaneous actions: the IR heater is shut off, the shutter is closed, and cold air is blown over the sample. Cooling to the vicinity of room temperature is completed within 2s. In the experiments of the present study, there was in fact a 120 ms delay between the end of draw and the onset of cooling. Adjustments to the operation of the optical sensor have now eliminated this delay, and in future studies we will be able to determine whether l20ms of constant-length heating at the draw temperature has a significant influence on microstructure development.
5. Throughout the drawing experiment, data points are acquired at a rate of up to 20,000 per second. Plots of total extension vs. time, temperature vs. time, and force vs. time are output on the computer screen.
Crystallinity from Density
The density [rho] of the film specimens was measured at 23[degrees]C in a density gradient column containing n-heptane and carbon tetrachloride. The volume fraction crystallinity x was estimated from the usual relationship (see, for example Ref. 10) using a value for the crystalline density [[rho].sub.c] of 1457 kg/[m.sup.3], based on unit cell dimensions close to those reported by Daubeny et al. . It is recognized that crystalline density can sometimes vary, but measurement of lattice spacings by X-ray diffraction indicates that it does not do so under the range of crystallization conditions applied in the present study. The value for the amorphous density [[rho].sub.a] was taken as the measured density of the undrawn film (see Figures 11 and 12a of Ref. 10). Based on the estimation of Nobbs, Bower and Ward  and on our own experimental evidence  we would not expect orientation in the noncrystalline regions to influence the value of [[rho].sub.a] unless the degree of noncrystalline ("amorphous") orientation were to exceed 0.45, which it rarely does in the present study. The values of x that will be reported represent an average of at least three density determinations.
Orientation in the Noncrystalline Regions From Chain-Intrinsic Fluorescence
The polarized chain-intrinsic fluorescence method for determining molecular orientation in the noncrystalline regions of PET has been described in detail elsewhere . In the present study, the fluorescence emission intensities [I.sub.ij] for one-way drawn specimens (along [X.sub.3]) were measured with the polarizer along [X.sub.i] and the analyzer along [X.sub.j] to obtain [I.sub.33], [I.sub.31] = [I.sub.13] and [I.sub.11]. Using the normalization condition [I.sub.33] + 4[I.sub.31] + (8/3)[I.sub.11] = 1, the second moment of the orientation distribution [[less than][P.sub.2](cos [theta])[[greater than].sub.a/fl] was obtained. It should be emphasized that since the samples were drawn at constant width, they may possess some degree of planar orientation, such that the plane of the benzene rings tends to lie parallel to the film surface. This lack of cylindrical symmetry means that the [[less than][P.sub.2](cos [theta])[[greater than].sub.a/fl] values in these samples represent the in-plane axial orientation o f the uncrystallized chains; they do not describe the orientation distribution in the thickness direction of the film.
RESULTS AND DISCUSSION
Single Stage Drawing at High Temperature
Results from the high temperature experiments show a delayed onset of molecular orientation (Figure 9a). For 120[degrees]C drawing, the chains start to orient when the draw ratio reaches about 2.3, and for 132[degrees]C drawing, orientation begins at a draw ratio of about 3.4. Thus, at high temperatures, the initial stage of constant load deformation is characterized by flow drawing where extension occurs by slippage of entanglements due to the high rate of molecular relaxation. Once orientation begins, however, it develops rapidly before abruptly reaching a saturation level of about 0.3 for 120[degrees]C drawing and 0.2 for 132[degrees]C drawing.
This behavior clearly differs from that observed at lower draw temperatures. In CER drawing at temperatures in the range 80 to 100[degrees]C, molecular orientation increases as soon as drawing begins (Figure 1). The exception to this is when strain rate is very low (flow drawing), in which case orientation never increases significantly however high the draw ratio might be, because the time available for relaxation is too long. For CF drawing at temperatures in the range 80 to 96[degrees]C, orientation also appears to increase at the onset of drawing (Figure 7), although there is not complete certainty about this due to a lack of data points at draw ratios less than 2. At any rate, if there is an induction period (flow regime) at these draw temperatures it must be much shorter than in CF drawing at temperatures in the region of 120-132[degrees]C.
The higher orientation values reached at 120[degrees]C than at 132[degrees]C might be assumed to result from a lower rate of molecular relaxation at the lower temperature. This is certainly the case in the temperature range 83-96[degrees]C, where the rate of orientation development decreases with increasing temperature. However, the rate of development of orientation at 120[degrees]C and at 132[degrees]C are actually very similar to each other, and the final orientation at 132[degrees]C is lower because the orientation process stops sooner at this temperature. The sudden cessation of orientation development probably arises from the formation of a crystallite network. The network may constrain orientation of tie chains that are not already fully extended, and force slippage between groups of interconnected crystallites by efficient transfer of stress from the rigid (permanent) crystallite junction points to the weaker (impermanent) entanglement junction points. Since the crystallization rate at a given level of noncrystalline orientation would be faster at the higher temperature, the network would be expected to form at a lower orientation level.
Direct comparison of noncrystalline orientation development over the temperature range 83 to 132[degrees]C may be somewhat doubtful because the two higher draw temperatures involved higher average strain rates than the lower draw temperatures. On the other hand, orientation development during CF drawing in range 83[degrees]C-96[degrees]C appears to be independent of strain rate when the (average) strain rate is in the region of [5S.sup.-1] and higher . Figure 10 can therefore be considered to provide a useful overview of the influence of temperature on noncrystalline orientation development, from which the decrease in the orientation maximum with increasing temperature is rather striking. It should be noticed that instead of drawing the curves at the lower draw temperatures from the origin (as was done in Figure 7 and in Ref. 18), we have implied a short flow region, which increases with temperature, only to point out that this is an alternative possibility in need of further investigation.
The data in Figure 9a may be interpreted as follows. Compared to the rate of deformation imposed by the applied load, the rate of molecular relaxation at the draw temperature is too fast to permit orientation, so that flow drawing occurs. However, at some point during constant load drawing, the strain rate starts to increase sharply due to a decrease in the entanglement network density. Lorentz and Tassin  have argued that the predominant relaxation process active over the range of strain rates involved in a single CF deformation is likely to be that of chain retraction. In this relaxation process, entanglement constraints are released and the deformed contour length of the chain retracts back towards its equilibrium contour length within the deformed "tube" . Thus, the retraction process results in a greater number of monomers between entanglements, and the increase in strain rate resulting from the less resistant network reduces the time available for relaxation to a point where chains can start to orient.
As with orientation, crystallization onset occurs at a higher draw ratio at 132[degrees]C than at 120[degrees]C (Figure 9b). In fact, crystallization onset and orientation onset occur almost simultaneously. In CER and CF drawing below 100[degrees]C, it is necessary to reach a critical molecular orientation in the range 0.1-0.3 before crystallization is induced [17, 18]. This is because thermally induced crystallization in unoriented PET does not occur at temperatures below 100[degrees]C, or at least the induction period is of the order of 10 minutes to several hours (see, for example ref. 31), and a significant degree of molecular orientation is required before crystallization can occur during drawing. However, at 120[degrees]C the induction period in unoriented PET is about 3 min, and at 132[degrees]C it seems to be in the order of seconds . It can be assumed that the induction period at these high temperatures would be substantially reduced by even a minimal degree of molecular orientation [32-34].
Because of high molecular mobility at 120 and 132[degrees]C, very little molecular orientation is required to induce crystallization, and therefore crystallization onset lags orientation onset by only a small margin. The strain rate is so fast during the stage where rapid orientation and crystallization occur that it is very difficult to "capture" a segment of film in this draw ratio region. The consequent lack of data in the region where rapid microstructural changes are taking place makes it difficult to determine the critical orientation values for onset of crystallization at these temperatures, but they can be roughly estimated to lie in the range 0.03 to 0.05.
The shape of the plots of volume fraction crystallinity vs. draw ratio in Figure 9b are more similar to those generated in CER drawing at temperatures below 100[degrees]C than to those obtained from CF drawing in that temperature range (see, for example Figure 8). On the basis of arguments made elsewhere , and discussed in the Review section, this implies that at the strain rates applied in the present high-temperature CF experiments, there is sufficient time for crystallization to reach pseudo-equilibrium values at a given level of noncrystalline orientation. Thus, although the strain rates are somewhat higher in the present study than in the CF studies at lower temperature, it seems that the higher crystallization rates at 120[degrees]C and 132[degrees]C more than compensate for this. However, as mentioned in the Experimental section, the drawing sequence in the constant load instrument involved a 120 ms delay between the end of drawing and the start of cooling. At the high temperatures and high strain rates involved, it is certainly possible that some significant crystallization occurred during this period. For future experiments we have virtually eliminated this delay, and it will be interesting to see if structure development during high temperature drawing is influenced by the change.
It may be recalled that in CER drawing the critical orientation for onset of crystallization was found to decrease with temperature whereas in CF drawing it was found to be independent of temperature in the range 83-96[degrees]C and to have a value of 0.18. We believe that this absence of temperature dependence in CF drawing is related to the lack of time available to reach pseudo-equilibrium crystallinity values . Since there is evidence that the current CF drawing conditions permits crystallinity to proceed to pseudo-equilibrium values, temperature dependence of the critical orientation would be expected.
Two Stage (High Temperature/Low Temperature) Drawing
Films segments drawn at 132[degrees]C to draw ratios ([[lambda].sub.1]) of 2.6, 2.9, 3.9 and 4.1 were redrawn at 100[degrees]C to draw ratios ([[lambda].sub.2]) of 3.3, 3.1, 1.9 and 1.6 respectively, resulting in final draw ratios ([[lambda].sub.f]) in the range 6.6-9.0. The level of noncrystalline orientation in the films prior to the second draw ranged from zero, at [[lambda].sub.1] = 2.6 to almost 0.2 at [[lambda].sub.1] = 4.1. and the crystallinity levels ranged from 0 to 0.12.
It is evident from Figure 11a that when the films segments drawn at 132[degrees]C are redrawn at 100[degrees]C, the levels of noncrystalline orientation obtained are substantially higher than the levels obtainable from single-stage drawing at 132[degrees]C. Although similar orientation values were reached at the end of the second draw, irrespective of the initial orientation level, it is not yet clear whether "saturation" values of orientation for these drawing conditions have been reached in all four specimens, or whether the orientation levels would continue to increase on separate curves of orientation versus [[lambda].sub.f] (or [[lambda].sub.2]).
It is interesting that the second draw increases orientation in the sample with an initial draw ratio of 4.1. Since a [[lambda].sub.1] of 4.1 is on the edge of the plateau region, further drawing of this sample at 132[degrees]C would not have resulted in additional orientation. It appears, therefore, that reduced molecular mobility during the 100[degrees]C draw diminishes the slippage processes that take place at the higher draw temperature. It should be noted moreover, that at only slightly higher values of [[lambda].sub.1], the films broke almost immediately, or drew very unevenly, during the second draw. It is for this reason that segments with [[lambda].sub.1] [greater than] 4.1 were not included in microstructure studies.
The degree of crystallinity increases during the second draw to values that are lower than those in single-stage (132[degrees]C) drawn film at an equivalent final draw ratio (Figure 11b). Apparently the lower crystallization rate (for a given noncrystalline orientation) at 100[degrees]C is not sufficiently enhanced by higher orientation levels to match the degree of crystallinity obtained during 132[degrees]C drawing.
A result of practical interest from the two-stage drawing study is that high draw ratios can be achieved (e.g. [[lambda].sub.f] = 9) at high strain rates without compromising the final levels of orientation and crystallinity.
When amorphous PET film is drawn under a constant nominal stress of 7 MPa at temperatures in the range l20-132[degrees]C, the initial stage of the deformation is characterized by polymer flow, where neither molecular orientation nor crystallization take place as a result of high rates of molecular relaxation. However, strain-rate increases sharply in the course of the deformation, reducing the time available for relaxation, and the chains start to orient at a draw ratio that depends on temperature. At 120[degrees]C the orientation begins at a draw ratio of 2.3, and at 132[degrees]C it begins at a draw ratio of 3.4. Orientation proceeds very rapidly and then abruptly reaches a saturation level of 0.3 for 120[degrees]C drawing and 0.2 for 132[degrees]C drawing. Crystallization onset seems to lag only slightly behind orientation onset, because the critical orientation for onset of crystallization is very low at these temperatures. It appears that under the strain rate conditions resulting from the load applied, there is time for crystallization to proceed to pseudo-equilibrium values corresponding to a particular orientation level, which differs from results obtained in CF drawing at lower temperatures. This may be because the strain rates of the present study are not high enough to compensate for the faster crystallization kinetics at the higher temperatures, or it my be due to a l20ms delay between the end of drawing and the start of quenching.
In a two stage drawing procedure, where film drawn at 132[degrees]C was redrawn along the same axis at 100[degrees]C, total draw ratios as high as 9 were obtained despite the high strain rates involved. The levels of noncrystalline orientation reached in the two stage process are substantially higher than the levels obtained from single-stage drawing at 132[degrees]C, whereas the degree of crystallinity attained is lower.
These studies were undertaken in connection with the TRI project "Structure and Properties of Poly(ethylene terephthalate) Film," supported by a group of Corporate TRI/Princeton Participants. The author is indebted to Harry Buvel and Carl Gormen for construction of the high temperature drawing instrument, to Diane Buvel for writing the instrument software, and to Dermis Briant for his careful experimental work.
(1.) F. Rietsch, R. A. Duckett and I. M. Ward, Polymer, 20, 1133 (1979).
(2.) I. M. Ward, Proc. Phys. Soc., 80, 1176 (1962).
(3.) W. Kuhn and W. Grun. Kolloid Z., 101, 248 (1942).
(4.) S. W. Allison and I. M. Ward, Brit. J. Appl. Plugs., 18. 1151 (1967).
(5.) P. R. Pinnock and I. M. Ward, Trans. Faraday Soc., 62, 1308 (1966).
(6.) I. M. Ward, Polym. Eng. Sci., 24, 724 (1984).
(7.) J. E. Spruiell, D. E. McCord and R. A. Beuerlein, Trans. Soc. Rheol, 16, 535 (1972).
(8.) G. Le Bourvellec, L. Monnerie and J. P. Jarry, Polymer, 27, 856 (1986).
(9.) G. Le Bourvellec, L. Monnerie and J. P. Jarry, Polymer, 28, 1712 (1987).
(10.) D. R. Salem, Polymer, 33, 3182 (1992).
(11.) D. R. Salem, Polymer, 33, 3189 (1992).
(12.) B. Clauss and D. R. Salem, Polymer, 33, 3193 (1992).
(13.) D. R. Salem, Polymer, 35, 771 (1994).
(14.) B. Clauss and D. R. Salem, Macromolecules. 28, 8328 (1995).
(15.) A. Ajji, J. Guevremont, K. C. Cole and M. M. Dumoulin, Polymer, 37, 3708 (1996).
(16.) A. Ajji, K. C. Cole, M. M. Dumoulin and I. M. Ward, Polym. Eng. Sci., 37, 1801 (1997).
(17.) D. R. Salem, to be published. An abbreviated version appears in Book of Extended Abstracts: Twelfth Annual Meeting of the Polymer Processing Society, p. 5.6/1, Sorrento, Italy (1996).
(18.) D. R. Salem, Polymer, 39, 7067 (1998).
(19.) D. R. Salem, to be published. An abbreviated version appears in Book of Extended Abstracts: Ninth Annual Meeting of the Polymer Processing Society, p. 83, Manchester, U.K. (1993).
(20.) D. R. Salem, Polymer, 36, 3605 (1995).
(21.) R. Pearce, K.C. Cole, A. Ajji and M. M. Dumoulin, Polym. Eng. Sci., 37. 1795 (1997).
(22.) C. P. Buckley, D. C. Jones, and D. P. Jones, Polymer, 37, 2403 (1996).
(23.) M. Vigny. J. F. Tassin, A. Gibaud and G. Lorentz, Polym. Eng. Sci., 37, 1785 (1997).
(24.) G. Le Bourvellec, J. Beautemps and J. P. Jarry, J. Appl. Polym. Sci., 39, 319 (1990).
(25.) G. Le Bourvellec and J. P. Jarry, J. Appl. Polym. Sci., 39, 329 (1990].
(26.) G. Lorentz and J. F. Tassin, Polymer, 35. 3200 (1994).
(27.) P. Lapersonne, J.-F. Tassin, L. Monnerie and J. Beautemps. Polymer, 32, 3331 (1991].
(28.) R. de P. Daubeny, C. W. Bunn and C. J. Brown, Proc. Roy. Soc., A226, 531 (1954).
(29.) J. H. Nobbs, D. I. Bower and I. M. Ward. Polymer, 17, 25 (1976).
(30.) M. Doi and S.F. Edwards, The Theory of Polymer Dynamics, Charendon Press, Oxford, England (1986).
(31.) K. G. Mayhan, W. J. James and W. Bosch, J. Appl. Polym. Sci., 9, 2605 (1965).
(32.) T. Sun, J. R. C. Pereira and R. S. J. Porter, J. Polym. Sci, Phys.Edn. 22, 1163( 1984).
(33.) A. Ziabicki in Fundamentals of Fiber Formation, Ch. 2, p. 125, John Wiley & Sons (1976).
(34.) F. S. Smith and R. D. Steward, Polymer, 15, 283 (1974).
|Printer friendly Cite/link Email Feedback|
|Author:||SALEM, D. R.|
|Publication:||Polymer Engineering and Science|
|Date:||Dec 1, 1999|
|Previous Article:||Physical and Microstructural Effects of Heat Setting in Polyester Films.|
|Next Article:||Influence of Interchange Reactions on the Crystallization and Melting Behavior of Nylon 6,6 Blended With Other Nylons.|