The strange non-Newton world inside the barrel

Upcoming metallocene elastomers offer better properties

If your resin melt processes easily without thermal degradation, sharkskin or melt fracture, how would you describe it? “It flows like water” is a popular expression. The interesting thing about what’s happening inside the barrel, however, is that the resin isn’t flowing like water, it’s flowing like a “Non-Newtonian Fluid”.

Water is a Newtonian fluid (the same guy of high-school physics, falling apple fame) which means that it behaves like, well, water. Take a rotating rod, for example and stick it into a glass of water. The spinning water is flung to the outside of the glass, and a familiar bathtub drain vortex appears. Repeat the experiment with a glass of molten resin, however, and the liquid will bulge at the spinning rod, a totally opposite effect.

This proves two things: The first is that you have far more time on your hands than anyone in the plastics (or magazine) industries does. Second, it shows that moving that non-Newtonian liquid resin changes its viscosity.

And it’s not just flow you’re dealing with either. Temperature also changes viscosity, and while we all know that intuitively, for the mathematicians out there, it follows an Arrhenius-type relation. For the rest of us, that means that polymer melt flows are much more temperature sensitive that we realize.

How sensitive? The formulas suggest that a 1% error in temperature measurement can cause a 30% change in the melt viscosity! Feedback control through computers, combined with tighter sensing, has lulled us into a false sense of complacency about temperature and its effects on melt flow.

Now that you’re sufficiently impressed with the temperature dependence of the melt, how about pressure? There’s lots of pressure in the barrel, and it does more than heat and convey the resin. This time the math links viscosity and pressure in different ways depending on the process. For extrusion running at say, 7000 psi, (about 500 atm for the metric generation), viscosity varies roughly linearly with pressure, so expect a typical resin viscosity to be about five times the atmospheric value at a given temperature. This holds all the way down the pressure scale to ambient.

Monitoring that parameter can tell you something useful during troubleshooting. Unfortunately, you can’t extrude a resin garbage can. At a hypothetical pressure of 15,000 psi (about 1000 atm), pressure and viscosity vary exponentially; and the melt might see a viscosity thirty times higher than ambient. Triple the pressure and the viscosity parameter increases by over twenty thousand times! Obviously, these are not real-world conditions, but the point is that highly non-linear dependencies mean that how closely you look at (and control) any given parameter depends on how linear its relationship is to the thing you’re trying to control.

And productivity means speed, which generally translates to higher pressures, temperatures and shear speeds, so it gets tougher as you try to squeeze that last five percent out of your equipment. Exponentially tougher.