Weld Lines in Injection Molding: Why the Part Breaks There and How to Move It

What a weld line is

A weld line is the seam where two flow fronts meet inside a cavity. When the melt splits around a core, a pin or a hole and rejoins on the far side, the two fronts have cooled and oxidised slightly during their separate journeys, and when they meet they form a bond that is weaker than the bulk material.

The mechanism is straightforward. Polymer chains bond by diffusion: two surfaces fuse when chains from each side are mobile enough to cross the interface and entangle. At a weld line the interface has been exposed to the mould and to the air in the cavity, so it has a skin of oriented, cooler, partially solidified material. The chains cannot cross that skin, so the bond is mechanical rather than molecular.

There is also a geometric effect. Where two fronts meet, the entrained air and any volatiles have to go somewhere, and if there is no vent at the meeting point they end up in the seam as a V-notch. That notch is a stress concentration on top of a weak interface, which is why weld lines are where parts break rather than where they merely measure low.

Strength is set at the interface

The two variables that dominate weld strength are the temperature of the melt when the fronts meet and the speed at which they meet. Hotter melt means more mobility and deeper diffusion across the interface; faster fill means less time for the front to cool before it joins.

Measured against bulk strength, a weld line in polypropylene runs 55 percent at a 210 °C melt, 68 at 230, 78 at 250, 85 at 270 and 90 at 290. The curve flattens at the top because there is a limit to how much diffusion can be recovered: even at the highest practical temperature the interface is not bulk material.

Fill velocity does something similar from the other direction. At a flow front velocity of 20 mm per second the weld is 58 percent, at 50 it is 70, at 100 it is 80 and at 200 it is 86. Faster fill gets you most of the way to the ceiling with much less thermal risk than raising the melt temperature, which is why high injection speed is the first lever experienced moulders reach for.

Table 1 — The five levers, and what each one buys.

The four things that actually move it

Of the five levers in the table, the two process ones are the easiest to pull and the two design ones are the most effective. A part whose weld line runs across a loaded boss cannot be saved by process, and a part whose weld line is in a non-structural web is fine even at 60 percent strength.

Melt temperature is limited by the material. Above the recommended range, polypropylene degrades, polycarbonate yellows, and polyoxymethylene releases formaldehyde. The gain from the last 20 degrees is worth less than the risk. Mould temperature is under-used: raising a mould from 40 to 80 °C adds perhaps 10 percentage points of weld strength and costs cycle time, and it is often the cheapest improvement available.

Injection speed is the lever with the best return, but it has a limit. Above a certain velocity the flow becomes unstable, jetting appears, and the weld moves but the part gains a new defect. The practical approach is to run a velocity ramp that fills fast and packs slowly, so the fronts meet at high speed and the packing is still controlled.

The materials that weld badly

Weld strength as a fraction of bulk varies enormously between polymers, and the variation is bigger than the variation between process settings. Semi-crystalline materials with a sharp melting point weld worse than amorphous ones, because the interface has to re-nucleate crystals rather than just diffuse.

Table 2 — Weld strength by material, and what recovers some of it.

Glass-filled materials deserve the emphasis they get. The fibres do not cross the weld interface; they orient parallel to it, which means the seam is not only unbonded but also unreinforced. A 30 percent glass polypropylene part can lose two thirds of its strength at a weld line, and that is a design assumption that has to be made explicitly rather than discovered in a test.

Moving the weld line

The most effective fix is to put the weld where it does not matter. Weld lines form where flow fronts meet, and flow fronts meet where the cavity splits and rejoins, or where two gates feed the same region. Both are design decisions.

• Single gate in the centre of a symmetrical part puts the weld at the far edge, which is often the strongest place for it.
• Two gates on opposite ends put the weld in the middle, which is usually the worst place, because that is where the bending moment peaks.
• An overflow well at the meeting point lets the cold front material flow past the part and be trimmed off, which removes the notch.
• Sequential valve gates can be timed so that one front is still molten when the other arrives, which recovers most of the strength.
• A flow leader, a slightly thicker channel from the gate to the meeting point, keeps the front hot for longer.

The overflow well is the oldest trick and still one of the best. A small pocket at the point where the fronts meet gives the contaminated front material somewhere to go, and the part is trimmed off the well afterwards. The cost is a little extra material and a trimming operation, and the benefit is a weld that is nearly as strong as the bulk.

Designing around it

Since the weld line cannot be removed, the design has to assume it. That means treating the weld as a material with reduced allowable stress, and putting it somewhere that the reduced stress is acceptable.

1. Assume 50 percent of bulk strength at a weld for an unfilled semi-crystalline material and 30 percent for a glass-filled one, unless you have test data for the actual geometry.
2. Keep welds out of bosses, ribs at high load, snap-fit cantilevers and any feature that sees a stress concentration.
3. Put the weld in a region loaded in compression rather than tension, because a weld is much better in compression.
4. Orient the weld away from the principal stress direction, since strength is lowest perpendicular to the interface.
5. If the weld must carry load, thicken the section locally so the stress is lower, and add a radius at the notch.

The last point is the one that is easiest to miss. A weld line in a thin section is a notch in a thin section, and the stress concentration factor at a sharp V-notch in a thin wall can be three or four. Thickening the section around the weld reduces the stress twice: once by increasing the area and once by reducing the notch sharpness.

Testing it

There are two ways to test a weld line and they answer different questions. A standard tensile bar moulded with a weld line in the middle, to ISO 527 or ASTM D638 type I with a double gate, gives a number that can be compared between materials and lots. Testing the actual part, loaded the way it will be loaded in service, tells you whether the design works.

Table 3 — The four ways to measure a weld line, and what each one is good for.

The microscopy option is worth using once per project, because it explains the other three. A polished cross-section through the weld shows whether the seam is a clean interface, a line of voids, or a notch open to the surface. A seam that is a line of voids will fail at low strength regardless of process; a seam that is a clean interface responds well to temperature and speed.