Printer Friendly

Thermoforming thermoplastic polyurethanes.

By combining a simple laboratory procedure with rheological measurements, the optimal thermoforming conditions can be determined and correlated with material properties.

Thermoplastic polyurethanes (TPUs) find applications in injection molding, thermoforming, and extrusion of profiles, blown film, and sheet. For many applications, conversion of TPUs into parts by thermoforming provides an attractive alternative to injection molding and extrusion, as thermoforming offers relatively low capital costs and high productivity. A variety of specialized applications, such as food packaging, high-tech electronics, automotive and aerospace, utilize this technique with other thermoplastics.

In thermoforming, a material usually in sheet or film form is shaped after it has been softened. While the technology of the mechanical portion of the process is well developed and understood, many trials must be performed in order to understand the material contribution to the process and compensate for it. A direct measurement of material behavior in situ during a thermoforming cycle is extremely complex, owing to the difficulty of instrumenting commercial equipment for accurate reading and control of temperature and stress. Consequently, thermoforming is often viewed as an art rather than a science.

Past studies of thermoforming focused on amorphous resins such as PMMA, PS, and ABS, and crystalline resins such as PET, PP, and PE. These studies concluded that the best thermoforming conditions are attained when the material is in its rubbery solid state above its glass-transition temperature ([T.sub.g]), but below its crystalline melting temperature. However, none of these investigations included thermoplastic polyurethanes. Unlike the other resins, TPUs are two-phase systems in which one phase is hard and solid at room temperature while the other phase is soft and elastic. Although TPUs can exhibit a wide rubbery plateau, ranging from about -40 [degrees] C to about 100 [degrees] C, this plateau region may not be suitable for thermoforming, as materials must flow in order to be shaped irreversibly. TPUs will flow only when the hard phase is softened or melted, which takes place some 150 [degrees] C to 200 [degrees] C above the [T.sub.g]. In comparison, normal forming temperatures are usually 70 [degrees] C to 100 [degrees] C above the [T.sub.g].

The objective of this study was to identify the material parameters important to thermoforming TPUs. The thermoforming process was simulated in the laboratory using the experimental approach developed by Throne(1) (see the Box). Dynamic mechanical analysis (DMA) was used to obtain the viscoelastic parameters of a TPU over a wide weight-average molecular-weight ([M.sub.W]) range. Correlating the thermoforming experiments with the viscoelastic measurements provided insight into the material behavior and material parameters influencing the thermoforming process, allowing the material parameters governing thermoforming to be identified and quantified.


This study used a polyester and MDI-based TPU having a [T.sub.g] of the soft segment at about -35 [degrees] C and a melting peak of the hard segment at about 180 [degrees] C. The thermal characteristics of the TPU, as determined using a Mettler TA-3000/TC-10A differential scanning calorimeter (DSC) at a 10 [degrees]/min heating and cooling rate, are shown in Fig. 1. Ten different molecular weights of the TPU were evaluated. The products with their respective [M.sub.w]s as determined by gel permeation chromatography (GPC) are shown below.
Sample ID Mw

TPU 21 212,000
TPU 27 168,000
TPU 33A 162,000
TPU 30 152,000
TPU 40 152,000
TPU 33B 148,000
TPU 35 132,000
TPU 43 127,000
TPU 45 123,000
TPU 54 115,000

Temperature-dependent viscosities were measured on a Rheometrics Dynamic Mechanical Spectrometer equipped with oscillating parallel plates. An angular frequency of 10 Hz was used to provide a cyclic shear deformation. The extruded or compression-molded samples were dried in a vacuum oven at 105 [degrees] C for more than 2 hrs and then loaded into the parallel-plate fixture at 170 [degrees] C. The compartment was then purged with dry nitrogen to reduce the temperature to 140 [degrees] C and the viscoelastic properties were measured as a function of increasing temperature.

The material parameters determined from these measurements are G[prime], G[double prime], [[Eta].sup.*], and tan [Delta]. G[prime], the shear modulus, is the elastic-response component associated with the tendency of the material to remember its predeformation dimensions. The loss modulus, G[double prime], is the energy-dissipative-response component and is associated with flow during deformation. Tan [Delta] describes the relative response of the two components: when tan [Delta] [much less than] 1, the material behaves as an elastic solid; and when tan [Delta] [greater than] 1, the material behaves as a viscous fluid. [[eta].sup.*], the complex viscosity, is similar to the steady-state shear viscosity as measured, for example, by capillary rheometry.

Effect of Temperature on Thermoformability

Attempting to gain insight into material properties critical to thermoforming, a number of investigators have studied material behavior in stretching modes. Most researchers have measured tensile strength near the thermoforming temperature at constant strain rates(2-5) or the material creep.(6, 7) These measurements yielded, however, uniaxial stretching data, while actual thermoforming is a biaxial stretching process. Some investigators cautioned against using uniaxial data to predict biaxial performance.(1, 8) Others cautioned also against predicting material performance in an actual draw-down situation from the above tests.(9)

Throne's biaxial-stretching laboratory test (see the Box), used in this study, more closely mirrors the thermoforming process. In drawing a flat sheet into a simple 60 [degrees] cone, the temperature was varied while the deformation force and the rate of deformation remained constant. The areal draw ratio, [Ra.sub.max] (Equation 1), is a measure of the thermoformability of the polymer, and the experiments at different temperatures helped to determine the forming window.

A plot of the maximum areal draw ratio (expressed as [t.sub.o]/[t.sub.cap]) as a function of forming temperature [ILLUSTRATION FOR FIGURE 2 OMITTED] shows, as expected, that the areal draw ratio is strongly temperature dependent. Materials can be drawn to a larger extent as the temperature is increased. It is also apparent from Fig. 2 that a higher [M.sub.w] material must be heated to a higher temperature to attain an extent of draw similar to that of a lower [M.sub.w] material.

Figure 2 can also be used to determine a forming window for each material. For example, the temperature window to attain a draw ratio [t.sub.o]/[tc.sub.ap] [greater than] 5 appears to be from 149 [degrees] C to 160 [degrees] C for the TPU of [M.sub.w] = 152,000, and from 162 [degrees] C to 177 [degrees] C for the TPU of [M.sub.w] = 212,000. In each case, above these temperatures the forming became unstable, i.e., during the draw-down, the sheet was frequently split or perforated in the cap region (designated as "F" on [ILLUSTRATION FOR FIGURES 7 AND 8 OMITTED]).

Effect of Temperature on Viscoelastic Properties

A typical plot of the viscoelastic parameters over the temperature range of interest is shown in Fig. 3 for a TPU of [M.sub.w] = 152,000. These plots show that as the temperature increases above 140 [degrees] C, tan [Delta] slowly rises and G[prime] and [[Eta].sub.*] decrease. This behavior is attributed to a loss of network structure. Ultimately a temperature is reached when tan [Delta] [greater than] 1, at which point the molecular ties become ineffective and the polymer behavior is dominated by its dissipative characteristics. Henceforth, the temperature at which tan [Delta] = 1 will be referred to as the solid-rubber-to-viscous-liquid (SR-to-VL) transition temperature.

A good correlation is found between the temperature at which tan [Delta] = 1 and the [M.sub.w] of the TPU [ILLUSTRATION FOR FIGURE 4 OMITTED]: the lower the [M.sub.w], the lower the transition temperature. A good agreement also exists between [[Eta].sub.*] and the [M.sub.w] [ILLUSTRATION FOR FIGURE 5 OMITTED]. The dependence of the SR-to-VL transition temperature on molecular weight may be viewed, therefore, as a dependence on polymer viscosity. Apparently, a specific viscosity must be reached for the transition to occur. To test this hypothesis, the complex viscosity at the transition temperature (tan [Delta] = 1) was plotted as a function of the transition temperature [ILLUSTRATION FOR FIGURE 6 OMITTED]. From this plot, the SR-to-VL transitions of all the materials studied occur when their viscosity is about 10(5) poise.

Viscoelastic Properties and Thermoformability

Thermoforming involves a biaxial deformation of a molten polymer. At the forming temperature, the polymer should possess a strong viscous component that allows for flow when sufficient stress is applied, and a significant elastic component to resist flow and impart integrity. As formability of a material is defined by the combination of these two properties, it is only logical to conclude that the optimal conditions for thermoforming must occur at a temperature corresponding to the material's transition from a solid-rubber state to a viscous-liquid state. Conversely, the further the forming temperature is from the transition temperature, the more difficult the forming will be.

To test this assumption, the maximum areal draw ratio was plotted against the temperature difference ([Delta]T) between the draw temperature, T(draw), and the SR-to-VL transition temperature at which tan [Delta] = 1 [ILLUSTRATION FOR FIGURE 7 OMITTED].

[Delta]T= T(draw) - T(tan [Delta] = 1) (2)

Although the data in Fig. 7 are scattered, the trend appears to be well defined - evidently, materials can be drawn to a larger extent when the sheet temperature approaches the SR-to-VL transition temperature, i.e., [Delta]T = 0.

Figure 7 also provides information on the thermoforming windows of all the TPUs studied with the exception of the TPU of [M.sub.w] = 212,000. If a shallow draw is required ([t.sub.o]/[t.sub.cap] [less than] 5), the forming window appears to be about 10 [degrees] C wide, ranging from about 4 [degrees] C below the transition temperature to about 5 [degrees] C above it. For a deep draw, ([t.sub.o],/[t.sub.cap] [greater than] 10), the forming window becomes narrower, only 2 [degrees] C to 5 [degrees] C wide. Then the forming range extends from the transition temperature to about 5 [degrees] C above it.

The results for the TPU of [M.sub.w] = 212,000 are puzzling. In contrast to the other samples, this material could be formed at temperatures significantly below the SR-to-VL transition temperature. A maximal draw ratio of [greater than] 15 was attained about 7 [degrees] C below the transition temperature. On cooling. however, these samples shrunk considerably, suggesting that the molecular ties were still present when the material was formed, possibly causing the material to retract when released. Such behavior only supports the earlier conclusion that the best condition for thermoforming is at the SR-to-VL transition temperature, i.e., [Delta]T = 0.

To quantify the optimal intrinsic material properties for forming TPUs, the complex viscosity at the temperature of the forming experiments was plotted as a function of the elastic modulus at the same temperature. Figure 8 shows that the most stable conditions for thermoforming are attained when the elastic modulus is above 7.5 x [10.sup.5] dyne/[cm.sup.2] (7.5 x [10.sup.4] Pa) and the complex viscosity is above 1.05 x [10.sup.5] poise. These properties constitute the lower limits of the viscosity and modulus windows. The upper limits can be determined from the DMA measurements at 4 [degrees] C to 5 [degrees] C below the SR-to-VL transition temperature, and are supported by the forming data in Fig. 7. Based on DMA and the thermoforming experiments, the upper limits for elastic modulus and viscosity are [10.sup.6] dyne/[cm.sup.2] and 1.4 x [10.sup.5] poise, respectively.

The optimal elastic-modulus range for forming TPUs, 7.5 x [10.sup.4] to [10.sup.5] Pa, is in good agreement with Throne's conclusion that the elastic modulus for any material at the thermoforming temperature ideally should range from 7 x [10.sup.4] to 7 x [10.sup.5] Pa.(1) An elastic modulus of [10.sup.6] Pa has been reported as the ideal modulus for forming ABS(2) and polypropylene.(7)


A nonreversible deformation and stable forming of the TPUs studied can be attained 15 [degrees] C to 40 [degrees] C below the final melting temperature (about 190 [degrees] C) of the hard-segment phase. The optimal conditions for thermoforming occur at a temperature close to the SR-to-VL transition temperature, i.e., the temperature at which G[prime] = G[double prime], or tan [Delta] = 1.

The SR-to-VL transition temperature is influenced by the [M.sub.w] of the TPU. Generally, a higher [M.sub.w] material will exhibit a higher transition temperature. Similarly, a higher [M.sub.w] TPU will require a higher forming temperature to attain the same maximum draw ratio.

From the thermoforming experiments and viscoelastic property measurements, the optimal temperature window for forming TPUs was found to range from about 4 [degrees] C to 5 [degrees] C below to 5 [degrees] C above the transition temperature.

Although these observations are based on the study of only one TPU, a brief examination of the viscoelastic behavior of other TPUs suggests that these conclusions may be general for polyether-based and polyester-based TPUs with a relatively low-to-medium degree of crystallinity of the hard-segment phase. With high-hard-segment crystalline TPUs, the recommended forming temperatures are expected to be within a few degrees below the hard-segment melt temperature.


1. J.L. Throne, Thermoforming, Hanser Publishers, Munich (1987).

2. V.E. Malpass and C.H. White, SPE J., November 1971, p. 23.

3. J. Meissner, Rheol. Acta, 8, 78 (1969).

4. J.L. White and Y. Ide, J. Appl. Polym. Sci., 22, 260 (1978).

5. D.C. Hylton and C.Y. Cheng, SPE ANTEC Tech. Papers, 34, 34 (1988).

6. J. Guignard, 26th SPI Conf., New York (1968).

7. P. Lund, SPE ANTEC Tech. Papers, 34, 496 (1988).

8. L.R.G. Treloar, The Physics of Rubber Elasticity, Oxford University Press, Britain (1958).

9. S. Middleman, Fundamentals of Polymer Processing, McGraw-Hill, New York (1977).

RELATED ARTICLE: Determination of Thermoformability

In thermoforming, the maximum areal draw ratio, [Ra.sub.max], defined as the ratio of the initial sheet area to the area of the cap, is an important design parameter, as it is a direct measure of the ability of a polymer to be drawn into a three-dimensional corner. J.L. Throne(1) showed that in the thermoforming of a simple straight-sided cone with a cone angle of [pi]/3 (60 [degrees]), [Ra.sub.max] can be calculated from the ratio of initial sheet thickness to the cap thickness, [t.sub.o]/[t.sub.cap], using the following equation:

[Ra.sub.max] = 3/4 ([t.sub.o]/[t.sub.cap] (1)

In the setup shown in the Figure, the flat sheet was held in a book frame clamped together with vise grips. The outside frame dimensions were 5 x 8 inches, and the inside dimensions 3 x 6 inches. A commercial metal funnel with an angle of about 60 [degrees] was used. Vacuum was supplied by a commercial shop vacuum cleaner attached to the bottom of the funnel. Typically, such vacuum cleaners develop a pressure of 100 to 130 mm Hg. This was sufficient to draw most of the sheet into the funnel.

The flat sheet, preheated up to the desired temperature, was placed for 3 min into a standard forced-air convection lab oven equipped with an adequate temperature controller. Sheet temperature was determined with an infrared thermometer. Immediately upon removal from the oven, the hot sheet was placed against the funnel rim while the vacuum was activated, and held in this position for 30 to 60 sec until it cooled off. Then the vacuum was shut off, and the framed sheet was removed from the funnel and the sheet from the book frame. The cap was then split and its thickness was measured.

Initial sheet thicknesses were typically 0.055 to 0.070 inch, and the precision of the thickness measurements was [+ or -]0.0005 inch. Three or four experiments were run for each material over a range of 140 [degrees] C to 170 [degrees] C increments.
COPYRIGHT 1995 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1995 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:includes related article
Author:Eckstein, Yona; Jackson, Robert L.
Publication:Plastics Engineering
Date:May 1, 1995
Previous Article:Switching from metal to plastic components requires value-added capabilities.
Next Article:Annual report 1994.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters