Injection Molding Cycle Time Explained: Fill, Pack, Cool, and Eject

A molding cycle runs through four main stages: filling the cavity, packing it under pressure, cooling the part, and opening the mold to eject it. Their durations depend on the part, resin, and mold. Screw recovery happens too, but it can overlap cooling, which is why adding every operation’s time can overstate the cycle.

What Is Injection Molding Cycle Time?

Autodesk defines cycle time as the total time required to complete all the stages of the molding cycle. In practice, a machine timer needs a start and a stop, such as from the start of one injection to the start of the next. Pick one and use it consistently, since a shop that measures differently cannot compare numbers with you.

Machine cycle and part handling are separate. A robot that takes parts away and places them on a conveyor may do most of that work after the mold closes again, so its motion does not always add to the machine cycle.

Overlap is the third piece of the definition. A cycle is not a simple queue of operations, as the machine runs several at once.

Filling Stage

Melt moves from the barrel into the cavity. Autodesk describes fill time as the time needed to fill the mold, with the machine controlling the velocity (flow rate) of the melt.

Injection speed is a process choice with consequences on both sides. Too slow, and the melt front cools before the cavity is full. Too fast, and shear heating or trapped air can cause defects.

Flow resistance depends on wall thickness, flow length, gate size, and resin viscosity. Fill completion is the switch point, which can be triggered by screw position or cavity pressure, at which the machine stops controlling velocity and moves on to pressure control.

Packing and Holding Stage

Pressure continues after the cavity is volumetrically full. The aim is to compensate for shrinkage as the polymer cools. Autodesk notes that from 5 to 25 percent more material can be added to the mold during this stage.

The pressure profile can stay flat or step down, and it changes how much material enters and where. Autodesk also treats packing time and holding time as the same stage.

Gate freeze marks the useful end. Autodesk states that the gate should freeze during the packing time to prevent material from exiting the mold. Holding beyond gate freeze adds cycle time without adding material, while ending before gate freeze can let melt flow back out. A gate-freeze study (weighing parts at increasing hold times until the weight stops rising) finds this point directly.

Cooling Stage

Physical cooling begins as molten polymer contacts the mold surface during filling. The separately timed cooling stage comes later: Autodesk defines it as the stage when no more pressure is being applied, with the mold held shut while the polymer cools until the part can be ejected. Autodesk says this stage is normally the longest part of the cycle and can account for up to 80 percent of it. That is an upper bound in Autodesk’s description, not a target or a typical value for your job.

Four factors shape it:

• Part wall thickness: the thickest section controls, since heat has to travel out through the plastic itself
• Mold temperature: a colder mold removes heat faster, with possible effects on surface finish and shrinkage
• Cooling channels: their distance from the cavity and their flow condition determine how well the steel draws heat
• Temperature at ejection: the part has to be stiff enough to survive the ejector pins

Thickness dominates because plastic conducts heat poorly. If you thicken a wall during design, expect cooling time to rise disproportionately.

Mold Opening and Ejection

Autodesk groups the following under mold open time:

• Opening the mold
• Ejecting the part
• Preparing for the next cycle, such as loading inserts, which is not always part of the cycle
• Closing the mold

Mold movement speed is limited by mold size, slides, and how carefully the mold has to close to protect the parting line. Ejector operation has to clear the part from the core. A robot or part drop has to confirm the part is gone before the mold closes, because closing on a part can damage the tool.

Which Operations Can Overlap?

Screw recovery is the clearest overlap. The screw rotates and retreats to prepare the next shot, and this happens while the part in the mold is cooling. Fictiv, a manufacturing supplier, says in its cycle time article that recovery usually runs in parallel with cooling and adds nothing to total cycle time, unless plasticating takes longer than the cooling phase. When it does, the cycle can stretch to fit, a condition RJG calls recovery limited. How much depends on what the machine can run at the same time, as the worked timeline below explains. Shot size sets how much the screw has to prepare, so how barrel capacity affects recovery and residence time is the place to go if recovery is the constraint.

Robot timing can overlap as well. Machine-control timing differences are another factor, as controllers can start or end timers at different events, so two machines displaying the same “cooling time” may not be measuring the same interval.

Reading a Cycle-Time Chart

A cycle chart is useful only if each bar uses real start and end events.

1. Mark when each operation begins and ends, using the machine’s event log or a process monitor.
2. Draw each on its own row against one time axis.
3. Find the total by looking at the first start and the last end.

Do not add the rows. Any operation that runs alongside another would be counted twice, and the sum would describe a cycle that does not exist.

Why Minimum Cycle Time Is Not Always the Best Cycle Time

Shortening a cycle can move cost and problems somewhere else.

• Incomplete packing: ending hold early can let parts come out light or with sink
• Premature ejection: a warm part can take pin marks or stretch on ejection
• Part distortion: parts ejected warmer have more shrinkage left to happen outside the mold, which is one route to the shrinkage differences that cause warpage
• Excessive residual heat: a part that is too warm can deform when stacked
• Quality variation: a process running near a limit can pass on one shift and fail on the next

The better question to ask is how short a cycle can run while the parts stay inside specification with some margin. Finding that out takes a controlled reduction, measuring parts at each step.

Worked Cycle Timeline

These times are assumed for illustration and are not process data for any part.

Operation	Start (s)	End (s)	Adds to cycle?
Fill	0	2	Yes, 2 s
Hold and pack	2	8	Yes, 6 s
Cooling	8	26	Yes, 18 s
Screw recovery	10	20	No, inside cooling
Mold open, eject, close	26	30	Yes, 4 s
Ejection stroke	27	28	No, inside mold open

Machine cycle: 2 + 6 + 18 + 4 = 30 seconds. Adding every row would give 2 + 6 + 18 + 10 + 4 + 1 = 41 seconds, which is 11 seconds too long. At that error, any plan built from the total would be off by more than a third.

In this example, the timed cooling interval takes 60 percent of the cycle, though the plastic has been losing heat since filling began. The table assumes one sequence: the mold opens when cooling ends, and the next injection starts at 30 seconds. Recovery finishes at 20 seconds, so it does not control this cycle. If it ran past 26 seconds, the result would depend on the machine. A controller that holds mold opening until metering ends would delay the opening. A controller that runs both at once, as US patent 4,847,779 describes, could open on time, and recovery would then limit the cycle only if it ran into the next injection. Reducing recovery time here would not shorten the cycle at all, while taking 2 seconds off cooling would.

Cycle Time FAQ

Can screw recovery continue after the mold opens?

It can on a machine built to do so. The patent cited above describes a control method that opens the mold while metering continues, when cooling ends first. Whether your press allows this is a question for its manual and controller settings. Recovery still has to finish before the next injection can start.

Why can two molds using the same resin have very different cycle times?

The resin is only one input. Maximum wall thickness, cooling channel layout, mold temperature, and how much the part can tolerate at ejection all differ, and wall thickness alone can make one part take much longer. A slide or insert loading step can add mold-open time that the other tool does not have.

Where to Look First for Time

Chart the cycle from real machine events, then find the operation that sets the length. If it is cooling, look at the thickest section and the channel layout before touching machine settings. If the next step is waiting on recovery, the screw is the limit, and work on cooling will not help. If mold-open time looks long, check whether slides, robot motion, or a conservative mold-protect setting account for it. Test each reduction against measured parts, and stop where dimensions or appearance begin to drift. Chart the cycle again after each change, since shortening one operation can simply hand control to another. Autodesk’s 80 percent cooling figure is an upper bound, so your own chart decides where the time sits.