Injection molding: the injection spin process.
In the usual injection molding process, polymer is pressurized to flow into a cavity. Orientation of macro-molecules develops in the flow direction as a result of shear stress. This pressure-flow-induced orientation leads to a skin-core structure in the injection molded articles. Orientation is strong in the skin layer because shear stress is high near the wall of the cavity. Orientation in the core layer is random because of lower shear stress and longer time for the molten polymer to relax. Morphology of the skin-core structure has been documented in many cases.[1-3] Although mechanical performance of the polymer is usually dominated by its chemical structure, the morphological structure--such as orientation--may play an important role. Fujiyama and Wakino reported that the skin layer demonstrated better mechanical performance than the core layer because of a higher degree of orientation. To increase bulk material strength, we have developed a process to introduce orientation in the core layer. This extra orientation was drag-flow-induced, implemented by rotating a core immediately after the cavity was filled. Because the direction of core rotation was perpendicular to the direction of polymer flow into the cavity (machine direction), we expected to create a "double-cross-orientation" in the article. Our design is shown schematically in Fig. 1. Pressure-flow-induced orientation was frozen in the two skin layers and in the machine direction. Drag-flow-induced orientation was added subsequently to the core layer at a right angle to the machine direction.
Molding Process. The injection-spin process was demonstrated in our four-cavity syringe barrel mold, in which one core was able to rotate inside one of the four cavities. Rotation was driven by a 0.5-hp motor controlled by a computer in a programmed timing sequence. Maximum speed of rotation was 195 rpm. Acceleration of the rotation from zero to maximum speed was approximately 0.5 sec. A mechanical-electronic switch was installed in the molding machine so that when the mold closed, an electronic signal was sent to an analog/digital board in the computer to give reference time. Using a computer timer, we can specify when to start and stop the spinning.
Isotactic PP resin was used to mold 20-cc syringe barrels, about 10 cm long and 2 cm in diameter. To evaluate the effect of core rotation, syringe barrels were molded at two conditions--spinning the core at maximum speed (195 rpm) for 1.5 sec, or no spinning. All other molding parameters were the same: injection temperature, 238 [degrees] C; packing pressure, 34 MPa for 1 sec, then 14 MPa for 0.2 sec; mold temperature, 27 [degrees] C; cooling time, 10 sec; back pressure, 1.4 MPa; and screw speed, 200 rpm. The molded syringe barrels were visually identical.
In this article, we denote iPP-S and iPP-N as the samples that were made with and without core spinning, respectively.
Scanning Electronic Microscopy (SEM). Fracture surfaces of the syringe barrels were examined using a Leica 360 SEM. To avoid damaging the morphology, barrels were cooled with liquid nitrogen and then broken into pieces. Figure 2 shows an SEM micrograph of the iPP-S barrel in a longitudinal section.
X-ray Diffraction. Samples of iPP-S and iPP-N were tested at six positions: three evenly spaced along the syringe barrel (in the machine direction) from tip to flange, and three in the same relative positions but 90 [degrees] around the barrel from the gate position. X-ray azimuthal plots of the 040 and 110 reflection planes were obtained in tests of iPP-S and iPP-N samples.
Birefringence. We attempted to measure birefringence of the barrel using a polarizing microscope equipped with a compensator. Clear fringe patterns were seen in the iPP-N sample; no fringe pattern was visible in the iPP-S sample.
Another birefringence test was conducted with a pair of polarizer plates. A weld line was observed in every iPP-N sample as a dark line running the length of the syringe barrel, but was absent in iPP-S samples.
Tensile Evaluation. Tensile testing was performed with an Instron 1122. Because properties may be anisotropic, we evaluated tensile strength in the machine and hoop directions. Two types of test pieces were cut from the barrels: strips, 10 x 1 cm, and arcs, 1 cm wide. A special fixture was made for clamping the arc specimen. Five replicates were averaged in each experimental run.
Dynamic Mechanical Test. A Rheometrics RSA-II analyzer was employed to examine dynamic mechanical properties of the barrels. Specimens of about 1 x 1.2 x 10 mm were cut from the barrels and tested at four evenly spaced positions along the barrel. The RSA-II was operated in a temperature sweep mode, ranging from -35 [degrees] C to 35 [degrees] C, 30 rad/s frequency, and 300 g static force (tension mode). Dynamic strain was set at 0.001%.
Thickness. The wall thickness of the barrel was measured, using calipers, at the positions near the gate. A ratio of maximum to minimum wall thickness was used to indicate uniformity of the barrel wall thickness. A ratio of 1.0 indicated uniform wall thickness around the barrel.
Morphology. The purpose of the injection-spin process was to produce double-cross-orientation morphology in cylindrical articles. Figure 2 shows the laminate structure of the barrel that was made in the injection-spin process. The fracture surface of the core layer was very rough. Domains protruding out or sunken in represent traces of crystalline texture in which many molecular chains were oriented in the hoop direction. When the barrel was fractured, connections between crystal textures were broken. Thus, some hollow and protrusive spots were left on the fracture surface. However, the iPP-N sample showed very little difference in the SEM micrograph. We believe a multilayer structure was formed in the injection-spin process. Orientation in each layer was perpendicular to the adjacent layer.
X-ray diffraction illustrated a significant difference between spun and no-spun samples Figures 3a and 3b are generated from azimuthal plots of X-ray intensity patterns with an arbitrary baseline and one-quarter (0 [degrees] -90 [degrees]) azimuthal scan. The 040 plane reflection represents the b-axis orientation of the crystal of polypropylene. Figure 3a shows that spinning led to a rotation of this crystal axis. Without the spin, the b-axis was oriented to the circumferential (transverse) direction; with the spin, b-axis orientation in the machine direction increased. In general, orientation decreased from tip to flange in the barrel, as indicated by the azimuthal intensity pattern's becoming more circular.
TABLE 1. Orientation Functions Calculated from X-ray Diffraction. [f.sub.MD] [f.sub.TD] [f.sub.ND] near tip No-spin 0.3031 -0.0487 -0.2544 Spun 0.2840 0.1194 -0.4034 middle No-spin 0.4424 -0.2838 -0.1586 Spun 0.2971 0.1639 -0.4610 near flange No-spin 0.2361 0.0424 -0.2785 Spun 0.2272 0.2352 -0.4624
The c-axis orientation (macromolecular chain) was determined using Wilchinsky's equation and data of 040 and 110 reflection plane. Orientation function [f.sub.MD] in the machine direction was calculated with a reference angle equal to zero in the transverse direction. Orientation function [f.sub.ND] in the normal direction was obtained by the equation [f.sub.MD] + [f.sub.TD] + [f.sub.ND] = 0. The range of the orientation function is from -0.5 to 1. The three orientation functions are all equal to zero if the specimen is randomly oriented. If the specimen is perfectly uniaxially oriented, one orientation function is equal to 1 and the other two are equal to -0.5. To evaluate the effect of core rotation on orientation function, we plotted [f.sub.MD] against [f.sub.TD], as shown in Fig. 4. The iPP-N sample demonstrated very small or negative [f.sub.TD] and relatively large [f.sub.MD]. This means that the c-axis of the crystals was oriented, by and large, in the machine direction. The iPP-S sample showed considerable increase of [f.sub.TD] and slight decrease of [f.sub.MD], indicating that the injection-spin process intensified orientation in the transverse direction without losing much orientation in the machine direction. Orientation functions are listed in Table 1.
Measured [f.sub.MD] and [f.sub.TD] were averages of orientation function over the two skin layers and core layer. Because [f.sub.TD] was very small or negative in all layers in iPP-N samples, we expected that introducing orientation in the core layer could lead to a considerable increase in [f.sub.TD]. Rotation of the core changed the orientation in the core layer from a random state to the transverse direction. However, rotation did not change the orientation in the skin layer. Therefore, [f.sub.MD] dropped only slightly in iPP-S samples.
TABLE 2. Tensile Property Ratios of iPP-S to iPP-N. Property TD MD Yield energy 1.226 1.026 Yield stress 1.038 0.960 Yield strain 1.140 1.011 Modulus 0.998 0.970
Equal-biaxially oriented material does not show a birefringence pattern, because no preferential principal direction exists in which polarizing light transmits much faster or slower than in the other direction. The injection-spin-made barrels did not show a birefringence pattern because double-cross-orientation, although not biaxially oriented, represented an average of layers having orientation perpendicular to one another.
The injection-spin process eliminated the weld line completely. Because the weld line is likely to have the weakest strength in the article, its elimination is expected to improve mechanical performance.
Mechanical and Dynamic Performance. Rotation of the core pin generated a drag flow of the polymer between the two frozen skin layers. The drag flow stretched the polymer chains in the circumferential direction such that the strength in that direction was intensified. Tensile tests showed that yield energy--defined as integrated stress over strain up to the yield point--in the circumferential direction increased by 20% as a result of rotation. Improved mechanical performance was attributed to drag-flow-induced orientation in the core layer.
Results of dynamic mechanical tests are shown in Fig. 5. The loss tangent, which is independent of the geometry of a specimen, increased by a significant amount as a result of core rotation. Because larger loss tangent is often associated with more favorable impact properties, we expect the barrels molded in the injection-spin process to better resist cracking from impact.
Wall Thickness Uniformity. Wall thickness of the syringe barrel was 1.14 mm. with maximum azimuthal variation of 0.12 mm in a normal injection process. In contrast, the injection-spin process led to a drop of circumferential variation of thickness to 0.02 mm. six times better than the normal process. We believe the core rotation contributed to balancing of the wall thickness. The thickness variation caused by core deflection was reduced.
The "injection-spin" process created a multilayer structure with double-cross-orientation in cylindrical articles. As suggested by morphological analyses using SEM and X-ray diffraction, a drag-flow-induced orientation in the circumferential direction was generated in the core layer. The process eliminated the weld line and reduced variation in wall thickness. As a result of this process, the article's yield energy in the circumferential direction increased by 20%, and the article's loss tangent was significantly greater.
The authors thank Dr. Roger Phillips, Himont Inc., for conducting X-ray diffraction measurement, and Professor Robert Samuels, Georgia Institute of Technology, for his help in analyzing data of X-ray diffraction. The authors also thank Mr. Suresh Roy for conducting dynamic mechanical measurements.
1. E.S. Clark, Plastics Engineering, March 1974, p. 73.
2. R.J. Samuels, Structured Polymer Properties, John Wiley & Sons, New York (1974).
3. J. Trotignon and J. Verdu, J. Appl. Polym. Sci., 34, 1 (1987).
4. M. Fujiyama and T. Wakino, J. Appl. Polym. Sci., 35, 29 (1988).
5. Z. W. Wilchinsky, J. Appl. Phys., 31, 1969 (1960).
|Printer friendly Cite/link Email Feedback|
|Author:||Cao, Bangshu; Shepard, Thomas (English clergy); Tollefson, Norris; Sugg, Harry|
|Date:||Nov 1, 1994|
|Previous Article:||Additives annual '94.|
|Next Article:||Materials: hydrolysis-resistant thermoplastic polyesters.|