Cooling Time in Injection Molding: The Calculation That Sets Cycle Time

Cooling is most of the cycle

On a typical injection moulding cycle, cooling accounts for 50–70% of the total time. Injection is 2–5%, packing and holding is 15–25%, and mould open, ejection and close is 10–20%. If you want to shorten a cycle, cooling is where the time is — and it is also where most of the quality risk is, because cooling too fast creates warpage, sink and internal stress.

What makes cooling unusual is that it is almost entirely predictable. Unlike packing, which depends on part geometry in ways that are hard to model, cooling follows a clean analytical relationship that you can calculate on paper before the mould is cut.

The calculation

The standard formula for cooling time of a flat plate is:

t = (h² / (π² · α)) · ln( (4/π) · (T_melt − T_mould) / (T_eject − T_mould) )

Where h is the wall thickness (in metres, but usually half-thickness is used depending on the convention — be consistent), α is the thermal diffusivity of the polymer (m²/s), T_melt is the melt temperature, T_mould is the mould temperature, and T_eject is the temperature at which the part is stiff enough to eject.

The practical version, which is what most moulders actually use, is:

t = (s² / (π² · α)) · ln( (8/π²) · (T_melt − T_mould) / (T_eject − T_mould) )

Where s is the full wall thickness. Both forms appear in the literature; the difference is the geometry assumption (plate cooling from one face vs two). For most parts, the second form with full thickness gives a realistic answer.

Table 1 — Thermal diffusivity and typical temperatures. The values are close enough across resins that thickness dominates the calculation far more than material choice.

A worked example

Take a PP part with a 2 mm wall, melt at 220 °C, mould at 40 °C, ejected at 90 °C.

Using α = 0.10 mm²/s, s = 2 mm:

t = (4 / (9.87 × 0.10)) × ln( (0.81 × 180) / 50 )

t = 4.05 × ln(2.92) = 4.05 × 1.07 = 4.3 seconds

Now double the wall to 4 mm:

t = (16 / 0.987) × ln( (0.81 × 180) / 50 ) = 16.2 × 1.07 = 17.3 seconds

The wall doubled, the cooling time quadrupled. That is the relationship in action, and it is why a part with a 4 mm section in a 2 mm wall has a cycle time set by the 4 mm section.

Table 2 — Calculated cooling times. These are theoretical minimums for a flat plate with ideal cooling; real moulds add 20–50% because of uneven cooling and ejection constraints.

The gap between calculated and actual is where most cycle-time improvement lives. A mould that calculates at 4.3 s and runs at 9 s has a cooling problem — usually unbalanced circuits, poor turbulence, or a hot spot from a thick section.

What actually shortens cooling

There are only four levers, and three of them are about the mould rather than the process.

Table 3 — The four levers. The turbulence one is the most underused: many moulds run water in laminar flow and lose 40% of the achievable heat transfer.

Turbulence deserves the emphasis. Water in a 10 mm channel flowing at 3 l/min is laminar; the same channel needs roughly 12–15 l/min to be turbulent. The heat transfer coefficient in turbulent flow can be three to five times higher, which directly reduces the mould surface temperature recovery time between shots.

The practical check is a flow meter on each circuit and a ΔT measurement across it. A circuit with a ΔT above 3 °C is either flowing too slowly or has a blockage. Both mean the mould is not cooling as designed.

Thick sections, bosses and the parts that set the cycle

A part's cooling time is set by its thickest section, not its average. This is the single most common reason a cycle is longer than expected.

Table 4 — Features that extend the cycle. The boss rule is the one worth memorising: a boss thicker than 1.2× the nominal wall starts to dominate.

The design response is always the same: replace thickness with geometry. A 4 mm boss that could be a 2 mm wall with four ribs is not just faster to cool — it is lighter, less prone to sink, and uses less material. The cooling time argument is usually the one that convinces a designer, because it translates directly into part cost.

Mould temperature is a quality decision, not a speed decision

Lowering mould temperature shortens the cycle and damages the part. The relationship is not linear and the failure modes are specific.

Table 5 — Mould temperature ranges. For semi-crystalline resins, a cold mould means incomplete crystallisation, which means the part keeps shrinking after it is measured.

The semi-crystalline resins are the sensitive ones. PA, POM and PBT need a hot mould not for surface quality but for crystallinity. A part moulded in a cold tool has an amorphous skin and a partially crystalline core, and it continues to crystallise and shrink for days. Dimensions measured at the machine will not match dimensions measured a week later, and the customer will find out first.

Verifying the cooling time in the tool

The calculation tells you what is possible. The tool tells you what is happening. Three measurements close the gap.

Cavity surface temperature. An infrared probe or a contact thermocouple on the cavity surface immediately after ejection. If it is above the resin's ejection temperature, the cooling time is too short or the circuit is not working.

Circuit ΔT. Temperature difference between inlet and outlet on each circuit. Above 3 °C means insufficient flow. Above 5 °C means a real problem.

Part temperature at ejection. Measured with a contact probe on the thickest section. If it is above the recommended ejection temperature, the part will warp after ejection and the cycle is being pushed too hard.