Troubleshooting · Defect guide
Sink marks and internal voids
Why injection-molded parts sink, and what to fix first
Almost every sink mark has one root cause: material in a thick section keeps shrinking after the skin has frozen. This guide covers how to confirm it really is a sink mark, the mechanism, four groups of causes, checks ordered from cheapest to most expensive, and an honest answer on how little additives can do here.
Marketing Department, ADVANCEDCHEM SOLUTIONS CO., LTD. · 22 October 2026 · about 7 min read
Confirm it is a sink mark
A sink mark is a shallow depression in the surface, easiest to see with light at a low angle. It sits opposite a thick section: the face behind a rib root, a boss root, or a corner where two walls meet. It appears in the same place on every shot.
- Sink mark: a real depression. Lay a straight edge across it and you see a gap. Always over a thick section.
- Gloss difference / ghosting: the surface is flat but shines differently. No gap under a straight edge. Usually uneven mould temperature.
- Flow marks: wavy bands following the flow direction, not tied to thick sections.
- Internal void: the surface is flat but a cut through the thick section shows a cavity. Same cause as a sink mark; the skin was just stiff enough not to be pulled in.
Mechanism: why thick sections sink
Melt cools from the mould wall inwards. The skin freezes first while the core of a thick section is still molten. As that core cools, its volume shrinks. If no fresh melt is packed in during that time, because holding pressure is too low or the gate has frozen, the shrinkage pulls the still-soft skin inwards. If the skin is already stiff, the core tears open into a void instead.
The practical takeaway: a sink mark is a volume problem. There are only three ways to fix it: make the thick section thinner, pack melt in for long enough, or reduce how much the material itself shrinks. The first is the most reliable; the last is the weakest.
Four groups of causes
1. Part design
The number one cause. A rib or boss that is thick relative to the nominal wall always creates a thick spot at its root. A common design guideline is a rib thickness of about 40–60% of the nominal wall, at the low end for high-shrinkage resins such as PP, and coring out thick sections instead of leaving them solid.
2. Mould
A gate that is too small or far from the thick section freezes before the section is packed. Areas with poor cooling, such as small cores without water, run hotter and sink more easily.
3. Process
Too little holding pressure or time, switching from velocity to pressure (V/P) too early, zero cushion, a melt temperature that is too high and so adds shrinkage, and a mould surface that is too hot on the side where the sink shows.
4. Material
Semi-crystalline resins such as PP and HDPE shrink more than amorphous ones. A shift in MFI from a new grade, a new lot or a different regrind ratio changes how much melt enters the cavity. See Polymer basics for the background.
Step-by-step checks, cheapest first
- Look at the cushion. If the screw bottoms out to zero every shot, the shot is too small and holding pressure has nothing to push. Increase the shot size before anything else.
- Weigh every part for about 20 shots. If weight drifts, find the source of instability first, such as the regrind ratio or the dosing unit.
- Run a gate-freeze study. Extend hold time step by step and weigh the parts. When weight stops rising, the gate has frozen; set hold time slightly beyond that point. If the gate freezes before the thick section can be packed, the fix is in the mould.
- Step up holding pressure while watching for flash. If flash appears before the sink disappears, you are beyond what the process can fix.
- Measure mould surface temperature at the sink. If it is clearly hotter than elsewhere, check the cooling circuit at that spot.
- Cut the part through the thick section. If you find a void, making the skin cool faster has only moved the problem inside.
- Review the design. If none of the above works, the rib or boss root must get thinner. That is mould work and the most expensive step.
Cause → check → fix table
| Cause | Check | Fix |
|---|---|---|
| Short shot size, zero cushion | Read the cushion on the machine display | Increase dosing stroke to leave a stable cushion |
| Hold time shorter than gate freeze time | Gate-freeze study (part weight vs hold time) | Extend hold time past the point where weight levels off |
| Holding pressure too low | Step pressure up, watch sink and flash | Raise pressure within the flash-free window |
| V/P switchover too early | Mould with velocity phase only (no hold); the part should be about 95–98% full | Adjust the V/P switchover point |
| Gate too small or far from the thick section | Gate freezes early even with long hold time | Enlarge the gate or move it closer to the thick section |
| Mould surface too hot at the sink | Compare surface temperature at several points | Clean or rework the water circuit; use a high-conductivity insert in small cores |
| Melt temperature too high | Measure the temperature of purged melt | Bring it into the resin supplier's recommended range |
| Rib or boss too thick | Measure root thickness against the nominal wall | Thin it, core it out, or use several thinner ribs |
| Inconsistent material | Log resin lot and regrind ratio against part weight | Hold the blend ratio constant; check MFI when the lot changes |
The 95–98% fill figure is a general V/P setting guideline; adjust it to the actual part.
How much can additives help?
The honest answer is not much. Sink marks are about volume and heat, and design or process fixes beat formulation changes in nearly every case. These are the material options that actually exist, including ones ACS does not sell:
- Mineral fillers such as talc or CaCO₃ reduce the share of shrinking polymer, so sinks can get smaller. Weight, stiffness, toughness and surface change with them; confirm the part still meets its specification.
- Chemical blowing agents build internal pressure that offsets shrinkage in thick parts. They suit parts that can accept a matte surface or a fine-celled core. ACS does not supply them.
- Nucleating agents in PP make the skin set faster and crystallisation more uniform, but their effect on sink marks is inconsistent: better in some cases, no different in others. Do not rely on them as the main fix. If your real goal is cycle time or warpage, see Cycle time and warpage.
What additives cannot fix: a short shot size, an early-freezing gate, too little holding pressure, and a design with oversized thick sections.
Technical caution
Let-down ratios and technical data on this page are typical reference values, not a guarantee of properties. Confirm by trial in your own process before production use.
FAQ
Sink mark questions we hear often
We raised holding pressure and the sink mark is still there. Why?
Most often the gate has already frozen, so the extra pressure never reaches the thick section. Run a gate-freeze study: if part weight stops rising as you extend hold time, the gate is closed. The fix is then gate size or location, mould temperature near the gate, or wall design, not more pressure. The other common cause is zero cushion: the screw bottoms out, so there is no melt left for holding pressure to push.
Why does PP sink more than ABS in the same mould?
PP is semi-crystalline. As it cools, chains pack into crystals that are denser than the melt, so its volume shrinks more than amorphous ABS. Typical textbook mould shrinkage is about 1.0–2.5% for PP and 0.4–0.7% for ABS, depending on grade, thickness and process conditions. A part designed for ABS and switched to PP therefore tends to sink at rib and boss roots.
Are sink marks and internal voids caused by the same thing?
Yes. Both come from molten core material in a thick section shrinking after the skin has set. If the skin is still soft, it is pulled in as a sink mark; if the skin is stiff enough, the core tears apart into a void. Cooling the surface faster to hide a sink can therefore turn it into a void, which is invisible but weakens the part. Gas bubbles from moisture are a separate cause; dry the material and mould again to tell them apart.
Sink marks got worse after we added regrind or recycled pellets. Is that related?
It can be. Regrind and recycled pellets often have a different MFI from the virgin grade. The viscosity change shifts the V/P switchover and how much melt enters the cavity. If the blend ratio is not constant, part weight drifts and sink marks appear on some shots only. Weigh every shot while the problem occurs and hold the blend ratio steady before adjusting the machine.
Sink marks still there after tuning the machine?
Send photos of the part, a cut through the thick section and your machine settings on LINE. Our technical team will help you separate mould, process and material causes.