Walk through a plastics plant and the same motion repeats all day. Two steel halves close. Melt goes in. Heat comes out. The tool opens. A part drops or a robot lifts it. Then the tool closes again.
That loop looks simple from the aisle. Inside the tool it is a timed conversation between pressure, temperature, and geometry. If you buy molds, run presses, or design parts, the value is not memorizing slogans. It is knowing which stage created the mark you see on the bench.
What the Process Is Really Doing
Injection molding takes solid pellets, turns them into a flowable melt, pushes that melt into a closed cavity, holds the shape while the polymer stiffens, then releases the part so the next shot can start.
The mold is not a passive box. It is a heat exchanger, a pressure vessel, and a forming die in one stack. The cavity and core copy the part. The runner and gate decide how melt arrives. Cooling circuits pull heat. Slides, lifters, and ejectors free undercuts and then push the part off the core.
A short cycle can feel automatic. The process still has a beginning that happens before the first clamp-up.
Before the First Shot: Material and Tool Ready
Pellets do not go straight from a bag into a good part. Many resins pick up moisture from air. Water that rides into a hot barrel turns into vapor. Vapor shows up later as splay, bubbles, or weak knit areas. Drying time and hopper management belong to the process even though they sit outside the “four-step cycle” posters on the wall.
Color and additives get blended before or at the throat. Poor mixing does not wait until assembly to appear. It shows as streaks and shade drift from cavity to cavity.
On the tool side, the stack has to be clean, aligned, and at a stable temperature. A cold tool on the first shots and a tool that has been running for an hour are not the same thermal object. Water circuits need flow. Parting faces need to meet. Ejector plates need a free stroke. If this homework is sloppy, later “process tweaks” chase a problem that started on the bench.
Stage 1: Clamping
The moving half travels in. The parting faces meet. The clamp builds force and holds the stack shut.
That force has one job: keep the mold closed against the pressure of incoming melt. If the clamp is light for the projected area and the fill pressure, the parting line breathes. A thin film of plastic escapes. People call that flash. If the clamp slams or loads one corner of the tool harder than the other, you wear parting edges and you can print thickness variation into the part.
Clamp time is usually short. It still sets the rest of the shot. A dirty parting line, a crushed vent, or a slide that is not home will not be fixed by turning a fill speed knob.
Stage 2: Plasticizing the Next Shot
While one part cools in the cavity, the screw is already working on the next one. Pellets drop from the hopper into the feed zone. Barrel heat and screw shear turn solid into melt. The screw recovers, building a measured cushion of melt in front of the tip.
This stage decides whether the melt arriving at the gate is even. Unmelted pellets, overheated streaks, or trapped gas travel with the shot. The nozzle must seal against the sprue bushing. A drool or a cold slug at that interface becomes a blemish or a blockage later.
Plasticizing overlaps cooling on purpose. That overlap is how the press keeps pace. If recovery is slow, the cycle waits. If recovery is rushed and the melt is not uniform, the cavity pays for it.
Stage 3: Filling the Cavity
The screw now acts like a plunger. Melt moves through the nozzle, sprue, runners, and gate, then across the cavity.
Fill is a race against freeze-off. Thin walls and long flow paths cool the front of the melt as it travels. If the front freezes before the cavity is full, you get a short shot. If the front jets into open space instead of spreading as a fountain flow, you mark the surface. Air that cannot leave through vents gets packed into burns or bubbles.
Gate location is not decoration. It decides weld-line positions, orientation of flow, and which features pack last. A tool with two cavities that fill at different times is not “almost balanced.” It is two different processes sharing one machine setting.
During fill, the clamp is already locked. The steel is already taking load. This is why vent depth, gate size, and runner layout belong in the mold discussion, not only in the press discussion.
Stage 4: Packing and Holding
When the cavity is nearly full, the machine switches from filling by speed to packing by pressure. Extra melt is pushed in as the material in the cavity begins to shrink.
Polymers lose volume as they cool. If that volume is not replaced while the gate is still open, the part sinks in thick bosses, voids form in heavy sections, and dimensions wander. If pack goes on after the gate has frozen, you are only compressing the runner. The cavity no longer sees the pressure.
Holding is the quiet stage that shows up on a caliper. Rib-to-wall junctions, logos that are too deep, and bosses that sit on thick pads all ask for pack they may not receive if the gate freezes early.
A simple way to think about it:
| Stage | What the melt is doing | What the tool is doing | What you usually see if it goes wrong |
|---|---|---|---|
| Clamp | Waiting | Faces closed, force built | Flash, crushed vents, uneven thickness |
| Fill | Racing through the cavity | Steel taking pressure, air leaving vents | Short shots, jetting, burns, visible weld lines |
| Pack / hold | Feeding shrink | Gate still able to pass melt | Sinks, voids, weight scatter, size drift |
| Cool | Stiffening in place | Channels pulling heat | Warp, residual stress, long cycle |
| Eject | Already solid enough to move | Pins / stripper / air releasing the part | Pin marks, cracks, sticking, drag |
Stage 5: Cooling
Cooling is where the clock lives. Melt hits steel and heat starts leaving. Channels in the mold carry that heat away with circulating water or another fluid.
The part cannot leave when the surface looks solid. The core of a thick wall may still be soft. Eject too soon and the part bends, pins punch through, or dimensions relax on the table. Cool longer than the geometry needs and you give away shots you never get back.
Uniformity matters as much as duration. A core that runs hot while the cavity runs cold pulls the part toward the hot side as it shrinks. That is one common path to warp. Deep pockets with no nearby channel become hot islands. Hot islands stay soft and then move after eject.
Wall thickness is the design choice that cooling cannot fully hide. A sudden thick pad next to a thin wall stores heat. The thin wall freezes and the thick pad keeps shrinking. The surface over the pad sinks or the whole face bows.
Cooling design is mold work: channel placement, baffles, bubblers, inserts with higher conductivity, and contact between insert and plate. Process work can trim a few seconds. It cannot invent a channel that was never drilled.
Stage 6: Opening and Ejection
When the part can hold its shape, clamp force drops and the moving half opens. The part usually stays on the core. Ejector pins, sleeves, a stripper plate, or air then push it free.
This stage looks mechanical. It is also a quality stage. Pin area that is too small prints a witness or cracks a thin floor. A pin that is too early in the stroke marks a part that is still tender. A core with a hard polish and no draft holds the part until something tears.
Undercuts need slides or lifters to clear before the pins fire. If a slide is late, the part shears. If an ejector plate cocks, one side of the part leaves first and the geometry twists.
Runners and sprues leave with the shot unless the tool is hot-runner. Those leftovers get trimmed, reground, or discarded according to the quality plan. Regrind that goes back into the hopper is part of the next cycle’s material story, not a footnote.
After eject, the ejector plate returns, slides home, and the clamp closes. The next shot is already waiting in front of the screw.
What Happens After the Cycle
The press cycle ends at eject. The part’s story does not.
Gates get trimmed. Flash, if any, gets removed. Some parts go to annealing, machining, welding, printing, or assembly. Inspection looks at dimensions, appearance, and function. A part that looked fine at the robot gripper can drift after it finishes shrinking on a tray.
That delayed shrink is why packing and cooling settings and tool temperature are discussed together. The cavity copies a size at eject temperature. The part continues toward room temperature on its own.
How Part Design Talks to the Process
A drawing that ignores the cycle forces the tool and the press to compensate.
- Keep walls as even as the function allows so heat leaves on a similar clock.
- Add draft in the pull direction so ejectors do not fight the steel.
- Place ribs thinner than the wall they sit on so they do not sink the face.
- Give bosses a chance to pack: gate strategy and local thickness work as a pair.
- Put weld lines where they do not sit on a snap or a seal.
- Leave a path for air. Dead-end ribs trap gas.
- Avoid knife-edge steel. Thin standing core steel heats up and is hard to cool.
None of this is decoration on a CAD model. Each choice shows up as a fill pattern, a pack window, or a cooling time.
How Mold Construction Talks to the Process
Two tools can close on the same press and behave like different machines.
Gate type changes shear and freeze-off. A pinpoint gate freezes sooner than a larger edge gate. Runner balance decides whether cavity 1 and cavity 4 see the same fill time. Venting at the end of fill decides whether the last corner packs or burns. Steel choice and polish change release and wear. Cooling layout changes both cycle and warp.
Maintenance is process control. A scratched cavity, a blocked channel, a worn vent, or a leaking manifold changes the shot without anyone touching a setpoint. Shops that treat the tool as a living heat-and-pressure device spend less time chasing ghosts on the controller.
Reading Defects Backward
Inspection is late in the story. The useful habit is to walk the defect back to a stage.
- Flash at the parting line points at clamp, parting condition, or fill pressure that the clamp could not hold.
- Short shots point at fill, vents, melt temperature, or a gate that froze too soon.
- Burn marks at the last place to fill point at air that had nowhere to go.
- Sinks over ribs and bosses point at pack, gate freeze, or local thickness.
- Warp after a few hours on the table points at cooling imbalance or orientation from fill.
- Pin push or cracks at eject point at cooling time, pin layout, draft, or stick on the core.
- Splay and silver streaks often point back to moisture or degraded melt, not to the ejector plate.
This backward reading keeps teams from turning every defect into a fill-speed argument.
Questions Teams Actually Ask
How many stages are there?
Count them as four if you only watch the clamp-fill-cool-eject poster. Count six or more if you include drying, plasticizing, packing, and post-trim. The physics does not care about the poster. Packing is not the same event as filling. Plasticizing is not the same event as injection.
Why is cooling so long compared with fill?
Fill moves liquid. Cooling moves heat through plastic and steel. Plastic is a poor conductor. Steel can only take heat as fast as channels and contact allow. That is why a thick handle takes more clock than a thin lid even when both fill in a short burst.
Does a hot runner change the steps?
It changes the leftover plastic and the thermal profile near the gate. It does not remove clamp, fill, pack, cool, or eject. You still need the cavity full, the shrink fed, the heat removed, and the part released.
Where should a new program start?
Start with the part geometry and the tool layout, then set a process that matches that tool. A press recipe copied from another job assumes the same flow length, the same gate freeze, and the same cooling map. Those assumptions fail quietly.
A Working Picture of One Cycle
Picture a small enclosure.
Pellets have already been dried and sit in the hopper. The screw has a shot waiting. The mold closes and the clamp builds. The screw drives forward. Melt races through a runner, hits a gate on the hidden face, and fans across the walls. Air leaves the far corners through vents. The machine transfers to hold and feeds the shrinking bosses. Water moves through channels around the core. The enclosure stiffens. The mold opens. Pins push the box off the core. A robot takes the shot. The runner is clipped. The screw has already recovered. The mold closes on an empty cavity.
That is the process. It repeats because each stage hands a stable condition to the next one. When a stage hands over a problem, the next stage can only hide it for a while.
If you design parts, ask where the gate will sit and how heat will leave the thickest region. If you buy tools, ask how the cavities are balanced and how the channels reach the cores. If you run presses, treat clamp, fill, pack, cool, and eject as five different conversations, not one slider labeled “quality.”
The cycle is short. The cause of a mark on the part is usually not.