Polyethylene secondary flows

Many polymers exhibit neither a measurable stick-slip transition nor flow oscillation. For example, commercial polystyrene (PS), polypropylene (PP), and low density polyethylene (LDPE) usually do not undergo a flow discontinuity transition nor oscillating flow. This does not mean that their extrudate would remain smooth. The often observed spiral-like extrudate distortion of PS, LDPE and PP, among other polymer melts, normally arises from a secondary (vortex) flow in the barrel due to a sharp die entry and is unrelated to interfacial slip. Section 11 discusses this type of extrudate distortion in some detail. Here we focus on the question of why polymers such as PS often do not exhibit interfacial flow instabilities and flow discontinuity. The answer is contained in the celebrated formula Eqs. (3) or (5). For a polymer to show an observable wall slip on a length scale of 1 mm requires a viscosity ratio q/q equal to 105 or larger. In other words, there should be a sufficient level of bulk chain entanglement at the critical stress for an interfacial breakdown (i.e., disentanglement transition between adsorbed and unbound chains). The above-mentioned commercial polymers do not meet this criterion. 

Olah (17a) has also reported the alkylation reactions (at -10 with 1 1 HS03F-SbF5) of n-butane with ethylene to yield 38 weight percent of hexanes and of n utane with propylene to yield 29 weight percent of heptanes. The former reaction has also been reported by Parker (31) at 60 , but the product in this case more nearly resembles polyethylene degradation products. In our work with 10 1 HF-TaF5 at 40 , in a flow system, ethylene (14.1 wt.%) reacted with rv-butane to form 3-methyl-pentane as the initial product of 94% selectivity (Scheme 6, path a). The alternative, i.e., the direct reaction of ethylene with a secondary-butyl cation (path b), can be ruled out since butane does not ionize under these conditions (vide supra). 

To help understand and quantitatively evaluate the secondary movement shown above, Debbaut et al. [75, 77] augmented this experimental work with a three-dimensional flow simulation that incorporated viscoelastic effects. The finite element method, using a 4-mode Giesekus model as the viscoelastic constitutive equation, was used for the simulation. The polymer used for the experiment and simulation was a low-density polyethylene. Figures 12.20 and 12.21 show the experimental observations and the numerical predictions of the deformations of the interface for the rectangular straight channel [78], and for the teardrop channel [75], respectively. 

---