Axial-Feed thermoforming of oriented polypropylene tubes.

INTRODUCTION

The main objective of the manufacturing process is to produce parts that satisfy design specifications at the most economical cost possible. The mechanical properties of the final thermoplastic part can be improved by modified structure via plastic deformation in a solid state. Numerous common methods exist for this purpose such as uniaxial and biaxial drawing, rolling, die drawing, combination of drawing with rolling, and solid state extrusion (1-5). The ram extrusion process is a process that allows for the achievement of uniaxial orientation (6). In this process, an isotropic polymer billet at a temperature close to the melting temperature is forced by a piston through a die of smaller cross-section. In this study, a combination of ram extrusion and drawing processes were used to produce tubes that have high modulus and high strength depending on the draw ratio (DR) (i.e., the ratio of initial cross-sectional area to the final product).

A brief description of metal tube hydroforming and polymer sheet thermoforming processes is presented below. Tube hydroforming is a metal forming process that has been used successfully in the last 10 years to produce many automotive components (7-13). During the hydro-forming process, a tube is placed in the forming die cavity and sealed at both ends. Axial force (end feed) is then applied at both ends of the tube as well as an internal pressure from a fluid (e.g., oil, gas. or water) inside the tube. As a result of the internal fluid pressure, the tube expands and deforms to the desired shape controlled by the geometry of the die-cavity (14). In sheet thermoforming process, a sheet is heated until it becomes pliable and subsequently deformed into the mold by an applied pressure, vacuum, a moving plug, or a combination of these. There are many types of thermoforming processes, the most common among them are vacuum forming, pressure forming, drape, and plug assist (15).

Axial feed thermoforming process (AFTF) for thermoplastic tubes is a new process recently developed at McMaster University (16), (17). This process is similar to metal tube hydroforming, except that a thermoplastic tube is heated to a softened state (as in sheet thermoforming process) and subsequently deformed by a combination of applied internal pressure and axial feeding from the two ends. The process has advantages over conventional metal tube hydroforming due to relatively low cost of the dies and for the forming machines than metal tube hydroforming arising from significantly lower pressure and force requirements for deformation of polymers at higher temperatures. Also, the advantage over polymer sheet thermoplastic includes ability to feed the tube into the die that gives a close approximation of the die shape and a more uniform thickness distribution.

The mechanical properties and deformation mechanisms of biaxially oriented polypropylene sheet produced by the hydrostatic extrusion process has been examined in the past (18). Also, the tensile behavior of unfilled polypropylene and a 40 wt% talc-filled polypropylene, prepared from injection-molded square plates, has been examined at different temperatures and strain rates (19). Most of previous experimental work in similar processes, for example, in stretch blow molding, has been based on unoriented thermoplastics, where biaxial orientation is achieved after the blowing process (20). Although a few studies are reported in literature on thermoplastic tube expansion, there are no published studies of oriented thermoplastic tube expansion during AFTF process. Taraiya and Ward (21) described methods for the production of biaxially oriented polyethylene tubes and studied their resulting mechanical and structural properties. All tests were carried out at room temperature, and the tubes were pressurized using water. Microstructural changes that occur in polypropylene as a result of solid state tube extrusion or subsequent thermoforming of the oriented polypropylene tube (OPP) have not been studied in the past. These changes are expected to affect in-service performance of polypropylene components and hold the key to optimizing the extrusion and AFTF process.

In this article, many of the mechanical, forming, and postforming characteristics of extruded and thermoformed OPP materials are presented in terms of process conditions and resulting microstructure. The goal of this study is to understand how the microstructure and mechanical properties of oriented tube relate to the expansion during AFTF.

EXPERIMENTAL

Solid State Extrusion

The starting tubular polypropylene billets were machined from solid cylindrical billets received from Polymer Sheet Applications Company (PSAC) in Guelph, Ontario, Canada. The solid cylindrical billets were produced in the melt state and cooled directly in a water-cooling bath. The tubular billets had a nominal inner diameter of 1.0 in. The outer diameter was machined to various sizes depending on the desired DRs. Solid-state extrusion experiments to produce OPP tubes from tubular billets were carried out at PSAC. To achieve good sealing of the ends during thermoforming, available extrusion dies were modified to produce the exact size (inner and outer diameter) with a clearance to match the plugs in the open tube bulge testing system utilized for AFTF as described in the next section.

Axial Feed Tube Thermoforming (AFTF)

An axial-feed, die-less, tube bulge testing system was designed, fabricated, commissioned, and then used to conduct forming tests on OPP tubes. The test system was built around a hydraulically powered 25-kip mechanical test frame from MTS. This frame is equipped with two vertically mounted servo-hydraulic actuators connected to a PC-based controller and data acquisition system. In this system, the tube was placed between the upper and the lower taper plugs attached to the actuators. Each end of tube was then clamped by two split half clamps. The middle of tube was then kept free (unsupported) and open to view to allow for continuous observation of expanding tube using an on-line ARAMIS optical strain measurement system (Fig. 1). For this purpose, test specimens were painted with water-based ink of white and black random pattern speckles by using an air brash. The dimensions of the workpiece and final bulged part are shown schematically in Fig. 2. The tube was heated from inside as well as outside to a forming temperature of 150[degrees]C by circulating hot oil through the tube and by blowing hot air on outer surface of the tube. The lower plug was kept stationary, while the upper plug (attached to the upper actuator) was moved downward against the top end of the tube, thereby creating an axial force on the tube for end sealing as well as to supply additional tube material to the forming region (so-called axial-feed). A series of bulging tests for various combinations of internal pressure and axial feed (so-called loading paths) were conducted while continuously recording the bulging process with the ARAMIS system (22). This system is based on digital image correlation method and has been described elsewhere (23-25). The parameters such as internal pressure, axial displacement (or feed), axial strain, and hoop strain were measured continuously during bulging.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

The input loading paths for the internal pressure and axial feeding displacements were controlled by PC-based test controller attached to the hot oil tube bulging system. Four different load paths (A. B. C, and D) with the same axial feed were utilized as shown in Fig. 3. For each of these cases, the internal pressure was applied simultaneously with the axial feed during the bulging test up to the maximum pressure.

[FIGURE 3 OMITTED]

Sample Preparation and Physical Techniques Used

To evaluate the morphological changes in OPP tubes, cryomicrotomy was performed at low temperatures (-100[degrees]C) to produce thin sections parallel to extrusion direction for examination in a polarizing microscope at high magnifications. For scanning electron microscopy (SEM) experiments, field emission scanning electron microscope (FE-SEM JEOL JSM-7000F) with lower voltage, which lessens damage to specimen surfaces, was utilized. Samples of thickness ~0.5 mm from extruded tube, perpendicular to the extrusion direction, were cut using a rotary steel cutter with cooling water. Each sample was dipped in liquid nitrogen and then fractured along the extrusion direction (sec Fig. 4). The fractured surface was coated by a very thin layer of gold using a sputter coater (S150B) and then observed in FE-SEM under a low voltage. The specimens for X-ray technique were cut out of extruded tubes and from the middle of the bulge area of thermoformed tubes in the extrusion direction; the size of these fibers was about (0.5 mm X 0.5 mm), as shown in Fig. 5. X-ray diffraction system, operating at 50 kV and 90 mA, with a Cu Ka ([lambda] = 1.54 [Angstrom]) line was utilized. The diffraction experiments were conducted in the transmission mode and the output signal was recorded on a flat CCD area detector at a distance of 50 mm. The specimen was mounted on an automated single crystal orienter. A rotation of [empty set] (0[degrees] - 360[degrees]) was made with 5[degrees] intervals.

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

A complete picture of the distribution of crystalline orientation within a sample was obtained through pole figure analysis. The results of pole figure measurements were plotted, with interpolation, by using GADDS Software (Bruker-AXS) (26). The preferred orientation was recorded using the Debye patterns.

Sample Preparation and Test Conditions for Tensile Tests

Straight and bulged tubes with a DR of 6.3 produced via solid state ram extrusion were used for carrying out room temperature uniaxial tensile tests. The outer and inner diameters of the tubes were 35 and 25 mm, respectively. All tensile samples for testing were machined from the wall regions, prior to and after thermoforming as shown in Fig. 6. The width of the gauge section of the tensile samples could be no larger than the tube wall. Therefore, specimens gauge length and width of 3 and 2 mm, respectively were chosen. Uniaxial tensile tests were carried out at a cross-head speed of 1 mm/min (equivalent to a nominal strain rate of 6 X [10.sup.-3]/s) on a MTS 250 kN servo-hydraulic test machine. The load versus displacement curves were continuously recorded during the tests. The strain measurements during the tension tests were also made using the ARAMIS system, continuously recording images during the tests. The ruptured tensile samples were also collected and mechanically polished to observe the possible damage using an optical microscope.

[FIGURE 6 OMITTED]

RESULTS AND DISCUSSION

Morphological Changes From Solid State Extrusion

The morphological changes in the unoriented PP billet and in extruded tube at a DR of 6.3 in terms of FE-SEM images are shown in Fig. 7. The fibrils were clearly visible on fracture surfaces and aligned to the extrusion direction. In a polarizing microscope, a spherulite structure was observed for unoriented PP billet, consisting of a combination of both crystalline and amorphous regions. Figure 8 shows polarizing microscope images of the spherulitic morphology of PP billet and OPP tube at DR = 6.3. The structure of the extruded samples at DRs 5 and higher was completely changed from spherulites structure to an oriented structure (fibrils). This observation of the transformation from spherulitic to fibrous structure is in agreement with other researches (27).

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

Diffractions associated with all Debye rings were identified from the inner ring to the outer ring as originating from crystallographic planes (110)[alpha], (040)[alpha], (130)[alpha], and (lll)[alpha] + (041)[alpha] in the billet, see Fig. 9. The diffraction pattern revealed full rings in the billet that changed to strong peaks in extruded samples. The (110)[alpha] pole figures for the billet and extruded tubes were plotted in stereographic projection as shown in Fig. 10 using GADDS software. In billet sample, the pole figure showed a random orientation around the extrusion direction. In extruded sample, however, the pole figure showed approximately uniaxial orientation; the width of the (110)[alpha] pole became narrow with the increase in the DR and was distributed symmetrically around the extrusion direction.

[FIGURE 9 OMITTED]

[FIGURE 10 OMITTED]

Axial Feed Tube Thermoforming (AFTF)

In AFTF experiments, an increase in axial feed resulted in higher formability (bulge height) and delayed failure (thinning and/or bursting) (see Fig. 11). When axial feed was applied after a certain level of pressure (to prevent buckling or wrinkling) the bulge height increased rapidly during the feeding stage, depending on the amount of feeding and feed rate. This is because more material was fed into the expansion zone to reduce the severity of the deformation. A balance of axial feed and internal pressure was required to obtain a good shape of bulge, free of failure. Excessive feeding of material in the bulge zone led to wrinkling or buckling failure. Figure 12 shows the results in terms of internal pressure versus axial feed displacement curves from a series of bulging tests for a DR of 6.3 with different loading paths (A, B, C, and D) (see Fig. 3). The tests were conducted at a forming temperature of 150[degrees]C and a pressure rate of 20-65 kPa/s. With the same amount of the axial feed of 18 mm and at different maximum pressure, cases (A) and (D) resulted in bursting and wrinkling respectively, while cases (B) and (C) resulted in uniform bulges.

[FIGURE 11 OMITTED]

[FIGURE 12 OMITTED]

Morphological Changes From AFTF Process

Figure 13 presents the results of the (110)[alpha] and (040)[alpha] pole figures of the bulged samples at a range of axial feeds. With no axial feed, the b-axis of the crystals shows a concentration in the transverse direction with a slight difference as compared to the extruded samples. However, the bulge samples exhibited broadly distributed intensity with increase in axial feed. With the axial feed of 8 mm, (110)[alpha] and (040)[alpha] poles are distributed around the transverse direction in a band making an angle of about 30[degrees]. At a large axial feed of 18 mm, the (110)[alpha] and (040)[alpha] poles are distributed in a broad band making an angle of about (35-60[degrees]) to the extrusion direction with a high concentration in the transverse direction making an angle of about 72[degrees] to the normal direction in the ND-TD plane.

[FIGURE 13 OMITTED]

Uniaxial Tensile Behavior of OPP Tube After Thermoforming

Figures 14 and 15 show the true stress-strain behavior of the tensile samples machined from the middle of the extruded and bulged tubes with a draw ratio of DR = 6.3 in the extrusion and hoop direction and tested at room temperature. Extruded and bulged samples without axial feed in the extrusion direction exhibited brittle failure and almost the same tensile strength at brake point at a true strain of 0.10. With increasing axial feed the curve started with initial deformation, which is elastic followed by yielding and a region of plastic deformation. The material strength and stiffness for 18-mm axial feed are significantly higher than that for an axial feed of 8 mm.

[FIGURE 14 OMITTED]

[FIGURE 15 OMITTED]

The stress-strain curves in the hoop direction exhibited trends similar to the axial direction, but with reduced strength and ductility (see Fig. 15). As earlier, the material strength and stiffness for 18-mm axial feed are improved compared with that for 8. In general, increasing axial feed of the bulged tube material exhibits ductile behavior, failing after hardening and extensive deformation.

In Fig. 14, the samples with axial feed show deviation from the apparent elastic behavior, lower yield stress and early occurrence of fracture compared to the samples from extruded tube and those from tests done with no axial feed. This is because in the as extruded and those bulged with no axial feed, the reorientation of the fibers is significantly smaller. Most of the fibers are still aligned to the extrusion direction that results in increased elastic response and higher stress at fracture. On the other hand, a significant reorientation of fibers, away from the extrusion direction, occurs as a result of axial feeding. Clearly, the axial feed aides in reorientation of the fibers by providing addition transverse (hoop) strain component as reflected in Fig. 15. The interpretation also consists with the results shown in Fig. 13, which clearly shows a change in the pole figure from being distributed symmetrically around the extrusion direction to broadly distributed intensity with axial feed.

The break point (or the end point of all stress-strain curves), referring to Figs. 14 and 15, corresponds to failure of the fibers or amorphous region between the fibers. A visual inspection of the failed oriented thermoplastic specimens showed a fibrous failure for all of the tensile samples. From the geometry of the dog-bone specimens machined from tubes along its longitudinal direction, two types of failures are suggested as shown in Fig. 16. First type of failure initiates where the fiber from the straight gauge line intersects the fiber from sample are (Type I). The fiber along the gauge line is longer than at the are portion of the tensile sample. With increasing applied tensile load, the region of localized plastic deformation leads to craze formation. Second type of failure originates along the straight gauge line where the fiber is reduced by machining (Type II). In this case, the fiber "diameter" is reduced by cutting and this results in a premature failure.

[FIGURE 16 OMITTED]

Figure 17 shows images from the gage regions of tensile samples machined from tube blank (Fig. 17a) and from the tube bulged without feeding (Fig. 17b). Fracture begins in the regions where there is localized strain. ARAMIS-based strain analysis revealed that both samples fail in a brittle manner where the fibers break at a local true strain of 0.14. However, ductile fracture is associated with test coupons that are machined from the tubes bulged with axial feeding (Fig. 17c and d).

[FIGURE 17 OMITTED]

Stress whitening shown in Fig. 17c and d corresponds to formation of fibrillar bridges in the craze region, similar to the results reported earlier for thermoplastic polymers; PE, PP, Nylon, and PMMA (28), (29). Figure 18a and b shows the distribution of epsilon Y at four points in the gage region of a deformed tensile test coupon (marked points 1, 2, 3, and 4), machined along the longitudinal direction from tube bulged with an axial feed of 8 mm. As shown, the local tensile strains are not uniform across the test coupon. The fibers are broken at point 3 at a local true strain of 0.205. The fracture process typically involved three steps. First, the fibers were broken and separated along the longitudinal direction. At local true strain of 0.269, fibrillar bridges were formed and subsequently crazing occurred at local true strain of 0.376. As the deformation proceeded, the crack began to propagate and the test coupon twisted at a true strain of 0.515. For tensile samples with an axial feeding of 18 mm, the damage mode was principally crazing (Type I). For this test, the fiber was broken at point 2 at a local true strain of 0.301. From these tests, we can conclude that the damage occurs at an equivalent true strain of 0.2 and 0.324, respectively for samples that were machined from tubes bulged with axial feeds of 8 and 18 mm, respectively. However, the damage occurs at an equivalent true strain of 0.06 and 0.07, respectively for samples machined from the as-received and bulged tubes without end feeding. Figure 19 shows results similar to Fig. 18 for test coupons cut along the hoop directions from 8-mm axially fed tubes. Failure of the specimen occurred at an equivalent true strains of 0.06, 0.08, 0.11, and 0.13, respectively for samples that were machined from tube blank, from the tubes bulged without end feed and with axial feeds of 8 and 18 mm. The results demonstrate significant improvement in the formability of the tubes with axial feeding during bulging.

[FIGURE 18 OMITTED]

[FIGURE 19 OMITTED]

CONCLUSIONS

In this article, the mechanical properties and the characteristics of microstructures of oriented thermoformed polypropylene tubes through AFTF are investigated and compared with the extruded products through solid state extrusion. The effect of AFTF on structure and mechanical properties of OPP tubes has been examined. The relationship between microstructure, mechanical properties, and process parameters are presented in terms of the change in preferred orientation for PP billet, extruded and bulged tubes for different load paths.

The spherulitic structure of machined tubular PP billet is completely changed to fibrillar after its solid state extrusion into an oriented PP tube. In terms of the process characteristics, an increase in axial feed during thermo-forming resulted in better formability, that is, increased bulge height, and delayed failure by thinning or bursting. On-line 2D strain maps from room temperature uniaxial tensile tests on thermoformed OPP samples showed that failure of the fibers occurred at strains between 0.06 and 0.08 for samples without the axial feed. On the other hand, with increased axial feed, the material exhibited higher failure strains of 0.324 and 0.13 in both axial and hoop directions, respectively. The axially fed material also exhibited enhanced stiffness and toughness.

ACKNOWLEDGMENTS

The authors thank Dr. Frank Maine of PSA Composites Inc. for supplying the PP material and for help in conducting solid-state extrusion experiments on their extrusion equipment.

REFERENCES

(1.) Z. Bartczak, J. Morawiec, and A. Galeski, J. Appl. Polym. Sci., 86, 1413 (2002).

(2.) I.M. Ward and D.W. Hadley. Introduction to Mechanical Properties of Solid Polymers, Wiley, New York (1993).

(3.) I.M. Ward, P.D. Coates. and M.M. Dumoulin, Eds., Solid Phase Processing of Polymers. Hanser, Munich (2000).

(4.) A. Ciferi and I.M. Ward. Eds., Ultra-High Modulus Polymers, Applied Science, London (1979).

(5.) A. Ajji and M.M. Dumoulin, J. Appl. Polym. Sci., 102, 3391 (2006).

(6.) C.J. Farrell and A. Kelleer, J. Mater. Sci., 12, 966 (1977).

(7.) E. Chu and Y. Xu. Int. J. Mech, Sci., 46, 263 (2004).

(8.) A. Kulkarnia, P. Biswasa, R. Narasimhana. A.A. Luob, R.K. Mishrab. T.B. Stoughtonb, and A.K. Sachdev, Int. J. Mech. Sci., 46, 1727 (2004).

(9.) B. Carleer, G. van der Kevie, L. de Winter, and B. van Veldhuizen, J. Mater. Process. Technol., 104. 158 (2000).

(10.) M. Imaninejad, G. Subhash, and A. Loukus. Int, J. Mech. Sci., 46. 1195 (2004).

(11.) G. Nefussi and A. Combescure, Int, J. Mech. Sci., 44, 899 (2002).

(12.) C.L. Chow and X.J. Yang, J. Mater, Process, Technol., 130-131, 107 (2002).

(13.) M. Brunet, S. Boumaiza, and G. Nefussi, J. Mater, Process, Technol., 148, 269 (2004).

(14.) N. Asnafi and A. Skogsgardh, Mater. Sci. Eng., A279, 95 (2000).

(15.) C. Maier and T. Calafut. Polypropylene: The Definitive User's Guide and Databook, Plastics Design Library, Norwich. NY (1998).

(16.) E. Azhikannickal, M. Jain, and M. Bruhis. Polym, Test., 26, 195 (2007).

(17.) E. Azhikannickal, M. Jain, and M. Bruhis, Int, J. Mech. Sci., 50, 1065 (2008).

(18.) S.J. Pan, H.I. Tang. A. Hiltner. and E. Baer, Polym. Eng, Sci., 27, 12 (1987).

(19.) Y. Zhou and P.K. Mallick. Polym. Eng, Sci., 42, 12 (2002).

(20.) D.V. Rosato, Ed., Blow Molding Handbook, Hanser, New York (1989).

(21.) A.K. Taraiya and I.M. Ward, J. Appl. Polym, Sci., 59, 627 (1996).

(22.) GOM Ltd, ARAM1S v4.7 Manual, GOM, Braunschweig, Germany (2001).

(23.) M.A. Sutton, W.J. Wolters. W.H. Peters, W.F. Ranson. and S.R. McNeil, Image Vis. Comput., 1(3), 133 (1983).

(24.) X. Duan, M. Jain, M. Bruhis, and D.S. Wilkinson, Adv, Mater. Res., 6-8, 737 (2005).

(25.) Q. Situ, M. Bruhis, and M. Jain, The 6th International Conference and Workshop. Detroit. Michigan. USA, 0715, 131 (2005).

(26.) K.L. Smith, Z. Jin, J.L. Chambers. B. He, U. Preckwinkel, and P.D. Moran. Bruker Analytical X-ray Systems. GADDS for WNT 4.1.10, Copyright 1997-2007, Bruker AXS.

(27.) Z. Stojanovic, Z. Kacarcvic-Popovic, S. Galovic, D. Milicevic, and E. Suljovrujic, Polym, Degrad, Stab., 87. 279 (2005).

(28.) W.D. Callister Jr., Material Science and Engineering: An Introduction, Wiley, New York (2007).

(29.) M. Ashby, H. Shercliff, and D. Cebon, Materials Engineering, Science, Processing and Design, Elsevier, Oxford (U.K.) (2007).

Mohamed Elnagmi, Moisei Bruhis, Mukesh Jain

Department of Mechanical Engineering, McMaster University, Hamilton, Ontario, Canada L8S 4L7

Correspondence to: Mukesh Jain; e-mail: [e-mail]

Contract grant sponsors: Natural Sciences and Engineering Research Council (NSERC) of Canada. Mr. Earlby Wakefield of Decoma International Inc.

DOI 10.1002/pen.21316

Published online in Wiley InterScience (www.interscience.wiley.com).