sophiecentaur said:
We've all seen that to a lesser extent when pulling cling film off the roll at home. It would be interesting to know what the actual charging mechanism is. Is it due to the original heating, stretching and rolling that one face of the film gets one charge and the other face gets the other? Or is it to do with the curvature?
Those long sparks are a great demonstration of Q = CV. The C between faces drops drastically as they separate and so the V has to increase accordingly. The user supplies the energy in this case (and the roll does, in fact, feel sticky due to the attraction)
To set up the scene requires a lot of verbiage.
It is difficult to measure static voltage at the die outlet (>600°F, and not very safe, being at the roll nip), but I never read anything higher than 200V per an older model Simco electrostatic field meter. However, web coming out of the roll stack could hit 150 kV to 200 kV before partially discharging. Surface charge dropped to 10 kV or less after the sheet passed through the next stage - a bath of diluted silicone release agent.
The sheet didn't rub against anything in between the roll stack, and silicone bath, so charge must have been generated within the roll stack itself. The one I have in mind is a conventional 3 roll vertical stack with 30" diameter chromed steel cooling rolls in a "downstack" configuration where melt enters the upstream nip between the middle and bottom roll, and sheet exits the stack from the top roll (moving downward, hence the "downstack" moniker). Haul-offs designed to process PET and PLA are generally in an "upstack" configuration (or a "J" stack variation, or for thin sheet, a horizontal roll arrangement may be used). I think this factors into it, because the only processing lines known for lightning bolts shooting across the sheet were both downstacks.
What makes a downstack problematic in PET/PLA processing is PET freezes very quick upon contacting the first chilled roll it touches. Gravity assures the 'bottom' side of the melt curtain hits the bottom roll first, whereas most of the cooling is done by the middle roll against the 'top' side of the sheet. Another issue is how much material 'bead' is at the roll nip. If there isn't enough bead (or it isn't uniform across the nip) the sheet will have flat spots and other surface defects. Polystyrene, and many other plastics are more forgiving in this regard, but with PET, too much bead demands considerably more drive torque than normal (or can stall the rolls outright), and creates an inordinate amount of roll deflection leading to a different form of surface defect. Finally, because the sheet has been shrinking all the time it has been in contact with the middle roll it adheres to it to a degree, and (with gravity assisting) tends to want to wrap around the middle roll rather than transfer to the upper roll before exiting the stack.
I suspect several mechanisms are at work, but my guess is most of the static generation occurs as the sheet peels away from the center roll. Static discharge appeared to be more severe when operators were running with a heavier-than-normal bead, but I don't know if this was a consequence of squeezing it in the nip, or some other effect.