HDPE tensile bars with Type I dimensions as specified in ASTM D 638 were cast via injection molding. 185˚C was taken as the temperature used in the melting zone during injection molding (BOY 22S, Dipronic).

Tensile tests were performed according to earlier studies [19] [20] , with the setup being shown in Figure 1. In brief, the Instron tensile testing machine was used in the range of 0 - 2 kN. The maximum load and the drawing force were examined over a range of cross head speeds (5, 50, 80, 100, 200 and 500 mm/min). The load-exten- sion curve was obtained for each of the cross head speeds adopted. Changes in temperature were examined by slightly touching the drawing tip during the process of cold drawing. The cross-sectional area before and after the cold drawing process was recorded by using a caliper.

The hardness values of the undrawn specimens, as well as those of the necking regions, were examined by using a durometer at ambient conditions according to the ASTM D2240 method as reported in an earlier study [21] .

Quantitative data were obtained in triplicate. They were reported as means ± standard deviations, where indicated. Student’s t-test was performed to assess the statistical significance. Differences with p-value < 0.05 were considered to be statistically significant.

In this study, six cross head speeds (5, 50, 80, 100, 200 and 500 mm/min) were examined. The HDPE specimens cold-drawn at different cross head speeds were shown in Figure 2. During cold drawing, necking of the polymer was observed. While the applied force remained almost constant during polymer stretching, the cross-section of the necking region became less than that of the remaining undrawn portion of the polymer. This was confirmed by Figure 3, in which the cross-sectional area of the cold-drawn specimen decreased significantly. During the analysis above, the specimen cold-drawn at 5 mm/min was not taken into account. This was because at that cross head speed, failure of the specimen could not be reached even at the limit of stretching applicable by the tensile testing machine, leading to failure of determination of the final cross-sectional area in that case.

During cold drawing, we found that at the cross head speed of 500 mm/min, the temperature at the necking region increased. The generation of heat was hypothesized to be due to the inherent energy loss during plastic deformation. When the necking region was touched by fingers, there was an increase in the load experienced by the specimen. Such an increase in load disappeared once the fingers were removed, and the load experienced by the specimen returned to the original level (Figure 4(a)). Similar observations were found at the cross head speeds of 50, 80, 100 and 200 mm/min. Contrary to the situation above, at the cross head speed of 5 mm/min, the strain rate was low, and the temperature at the necking region decreased. This was proposed to be due to the dissipation of energy under the extremely low strain rate. Moreover, the load experienced by the specimen reduced when the necking region was touched by fingers, though such a decrease disappeared when the fingers were removed (Figure 4(a)).

Based on the extension-load curves in Figure 4(a), the stress and strain experienced by the specimens during cold drawing were determined. The engineering stress (σ) and engineering strain (ε) were obtained by equations (1) and (2), where F is the magnitude of the applied force, A is the cross-sectional area of the specimen, l₀ is the initial length of the specimen, and Δl is the change in the specimen length:

Based on the engineering stress and engineering strain, the true stress (σ_T) and true strain (ε_T) were obtained by equations (3) and (4). The true stress-true strain curves were plotted in Figure 4(b), in which the yield strength tended to increase as the cross head speed increased. This suggested that the yield strength was positively related to the strain rate.

Semi-crystalline polymers, as well as amorphous polymers, generally have the yield behavior described by the Eyring’s viscosity theory [22] . Previous experiments demonstrated that under specific testing conditions, the

ratio of yield stress and absolute temperature is in a linear relationship with the natural logarithm of the strain rate [23] [24] . Based on this relationship, the activation volume (V₀) of the polymer can be determined by Equation (5), where C is a constant, T is the absolute temperature, ε₀ is activation energy, is the cross head speed, σ is the yield stress, and K is the Boltzmann’s constant:

By rearranging the equation, we have:

The plot of the yield stress against the natural logarithm of the cross head speed was presented in Figure 5. According to Equation (6), the slope of the plot is equal to 2KT/V₀, from which V₀ was approximated to be 4.98 × 10⁻²⁷ m³. As this activation volume is much larger than a single repeating unit, this suggests that the yield in HDPE may involve the cooperative motion of multiple repeating units.

As HDPE was originally crystallized from solution in the absence of external forces during injection molding, there was no preferred orientation along which the polymer chains lie. During cold drawing, the semi-crystalline HDPE specimen underwent large-strain plastic deformation in tension. Molecular alignment of the polymer chains, therefore, occurred in the principle direction of extension parallel to the axis of the polymer. Cold drawing after the yield point was expected to lead to a strain hardening process. Otherwise, the material would break without drawing at the reduced cross-section where necking occurred. The occurrence of strain hardening, followed by an increase in anisotropy, was due to the reason that during cold drawing, the originally soft intermolecular interactions, which were manifested by the isotropic elastic properties of HDPE, were replaced by stiffer intra-molecular interactions experienced by the backbone of HDPE. The strain hardening of HDPE after cold drawing was confirmed in Figure 6, in which the hardness of the necking region was significantly higher (p < 0.05) than that of the undrawn part. This indicated that the higher the strain provided to the polymer during cold drawing, the higher the level of anisotropy finally obtained.