Tube Extrusion Defects: Gels, Die Lines and Weld Lines

Defects sort into four families, and each has a different owner

When a tube line starts producing rejects, the argument usually starts before anyone has identified the defect. The fastest way to end it is to put the fault in the right family, because each family belongs to a different part of the process — and often to a different department.

Table 1 — Defect families. A defect that repeats at a fixed interval along the tube is almost always process stability or a roll; a defect that runs the full length is die or material.

That last distinction is the most useful diagnostic in the plant. A mark that runs the entire length of the tube, in the same angular position, is a die or a material problem. A mark that repeats every 200 mm is a roll, a belt or a haul-off issue. A mark that appears at random intervals is contamination or a feed problem.

Gels: the defect that is always the material's fault, except when it is not

A gel is a small lens-shaped lump of high-molecular-weight material that has not melted with the rest of the melt. It appears as a glossy dot, typically 0.1 – 1 mm, standing slightly proud of the surface. On a printed tube it shows as a white spot where the ink did not transfer.

Gels come from four different places and they look almost identical, which is why plants replace resin for months without solving the problem.

Table 2 — Gel sources. The fingernail test separates the two most common causes in under a minute.

The fingernail test is crude and reliable. Press a hard gel with a fingernail: if it will not dent, it is cross-linked and no amount of screw tuning will fix it. If it smears, it never melted and the barrel profile or screw speed is wrong. That one check has saved more resin changes than any analytical method.

Degraded material behaves differently again: it usually has a brown or black centre and a comet tail pointing downstream. That tail is the signature of a char particle that has been dragged along the die surface. It means there is a dead spot — a thread, a thermocouple pocket, a step in the adapter — where material sits and cooks.

Die lines: longitudinal, permanent, and always in the same place

A die line is a fine longitudinal groove or ridge running the full length of the tube. It does not move around the circumference, and it does not go away when you change speed — which is exactly what distinguishes it from a weld line.

Table 3 — Die line causes. A line that grows slowly during a run is carbon, not a scratch.

Polishing the die land is the fix for most of these, and it is a job that needs to be done properly. A die lip polished across the flow direction leaves microscopic scratches that become die lines; the polish has to run along the flow, and it has to end at a finish that is better than the surface you want on the tube. A 0.4 µm finish on the land is normal for a packaging tube die.

The draw-down effect is the one that catches people. At a draw-down ratio of 6:1, any 1 µm imperfection on the land becomes a visible line because it is stretched out and the polymer orientation makes it show. Reduce the draw-down and the same die produces an acceptable tube. This is why a die that worked at 3:1 "suddenly" fails when the wall is thinned to save resin.

Weld lines: where the melt rejoins

A weld line forms wherever two flow fronts meet. In a crosshead die there should be none — the melt flows around the mandrel as a continuous annulus. If you see a weld line on a crosshead tube, one of three things is happening.

• The die is a spider type. Support legs split the flow and it re-joins downstream. The line is faint but permanent. Spider dies are cheaper and they always leave a line.
• The die has a damaged or obstructed flow path. A build-up on one side of the mandrel creates a local stagnation and a visible knit line.
• The melt is too cold at the recombination point. Two fronts meet without enough mobility to interdiffuse, and the interface is weak as well as visible.

The third case is the dangerous one, because the line is not just cosmetic. A cold weld line is a mechanical weak point, and on a thin-wall tube it becomes the place where the tube splits when squeezed. If a weld line is visible and the tube also fails a burst test at the same angular position, it is a cold weld, not a surface mark.

Surging and bambooing: the periodic defects

If the tube diameter varies in a regular rhythm along its length, you have a stability problem, and the period tells you where.

Table 4 — Periodic defects and their period. Measuring the period is faster than any other diagnostic.

Bambooing — the tube looks like a bamboo culm, with regular bulges — is usually a haul-off problem rather than an extruder problem. If the belt tension is low, the tube slips and catches, and the resulting speed oscillation appears as a diameter wave. It is often blamed on the screw, which is expensive and wrong.

Real screw surging is different: the melt pressure trace shows a slow oscillation with a period of tens of seconds, and it is usually caused by a feed zone that is too cold, a compression ratio that is too high for the resin, or a screen pack that is progressively blocking. The pressure trace is the instrument that settles it — a stable screw with a wandering pressure is a die or screen problem, and a wandering screw speed with stable pressure is a drive or feed problem.

Surface defects: melt fracture, gloss bands and drag marks

Table 5 — Surface defects. Sharkskin appears above a critical shear rate, so it is a speed problem before it is a temperature problem.

Sharkskin has a specific signature: it starts at the die lip and gets worse toward the exit, and it disappears if you slow down or raise the temperature. That is because it comes from the stress at the die wall exceeding what the polymer can relax before it freezes. The fix is to reduce shear stress — larger die gap, higher temperature, lower output — not to polish the die harder.

The triage order that saves the most time

When a defect appears, work through this list in order. It runs from cheapest to most expensive, and most problems are solved in the first three steps.

1. Is the defect longitudinal and fixed in position? Die or material. Go to the die.
2. Is the period measurable? Process stability or a roll. Log the period and compare against belt circumference and screw RPM.
3. Does it disappear at a different melt temperature? Unmelt or cold weld. Adjust the profile before touching the tool.
4. Does it appear on every cavity or every lane? If only one, it is local tooling. If all, it is material or the main flow path.
5. Did it start at a lot change? Material. Check bulk density, melt index and gel count on the incoming resin.
6. Did it start after maintenance? Assembly error — a seal reversed, a thermocouple in the wrong pocket, a die not fully seated.

Step four is the one that is skipped most often on multi-lane lines. If a line runs four tubes from one die head and only one lane is affected, the material is innocent and the answer is in that lane's flow path. Checking all four lanes separately takes five minutes and eliminates half the possible causes.

What defects actually cost

Table 6 — Cost of detection point. The further downstream a defect is caught, the more value has been added on top of it.

The economics are brutal in one direction: a gel costs almost nothing to scrap as a plain tube, and it costs the full printed, welded, capped and boxed value to scrap after decoration. Every hour of production after a defect starts is value added to a part that will be thrown away. This is why the inspection point belongs as early in the line as it can be made to work.