Injection Molding Defects: 12 Problems, Their Causes and Fixes

1. Diagnose before you adjust

The most expensive habit in injection moulding is turning knobs before identifying the cause. A moulder changes the holding pressure, then the melt temperature, then the cooling time, and after two hours the part is different but nobody knows why — and the fix disappears on the next run.

Three questions identify the cause of almost any defect faster than any parameter change:

• Where on the part is it? Near the gate, far from the gate, at a weld line, at a rib, on the parting line. Position is a diagnosis. A defect far from the gate is a pressure or flow problem. A defect at the parting line is a tool clamp problem. A defect at a rib is a shrinkage problem.
• When does it appear? First shots after a stop, after a material change, after a tool clean, gradually over a run. "Gradually over a run" usually means temperature drift or material contamination, not a parameter that was always wrong.
• Does it affect every cavity? One cavity out of eight points at that cavity's gate, cooling or vent. All cavities point at the machine, the material or the common runner.

2. Short shot

The cavity does not fill completely. It shows as a rounded, smooth-edged void at the end of flow — quite different from a sharp broken edge, which would be damage rather than under-filling.

Causes. Insufficient injection pressure or volume; injection speed too low so the melt freezes before it arrives; melt temperature too low, raising viscosity; mould temperature too low, freezing the flow front; a gate that is too small or in the wrong place; inadequate venting, so trapped air stops the flow; wall sections too thin for the flow length; or simply a machine with too little shot capacity.

Fixes, in order. Check the shot size against the machine's rated capacity first — an undersized barrel cannot be fixed by any parameter. Then raise injection speed, which is the most effective single change. Then raise melt temperature in 10 °C steps. Then raise mould temperature. Then increase the gate. If the part has a thin-to-thick transition, the answer is usually a redesign of that transition, not a process change.

Do not. Do not simply raise holding pressure to fix a short shot. If the cavity is not full at the end of injection, there is nothing to hold, and the extra pressure only packs the near-gate area harder — which produces a part that is oversized at the gate and short at the end.

3. Flash

Material escapes the cavity at the parting line, around ejector pins, or at an insert. It appears as a thin film of plastic attached to the part edge.

Causes. Clamp force too low for the projected area; injection pressure exceeding what the clamp can hold; mould temperature too high, thinning the melt; melt viscosity too low (over-heated, or a grade with too high a melt flow index); worn or damaged parting surfaces; a tool that was never properly fitted; or insufficient vent depth allowing melt to escape where air should.

Fixes. Calculate the required clamp force from projected area and cavity pressure before touching anything — flash on a large flat part is often simply an undersized machine. Then reduce injection pressure and holding pressure. Then check the parting surfaces: a 0.02 mm step or a piece of flash left from a previous run will cause flash on every shot afterwards.

The detail that matters. Flash that appears only on one side of the part is not a pressure problem. It is a tool alignment or clamp parallelism problem, and no amount of parameter adjustment will remove it.

4. Sink marks

A depression on a surface opposite a thick section, a rib, a boss or a heavy corner. It appears after ejection as the still-molten core contracts and pulls the skin inward.

Causes. A wall or rib that is too thick relative to the surrounding wall; insufficient holding pressure or holding time, so the gate freezes before the core has packed; mould temperature too high on the thick side, or cooling that is unbalanced; melt temperature too high, which increases the amount of shrinkage the pack has to compensate; or a gate that is too small and seals early.

Fixes. The permanent fix is design: ribs at 50–60 % of the adjoining wall thickness, bosses with a wall no thicker than the nominal wall, and corners cored out rather than left solid. The process fix is longer holding time and higher holding pressure, plus gate freeze checked by a gate-seal test — weigh the part at increasing hold times until the weight stops rising. That point is the correct holding time, and it is usually longer than the current setting.

Why it keeps coming back. Sink is often "fixed" by over-packing the part, which removes the visible sink but leaves internal stress that shows up later as warpage or cracking. Fix the geometry.

5. Voids

Bubbles inside a thick section, invisible from outside until the part is cut or x-rayed. Voids and sink are the same phenomenon with different surface behaviour: thick sections shrink, and the material either pulls the surface in (sink) or pulls away from the inside (void).

Causes. A section too thick to pack from the gate; low holding pressure; a gate that freezes before the centre of the thick section has solidified; moisture in the resin, which produces vapour bubbles that look similar but are a different problem; or gas from thermal degradation of an overheated melt.

Fixes. Distinguish moisture voids from shrinkage voids first: moisture voids are usually smaller, more evenly distributed and appear with silver streaks. If they are shrinkage voids, increase holding pressure and time, reduce the melt temperature slightly, and cool the thick section more aggressively. If the wall is thicker than about 5 mm, the answer is usually a design change — core it out, or accept gas-assisted moulding.

6. Weld lines

A visible line where two flow fronts meet, or where melt flows around an insert or a core and rejoins. On a structural part it is a weak point; on a cosmetic part it is a visible mark.

Causes. Any obstruction in the flow path — a hole, an insert, a core pin. Multiple gates. Flow fronts that cool before they merge. Low melt or mould temperature, which prevents proper molecular diffusion across the interface. Air trapped at the meeting point, which stops the fronts from knitting.

Fixes. Raise melt and mould temperature to improve diffusion across the interface. Raise injection speed so the fronts meet while still hot. Move the gate so the weld forms in a non-critical area, or move it so the weld disappears entirely. Add a vent or an overflow well at the weld location so trapped air can escape.

The part buyers miss. A weld line at the surface is a cosmetic problem. The same weld line 0.5 mm below the surface is a strength problem — weld strength in glass-filled materials can be 40–60 % lower than bulk strength, and a part that passes a drop test in unfilled ABS can fail in glass-filled PA66 with an identical-looking weld.

7. Warpage

The part is dimensionally correct when it comes out of the tool and bends, twists or cups afterwards. It is the most expensive defect on this list because it usually requires a tool change, not a parameter change.

Causes. Uneven shrinkage, and uneven shrinkage has three sources: differential cooling between the two halves of the tool, differential flow orientation (especially in glass-filled grades), and differential wall thickness. Also: ejection force applied unevenly, packing that is heavier on one side, and parts ejected too hot.

Fixes, in order of effectiveness. Balance the mould cooling — measure the two mould halves and both sides of the cavity, and match them within a few degrees. Reduce the mould temperature difference across the part. Increase cooling time so the part is rigid at ejection. Reduce holding pressure on the side that is over-packed. For glass-filled parts, move the gate so the flow orientation is symmetrical about the part centreline.

What does not work. Adding more ejector pins to force the part flat. It will bend back as soon as it warms up, and it leaves ejector marks.

8. Burn marks

Black or dark brown scorch marks, usually at the end of flow, at a weld line, or at the base of a rib. Sometimes the part smells burnt.

Causes. Diesel effect — trapped air compressed so fast that it ignites. Overheated melt, or melt held in the barrel too long. A hot-runner zone set too high. A gate that is too small, producing shear heating at the gate. Insufficient venting at the point where the mark appears.

Fixes. Vent the location. This is the real fix and it is often a two-minute job with a venting insert or a shallow grind on the parting line. Then reduce injection speed for the final fill stage so the air has time to escape. Then check all barrel and hot-runner zones against the material datasheet, and check the residence time — a small shot in a large barrel means the melt is cooking.

The tell. Burn marks that appear only after the machine has been idle for a few minutes are a residence-time problem, not a venting problem. The material sitting in the barrel has degraded and the first shots push it through.

9. Silver streaks and splay

Silvery streaks radiating from the gate, or a general haze across the part. This is the single most common defect caused by moisture, and the single most common reason a good tool produces bad parts.

Causes. Undried resin — by far the most frequent. Moisture condensing on cold tool surfaces. Thermal degradation of the melt. Air drawn in at the nozzle or along the screw. A contaminated regrind stream.

Fixes. Verify drying before touching anything else: check the dryer's dew point, the drying temperature against the datasheet, and the actual drying time. Confirm the hopper is sealed and the dryer is running during the shift, not just before it. Then check for nozzle drool and suck-back settings. Then check the regrind ratio.

Why "we dried it" is not an answer. A tray oven at 120 °C for two hours does not dry polycarbonate. It needs desiccant-dried air at −20 °C dew point for three to four hours. And dried pellets sitting in an open hopper re-absorb moisture within a couple of hours in a humid plant.

10. Flow marks

Wavy, matte or glossy bands on the surface, following the flow direction, usually far from the gate. Cosmetically unacceptable on a visible surface; mechanically harmless.

Causes. The flow front cooling and hesitating as it advances, so successive waves leave a record on the surface. Low melt temperature, low mould temperature, injection speed too slow, or a sudden change in wall thickness that stalls the front. Sometimes a gate positioned so the melt has to flow a long way across a cosmetic surface.

Fixes. Raise mould temperature first — on a textured or high-gloss part this alone often solves it. Then raise injection speed so the front advances without hesitating. Then raise melt temperature. If the gate is on the cosmetic face, moving it to a hidden edge is the permanent fix.

11. Jetting

A snake-like, folded trail of material starting at the gate and running across the part, instead of a smooth radiating flow front.

Causes. The melt entering the cavity as a free jet rather than a spreading front, because the gate is too small relative to the cavity thickness, or the injection speed is too high at the start, or the gate does not face a wall.

Fixes. Slow the first-stage injection speed so the melt enters gently and establishes a front. Enlarge the gate. Re-angle the gate so it impinges on a cavity wall rather than firing into open space. Jetting is a tool-and-profile problem; there is no material change that fixes it.

12. Delamination

The surface layer peels away in flakes or sheets. It looks like the part is laminated, which it effectively is.

Causes. Incompatible materials mixed together — a purge that was not complete, or a regrind stream containing a different polymer. Excessive moisture. Contamination with a mould-release agent or oil. Shear stress from a cold melt and a small gate. In some cases, over-packed parts release surface layers under stress.

Fixes. Purge the barrel thoroughly and verify with a purge test. Segregate regrind by material and keep the ratio below 20 % unless the grade allows more. Check that no release agent is being sprayed into the cavity. Raise melt temperature and enlarge the gate if shear is the cause.

The one you cannot fix in the press. If the delamination comes from a mixed regrind stream, every part made from that batch is suspect. Quarantine it.

13. Colour streaks

Swirls, dark streaks or light patches in a coloured part, or colour that varies between cavities and between shots.

Causes. Masterbatch not dispersed — usually because the melt is too cold or the back pressure too low. Colourant incompatible with the base resin. Excessive regrind diluting or darkening the colour. A hot-runner manifold with dead spots where material degrades. Screw wear producing unmelted particles.

Fixes. Raise back pressure and melt temperature to improve dispersion. Increase the screw's mixing action with a higher back pressure and a slightly longer cooling time. Check the masterbatch let-down ratio against the supplier's recommendation. If the streaks are dark and appear after idle periods, the cause is a hot-runner dead spot, not the colour.

14. The troubleshooting matrix

The table below maps each defect to the four places a cause can live. Use it to decide where to look first, not to confirm a guess you already made.

Read across, not down. A defect with a cause in only one column is usually that cause. A defect with plausible causes in all four columns — warpage is the classic — needs the cooling and the geometry checked before any parameter is touched.

15. When it is the tool, not the process

Some defects cannot be processed out, and recognising them early saves weeks. If any of the following is true, stop adjusting parameters and call the toolmaker.

• The defect appears in one cavity only, consistently. That cavity has a gate, vent or cooling problem that the others do not.
• The defect is identical on two different machines with two different operators. The common factor is the tool.
• The defect moves when you change the mould temperature but never disappears. The cooling circuit is wrong, not the setting.
• Flash appears on one side only. Parting surface or alignment.
• The part is out of tolerance in a way that no pressure change affects. The cavity dimension is wrong — shrinkage was estimated incorrectly.
• A defect appears after a tool clean and never goes away. Something was damaged or a vent was blocked with residue.

The commercial point: a defect that requires a tool change should be identified before the tool is accepted, not after 50,000 parts. That is what a proper first-article inspection and a dimensional report on the T1 samples are for.

16. Frequently asked questions