Cable Extrusion Defects: Voids, Protrusions and the Die That Causes Them

Three defects, three different causes

Cable insulation fails in three ways that get lumped together as extrusion problems. Voids are bubbles inside the wall. Protrusions are lumps on or through the surface. Surface defects are roughness, die lines and melt fracture. They look related and they are not: voids come from the melt, protrusions come from contamination or from the conductor, and surface defects come from the die and the draw.

Getting that wrong is expensive, because the fixes point in opposite directions. A void problem is usually solved by running hotter. A surface defect problem is often made worse by running hotter, because a hotter melt is a lower-viscosity melt and it is more prone to melt fracture at the same output.

The first step on any defect is therefore to cut a section and look at where the defect is. A void is inside the wall and does not touch the surface. A protrusion is at the surface and often has a hard core. A die line runs continuously along the length at the same angular position. Each of those three observations eliminates two of the three causes.

The second step is to measure the defect rate rather than describe it. Voids per metre, protrusions per 100 metres, and a surface roughness number. A defect that is counted can be tracked against a process change; a defect that is described cannot.

Voids come from the die, not from the resin

A void is a bubble of gas trapped in the insulation. The gas is almost always air, and the air got in at the die. In a crosshead, the conductor enters through a guide tip and the melt flows around it at 90 degrees to the conductor axis. That turn creates a low-pressure region behind the tip, and any air that is drawn in there is carried into the insulation.

The counter-intuitive part is the temperature relationship. Void count falls from 42 per metre at a 170 C melt to 26 at 180, 15 at 190, 9 at 200, 6 at 210 and 5 at 220. A hot melt is a low-viscosity melt, it fills the die more completely, and it seals the gap between the tip and the die land so the air cannot be drawn in. A cold melt is stiff, it does not seal, and it carries air.

So the instinct to run cooler to avoid degradation is the instinct that produces voids. The correct move is to run hot enough to seal and to control the residence time so that the hotter melt is not also a degraded melt.

The other air source is the conductor itself. A stranded conductor carries air between the strands, and at the die that air expands and migrates into the melt. The fix is a vacuum box on the conductor before the die, or a compression tip that squeezes the strand closed. A solid conductor does not have this problem, which is why a solid conductor line runs cleaner than a stranded one on the same die.

Protrusions: contamination, and the ones that are not

A protrusion on the surface of an insulation is a particle that is harder or bigger than the melt. The usual source is contamination: a bit of cured compound from a previous run, a piece of the screw flight, a fragment of the filter pack. A 300 micrometre particle in a 1 mm wall is a bump; the same particle in a 0.3 mm wall is a hole.

The screening is the first line. A 40 to 60 mesh screen pack on a cable line catches the particles that would make a 0.3 mm protrusion, and it is changed on a pressure rise rather than on a schedule. A screen that is left in until the line stops is a screen that has been bypassing for a week.

The second source is the compound itself. Crosslinked polyethylene is compounded with a peroxide, and a peroxide that has started to decompose in the bag produces gel particles that will not melt. A compound stored warm or used past its date is a compound that makes protrusions, and no screen will catch a gel that forms downstream of the screen.

The third source is the conductor. A strand that is out of place, a burr on a wire that was cut, a kink in a solid conductor: all of them push the insulation out from the inside. Those protrusions are always at the same angular position and they repeat with the lay length of the conductor, which is how they are identified.

There is a fourth, and it is the one that gets missed: moisture. A compound with 0.05 percent moisture produces steam at the die, and steam makes a protrusion that looks like a blister rather than a bump. It is soft when it is cut, which is how it is told apart from a particle.

Draw-down ratio and the surface

The draw-down ratio is the area of the die opening divided by the area of the finished insulation. A ratio of 1.0 means the insulation leaves the die at its final size. A ratio of 2.0 means it is drawn to half the area after it leaves, which is the normal way to run a cable line because drawing closes the die marks.

The useful window is narrow. Below 1.4 the insulation is barely drawn, the die marks do not close, and the surface carries the machining marks of the die land. Above 2.1 the draw is so severe that the melt is stretched beyond its ability to recover, and any small disturbance in the die becomes a visible defect. Measured on a PVC insulation line, the surface defect rate is 0.4 percent at 1.2, 0.5 at 1.6, 0.9 at 2.0, 2.1 at 2.4, 5.6 at 2.8 and 11.8 at 3.2.

So a die is designed for a draw-down ratio, and running the same die on a smaller conductor or with a thinner wall moves the ratio. A die designed for 1.8 running at 2.6 because the wall was reduced is a die that will produce die lines that were not there on the original specification.

The die land length matters as much as the ratio. A land that is too short does not build enough pressure to smooth the melt, and the surface carries the memory of the screw. A land that is too long raises the pressure and the melt temperature, which brings back the degradation risk. The usual land is 8 to 15 times the wall thickness.

Table 1 — The draw-down window for a PVC insulation line

Temperature, shear and the melt fracture window

Melt fracture is a surface instability that starts when the shear stress at the die wall exceeds a critical value. It shows first as a loss of gloss, then as a fine roughness, then as sharkskin, then as a helical distortion. The critical stress depends on the polymer and on the die, and it is the ceiling on output for a given die and temperature.

The way out is to raise the melt temperature, which lowers the viscosity and the stress at the same output. That is why a line that is fracturing can be cured by 10 C rather than by slowing down, and why the void problem and the fracture problem both point the same way: hotter.

The limit on hotter is degradation. PVC degrades above about 200 C, crosslinked polyethylene is compounded to run at 130 to 150 C in the die and to crosslink after, and a polyolefin insulation can run to 230 C. So the window between fracture and degradation is wide for a polyolefin and narrow for PVC, and that is why PVC cable lines are run with a temperature profile that is deliberately flat rather than rising.

The other lever is the die geometry. A longer taper into the land, a smoother transition and a polished land all raise the critical stress, which buys output without temperature. A die that was made with a 30 degree included angle and a 2 mm land is a die that will fracture earlier than one with a 15 degree angle and a 6 mm land.

Table 2 — Reading a defect by where it is and how it repeats

A defect list that leads somewhere

A useful defect log on a cable line has four columns: what the defect is, where it is in the wall, how it repeats, and what the line was doing when it appeared. The last column is the one that is usually missing, and it is the one that turns a log into a diagnosis.

The variables worth logging are melt temperature at the die, line speed, draw-down ratio, screen pack pressure, and the compound lot. A defect that correlates with a lot is a compound problem. One that correlates with speed is a fracture problem. One that correlates with screen pressure is a contamination problem. One that correlates with nothing is usually a die problem that has been there since the last clean.

The economics are worth stating. A 1 percent defect rate on a line running 200 metres a minute is 2 metres a minute of scrap, plus the cost of the spark test failure that catches it downstream. Most of the defects in this article are cheap to fix once they are identified, and expensive to live with.