Last updated: 2026-09-04 · 6 min read

Why Thick Ribs Sink and Bosses Void — Sink Marks, Vacuum Voids and Thick-Section Cooling

A thin-walled part fails when you thicken a section, because plastic shrinks as it cools and a thick region shrinks differently from a thin one. In a thick boss, rib or corner the outside skins freeze first while a molten core persists; as that core cools and contracts, it pulls material inward. The result is a sink mark if the skin is soft, or a vacuum void if it has frozen stiff.

Key Facts

Sink and void are the same root cause wearing different clothes — unequal cooling and shrinkage in a thick section that never got enough packing to compensate.

A sink mark is a surface dimple visible under raking light; a vacuum void is an internal hollow invisible from outside but catastrophic where it weakens a boss or traps stress.

The fix is never simply “add more pressure” — it is a chain of design and process moves that either remove the thick section, feed it while it shrinks, or let it cool evenly.

Cavity — The hollow space inside the mold that forms the part shape. Multi-cavity molds produce multiple parts per shot. Cavity count determines production output but also mold cost and complexity. Each cavity must be balanced for uniform fill. The related guide covers adjacent topics.

Sink or Void — What Decides Which?

Five factors, each pushing toward one defect or the other:

Factor	Pushes Toward Sink	Pushes Toward Void
Surface freeze timing	Skin still soft, gets drawn in	Skin frozen rigid before core shrinks
Mold wall temp	Hotter tool keeps skin soft longer	Cold tool skins the surface fast
Packing pressure	Enough pack fills the shrink	Gate seals early, core starved
Semi-crystalline resin	Amorphous shrinks less overall	PP/PA/POM shrink hard, void readily
Section geometry	Thick rib near a face	Isolated massive boss or corner

The most important row is the first, because the sink-versus-void outcome is fundamentally a race between the freezing surface and the shrinking core. Pack and hold force extra molten resin into the cavity to replace what shrinks, but that only works while a liquid path exists from the gate to the section — a gate that freezes off too early seals the part and the internal shrink has nowhere to draw replacement material from.

A hotter mold wall keeps the skin soft longer, favoring sink; a cold wall skins the part fast and traps shrinkage as a void. Semi-crystalline resins like PP, nylon and acetal shrink more than amorphous ones like ABS or PC, so they demand more packing and are more prone to both defects. Geometry decides where the defect appears: a thick rib under a cosmetic face is a sink generator, while an isolated fat boss or corner cores out into a void.

How Do You Eliminate Sink and Void?

Four steps, in order:

Fix the design first — hollow the thick section. The most reliable cure is to remove the cause at DFM: a boss or rib that is two to three times the nominal wall is a defect waiting to happen, so core it out, add a recess, or use a gusseted rather than solid geometry so no isolated mass sits under a face. Keep nominal wall as uniform as you can and blend thickness transitions gradually. A design with no massive section has nothing to sink and nothing to void.
Size the gate and packing so the section stays feedable. Sink and void are often a starvation problem: the gate freezes before the thick core does, sealing the part while it still wants to shrink. Enlarge the gate or move it so a molten path remains into the heavy section until it has solidified, and tune hold pressure and hold time so resin keeps packing that section through its shrink. The goal is simple in principle — never let the gate close on a section that has not finished packing.
Manage the thermal balance around the feature. Cooling is the lever that chooses between sink and void. Keep mold-wall temperature even around the thick section so it freezes as uniformly as possible, and consider conformal or localized cooling near a heavy boss to pull its core down without over-chilling the surface. A hotter tool can actually reduce sink on a cosmetic face by keeping the skin soft long enough for packing to fill the dimple before it freezes.
Match the resin and confirm on sectioned parts. If the design genuinely needs a heavy section, pick a lower-shrinkage or nucleated grade where the duty allows. Then verify the answer physically: cut the thick feature open and look for internal voids that no surface inspection reveals, and check the cosmetic face under angled light for shadow-sinks along hidden ribs. A part that looks clean can hide a voided boss, so sectioning T1 samples through the heavy features is the honest acceptance test.

Shrinkage — The dimensional reduction from mold cavity to cooled part. Amorphous resins: 0.4–0.7% (uniform). Semi-crystalline: 1.0–2.5% (directional). Mold makers apply shrinkage factor when machining cavity. Actual shrinkage varies with wall thickness, gate location and packing pressure. The related guide covers adjacent topics.

Sink and void are the same root cause wearing different clothes — unequal cooling in a thick section. The fix is never “add more pressure”; it is removing the bulk, feeding the shrink, or balancing the cooling.

Key parameter identified — critical variable confirmed against spec.

Tolerance within achievable range — commercial ±0.15mm or tighter as needed.

Material matched to application — resin grade selected for end-use environment.

Process window validated at T1 — parameters locked after first trial.

Inspection method defined — CMM, visual, or functional test specified.

Frequently Asked Questions

Is a vacuum void always visible on the surface?

No, and that is precisely what makes voids dangerous. A sink mark is by definition a surface dimple, so you can see it under raking light; a vacuum void is an internal hollow that can hide entirely beneath an apparently perfect surface, especially in a thick opaque boss or corner where the frozen skin bridges over the shrinkage inside.

The part looks fine, weighs slightly light, and then fails a strength or fastener-pull test because the material you thought was there is a cavity. This is why sectioning heavy features on trial parts is a real check and not paranoia — a boss that must carry a screw or a load has to be cut and inspected once to prove it is solid.

Why does sink only show on one face of the part?

Because a sink appears on whichever face is nearest and softest above the thick feature, and that face is often the cosmetic A-side while the B-side hides the same geometry against a rib or a mating surface. The shrink pulls the closest free surface inward, so a boss or rib placed near the show face dimples that face, whereas the same bulk under a thick, stiff, non-cosmetic wall may not visibly sink at all.

This is a design lever: moving the thick feature away from the cosmetic surface, or bringing a rib to the far side, relocates or hides the sink even when you cannot eliminate it. Understanding that sink follows the nearest mobile surface is why wall placement and feature placement are settled at DFM.

Can more packing pressure fix every sink mark?

No, and over-packing creates its own set of defects, which is why “just add pressure” is not an engineering answer. Extra pack helps only while a molten path exists from gate to section; once the gate or the section skin has frozen, more pressure cannot feed shrinkage that is already sealed off, so it simply stresses the part, sticks it to the core, flashes the parting line, or locks it so hard it will not eject.

Beyond the feedable window, the real fixes are: thin and hollow the offending section, keep the gate open longer, balance cooling, or change resin. Packing is one of several coupled variables, and chasing a sink mark by pushing pressure alone usually trades a cosmetic dimple for a dimensional or ejection problem.

Key Takeaways

Sink and void are the same root cause — unequal cooling and shrinkage in a thick section — manifesting as a surface dimple or an internal hollow depending on whether the skin freezes soft or rigid.
Fix the design first — hollow the thick section, keep nominal wall uniform, blend transitions gradually; a design with no massive section has nothing to sink and nothing to void.
Size the gate and packing so the section stays feedable, manage thermal balance around the feature, and confirm on sectioned T1 samples — a part that looks clean can hide a voided boss.

Sink and void are not random defects — they are the predictable consequence of thick sections that cool unevenly and shrink without compensation.
Hollow the bulk, feed the shrink, balance the cooling, and verify on sectioned parts. A design with no thermally troublesome mass has nothing to sink and nothing to void; every process band-aid downstream is weaker than removing the cause in the CAD.