Cooling time follows the square of the wall
A moulded part has to lose enough heat to become rigid before it can be ejected. The heat has to travel from the centre of the wall to the steel, and heat conduction through a slab takes time proportional to the square of the distance. Double the wall and the cooling time quadruples. That is not a rule of thumb, it is the solution to the heat equation, and it is the single most important number in cycle time.
Measured: 4 seconds at a 1.0 mm wall, 9 at 1.5, 16 at 2.0, 25 at 2.5 and 36 at 3.0. The 1.5 mm to 3.0 mm step doubles the thickness and multiplies the time by four, and a designer who thickens a wall for strength without thinking about the cycle has just doubled the cost of the part.
This is why thin-wall moulding exists as a discipline. A 1.2 mm wall instead of 2.0 mm does not just save material, it takes cooling from 16 seconds to about 6, and on a 30 second cycle that is a third of the output.
Cooling time against mould temperature: 14 seconds at 20 C, 16 at 30, 19 at 40, 22 at 50, 26 at 60 and 31 at 70. The relationship is nearly linear, and it looks like free money: drop the mould temperature and the cycle shortens.
The invoice arrives as distortion. A cold mould freezes the skin before the core has shrunk, so the part is ejected with a temperature gradient through its thickness. The skin is rigid and the core is still shrinking, and the part warps as it equalises. A 20 C mould can save eight seconds a cycle and produce a part that will not sit flat on the customer's assembly fixture.
The gradient also sets moulded-in stress. A part cooled too fast has a stressed skin and a relaxed core, and it will move when it is later heated, welded or painted. That movement is often blamed on the material, and it is really a record of how fast the tool was running.
Where the cooling time actually goes
The heat equation gives the minimum. What a real tool adds is the difference between the ideal and the actual, and it comes from four places.
| Loss | Typical penalty | Fix |
|---|---|---|
| Cooling channels too far from the cavity | 20 to 40 percent | Move them, or add baffles and bubblers |
| Poor flow, so the coolant heats up | 10 to 25 percent | Check flow rate per circuit, not pressure |
| Hot spots at cores and deep ribs | Local, but sets the whole cycle | Bubblers, thermal pins, or split the circuit |
| Mould at the wrong temperature overall | Trade against distortion | Set it from the part, not from the cycle sheet |
Table 1 — What makes real cooling slower than the calculation
The channel distance is the biggest and the most commonly ignored. A channel 30 mm from the cavity removes heat far more slowly than one 15 mm away, and the difference is not linear. A tool designed with channels placed for manufacturing convenience rather than for heat removal will run 30 percent slower than the same part in a well-designed tool, for the whole life of the tool.
Four levers that shorten cooling without changing the part
The wall thickness is fixed by the design and the mould temperature is fixed by the tolerance. What is left is the tool and the machine.
- Coolant flow. Turbulent flow transfers heat several times better than laminar, and most tools run laminar because the circuits are long and the pump is small. Raise the flow before lowering the temperature.
- Channel placement. Closer to the cavity, and conformal where the geometry allows it.
- Ejection temperature. A part ejected 5 C warmer is still rigid enough for most geometries, and the last five degrees are the slowest to remove.
- Two-stage cooling. Cool in the tool to the point of rigidity, then cool the rest in a fixture, which is what a robot cell does when it takes the part out hot.
The fourth lever is the one that changes the economics most, because it moves cooling out of the constraint. A tool that cools to rigidity in 12 seconds and a fixture that finishes the job in 8 is a 20 second cycle that used to be 24, with no change to the part.
How to tell whether the cycle is cooling-limited
A cycle is cooling-limited when the part is still soft at ejection and the operator is waiting for it. It is not cooling-limited when the machine is waiting for the operator, or when the robot is the constraint, or when the tool is open for a second longer than the ejection needs.
The test is a temperature measurement, not a stopwatch. Measure the part surface immediately after ejection with a contact probe on three consecutive shots. If it is above the material's heat deflection temperature, the cycle is short. If it is well below, the cycle is being set by something else and shortening the cooling will buy nothing.
| Symptom | Diagnosis | What to change |
|---|---|---|
| Part warps after ejection | Ejected too hot, or gradient too steep | Longer cooling or warmer mould |
| Part sticks in the cavity | Over-packed, or too cold | Packing profile, mould temperature |
| Cycle runs long but part is cool | Constraint is elsewhere | Look at ejection, robot or operator |
| Ejection marks on the part | Ejected too hot and too hard | Cool longer, slow the ejector |
| Dimensions drift through a run | Mould temperature drifting | Check the chiller and the flow |
Table 2 — Reading a cooling problem from what the part does
What to specify
A cooling specification that can be checked names four things and one measurement.
- Wall thickness, with the tolerance that matters for cooling, not just for strength.
- Mould temperature, set from the material and the tolerance rather than from the cycle target.
- Coolant flow rate per circuit, not supply pressure.
- The ejection temperature, measured on the part rather than inferred from the timer.
- Channel distance from the cavity surface, which is a tool drawing item and is usually missing.
Then measure the part, not the clock. A cycle time that was optimised by shortening the timer until parts failed is a cycle time that will fail again on the first cold morning.
Why does cooling time quadruple when wall thickness doubles?
Heat conduction through a slab takes time proportional to the square of the distance. Double the distance and the time goes up four times.
How much time does a 1 mm wall save against 2 mm?
About 12 seconds: 4 seconds at 1.0 mm against 16 at 2.0 mm.
Does a colder mould shorten the cycle?
Yes, and it also increases the thermal gradient through the part, which causes warping and moulded-in stress.
What is the biggest cause of slow cooling in a real tool?
Cooling channels placed too far from the cavity, which typically costs 20 to 40 percent of the cooling time.
Should I raise flow or lower temperature?
Raise flow first. Turbulent flow transfers heat several times better than laminar, and most tools run laminar because the circuits are long.
How do I know the cycle is cooling-limited?
Measure the part temperature right after ejection. If it is above the heat deflection temperature the cycle is short; if it is well below, something else is the constraint.
A cycle that will not come down?
Send the wall section, the current cycle and the mould temperature. We will tell you whether the limit is cooling, ejection or something outside the tool.
Continue reading
- Mould Temperature Control: The Variable That Sets Everything Else
- Clamp Force: You Cannot Size a Tonnage From a Table
- Packing Pressure Profile: The Half Second That Decides the Part
- Warpage in Injection Moulding: Differential Shrinkage in Practice
- Custom Injection Moulding: Tooling, Materials and Tolerances
AR 
EN 
ES 
FR 
DE 
PT 
RU 
JA 
KO 
IT 
NL 
TR 
PL 
ID 