During the injection molding process, a high degree of fiber attrition occurs due to the large shear forces imposed during the plastication stage, leading to reduction in fiber length. Assuming that fibers break when the accumulated processing energy exceeds the elastic energy of the fiber, Eq. 1 shows that the fiber attrition occurs when:
E_p>E_s (1)
where E_p is the processing energy and E_s is the material’s strain energy. Traditionally, energy can be expressed as the product of force (stress) and displacement (strain) (E_p=σγ). From the rheological point of view, stress is the product of viscosity and shear rate (σ=ηγ ̇). Since strain can be expressed as shear rate multiplied by shearing time (γ=γ ̇t_s), the processing energy (E_p) can be derived in terms of stress (σ), shear rate (γ ̇ ), and shearing time (t_s ) as:
E_processing= σγ= ηγ ̇γ=ηγ ̇^2 t_s (2)
Assuming that the variation of each parameter is similar (i.e., in a completely molten state), Eq. 2 implies that the processing energy is more affected by the shear rate than by other factors, as it is squared. Based on this idea, the effect of shear rate on fiber attrition was systematically investigated, from 10 to 3,000 s⁻¹ using a capillary rheometer. Figure 1 shows (a) the fiber length distribution, (b) average fiber length (Lp), and (c) fraction of 9 mm fibers after extrusion using the capillary rheometer.
Figure 1. (a) Fiber length distribution, (b) average fiber length (Lp), and (c) fraction of 9 mm fibers as a function of shear rate, using a capillary rheometer equipped with a 1 mm die at 330 °C.
As the shear rate increases from 10 to 3000 s-1, the fiber length (Lp) exponentially decreases from 7000 to 4500 micron. Moreover, the volume fraction of 9 mm fibers, which are same length as the original fiber length in a pellet, also decreases exponentially. Interestingly, it was observed that some 9 mm fibers can survive after passing through a 1 mm orifice die at high shear rates, even above 1,000 s⁻¹, which is within the typical shear rate range encountered in the injection molding process. Another observation was the presence of bent fiber structures, as shown in Figure 2(a). Such bent and long fibers are not observed in Figure 2(b), which shows fibers from the injection molding machine. It has been commonly thought that glass fibers break during processing due to their brittleness. However, the presence of bent structures indirectly indicates that the glass fibers themselves are flexible enough, suggesting that there may be another factor causing fiber attrition.
Figure 2. Images of glass fibers obtained from (a) a capillary experiment and (b) an injection molding machine.
The major difference between the two methods, capillary rheometer and injection molding machine, was the precondition before mechanical shearing. The capillary rheometer allows a long time (> 10 minutes) at a high temperature of 330 °C to completely melt LFT pellets before extrusion, while the injection process imposes mechanical shearing to melt LFT pellets through frictional heating during the plasticization stage. Thus, the pellets are deformed in a liquid-like state in a capillary rheometer, whereas the pellets are forced to deform in a solid-like state in an injection molding machine. Therefore, it was hypothesized that this difference in preconditioning significantly affects fiber attrition. This observation also aligns with the hypothesis in Eq. 2 (E_processing=ηγ ̇^2 t_s), as the viscosity of unmelted resin is much higher than that of completely melted resin; the typical viscosity decreases with melting from >10¹⁰ Pa·s (pellet, T < Tg) to ~10³ Pa·s (melt, T > Tg), while the shear rate during injection molding varies by only several times ~10³ s⁻¹ under various processing conditions. Although the shear rate term is squared in Eq. 2, the difference in viscosity between the solid and liquid states exceeds the effect of the shear rate. Therefore, if the feed material is not completely melted, fiber attrition during the injection molding process is governed primarily by viscosity rather than by shear rate or shearing time. Since glass fiber is observed to be quite flexible, fiber attrition during injection molding is accompanied by the breaking of the solid polymer matrix holding the fibers, as illustrated in Figure 3.
Figure 3. Hypothesis of fiber attrition mechanism for LFT pellets.