Cycle Time Optimization — How Cooling, Injection, and Part Design Affect Per-Part Cost

A production manager looks at a press running a 28-second cycle and asks the classic question: can we get to 24? The answer is rarely found in the injection phase. Cooling — waiting for the melt to solidify to ejection temperature — consumes 60–80% of the total cycle on most parts. The injection, holding, and ejection phases are measured in tenths of seconds; cooling is measured in seconds, and it is the only phase that scales with the square of wall thickness.

This post is the cycle time playbook: where the time actually goes, the cooling time calculation with a worked example, and the levers — process, tooling, and design — ranked by payback. If you need the volume economics first, see cost reduction at 500K+ pieces per year; this post is the technical deep dive on the single biggest lever.

Where the Cycle Time Goes

A typical cycle decomposes into five phases:

Two facts drive everything: cooling dominates, and cooling time scales roughly with the square of wall thickness. A 2.0mm wall cools in about 44% of the time of a 3.0mm wall — not two-thirds. That square relationship is why wall thickness is the highest-leverage cycle variable in the entire process.

The Cooling Time Calculation

The standard engineering estimate for cooling time is derived from one-dimensional heat conduction through the part wall:

tc = (s² / (π² × α)) × ln((8/π²) × ((Tm − Tw) / (Te − Tw)))

Where:

• tc = cooling time (seconds)
• s = part wall thickness (mm)
• α = thermal diffusivity of the material (mm²/s) — for ABS ~0.09, for PP ~0.08, for PC ~0.11
• Tm = melt temperature (°C)
• Tw = mold wall temperature (°C)
• Te = ejection temperature (°C)

Worked example — ABS housing, 2.5mm wall:

• s = 2.5 mm, α = 0.09 mm²/s, Tm = 230°C, Tw = 50°C, Te = 90°C

tc = (2.5² / (π² × 0.09)) × ln((8/π²) × ((230 − 50) / (90 − 50))) tc = (6.25 / 0.888) × ln(0.8106 × 4.5) tc = 7.04 × ln(3.648) tc = 7.04 × 1.294 = 9.1 seconds

Now the same part at 3.0mm wall: s² scales from 6.25 to 9.0 — cooling time becomes 13.1 seconds. A 0.5mm wall change costs 4 seconds of cycle time. On a 1M-piece program at $20/hour machine cost, that single design decision is worth roughly $22,000 per year in machine time.

The formula is an estimate — it ignores runner thickness, mold temperature distribution, and conformal cooling effects — but it is the correct order-of-magnitude tool, and it explains why wall thickness dominates every other cycle lever.

The Levers, Ranked

1. Wall Thickness (Design — Highest Impact)

Cooling scales with s². Every design review should ask two questions: is the nominal wall as thin as the structural requirement allows, and are the thick sections necessary? Common offenders: bosses with excess material around inserts, thick ribs (should be 50–70% of nominal wall), and aesthetic bosses that can be cored out.

The design rules live in the wall thickness design guide. The cycle math in this post is why they matter financially.

2. Cooling System Design (Tooling — Highest Payback Investment)

Standard drilled cooling lines cool from discrete channels; the heat must conduct through the steel to reach them. Conformal cooling channels follow the part contour, cutting the conduction distance and evening the temperature — typically reducing cycle time 20–40% on thick or complex parts. The investment is in the insert; at volume, payback is months. See conformal cooling and cooling system design.

Baffles and bubblers are the intermediate step for deep cores — see the ejector system guide for how cooling and ejection interact in deep parts.

3. Mold Temperature (Process — The Fast Lever, With Limits)

Lowering mold wall temperature shortens cooling directly — the formula’s (Te − Tw) term. The limits are material-specific: too cold a mold freezes the skin before packing completes, producing sink, poor weld line strength, and dimensional instability. The optimization is to run the mold as cold as the part quality allows, and to use mold temperature controllers that hold ±1°C — a mold that oscillates ±5°C forces a conservative cycle to protect quality.

4. Packing and Holding Time (Process — Free Seconds)

Holding pressure is maintained until the gate freezes; after freeze-off, holding time adds cost and nothing else. Gate freeze time can be measured by weighing parts at progressively shorter holding times — the holding time where part weight stabilizes is the minimum. Many tools run 2–3 seconds of unnecessary holding because nobody ran the weight study.

5. Runner and Gate Design (Tooling)

The runner system cools with the part, and the thickest section governs total cooling — a thick cold sprue can add seconds to every cycle. Hot runners eliminate the sprue entirely and shorten the cycle, which is part of their payback case at volume.

6. Machine Speed and Automation (Equipment)

Mold open/close time, ejection speed, and part removal are machine-controlled seconds. A slow ejector retract or a hesitant robot adds 1–2 seconds per cycle that no amount of cooling optimization recovers. Automated part removal on 80% of our press fleet exists for exactly this reason — the machine should never wait for a human.

The Optimization Sequence

Optimizing in the wrong order wastes effort — cutting injection time by 0.2 seconds while cooling runs 9 seconds changes nothing meaningful. The correct sequence:

1. Measure the baseline — shot-by-shot process logs give the real phase split, not the assumed one
2. Run the holding-time weight study — free seconds, zero tooling cost
3. Optimize mold temperature — as cold as quality allows, held stable
4. Attack wall thickness in design — the square law makes this the biggest single lever
5. Upgrade cooling design — conformal or baffles for the remaining hot sections
6. Tighten the machine — open/close, ejection, and removal to their mechanical limits

Each step has a measured payback, and each compounds: a 28-second cycle at step 1 typically reaches 22–24 seconds by step 6 — a 15–20% capacity increase on the same press.

Frequently Asked Questions

How much of the cycle is actually cooling? 60–80% on typical parts. It varies with wall thickness and material, but the rule holds across commodity and engineering resins — which is why cycle optimization is mostly cooling optimization.

Is the cooling time formula accurate? It is the standard first-order estimate and typically lands within 10–20% of measured values. Its real value is comparative: it shows how wall thickness, mold temperature, and ejection temperature move cooling time before any tooling is built.

Why can’t I just run the mold colder? Because the skin freezes before the part packs. Sink marks, poor weld strength, and warpage appear when mold temperature drops below the material’s quality window. The optimum is the coldest mold that still packs the part — measured, not guessed. See warpage and sink marks for the failure modes.

Does conformal cooling pay on simple thin-wall parts? Usually not — thin walls already cool fast, and the conduction distance is short. Conformal cooling pays on thick sections, deep cores, and complex geometries where the cooling time is conduction-limited.

How do I verify the supplier’s claimed cycle time? Demand the trial data: shot log, phase times, and process parameters from the tool trial. That data is the production baseline — see supplier qualification and the mold trial process.

What is a realistic cycle reduction target? An unoptimized program typically finds 10–20% through the sequence above — holding time, mold temperature, and wall thickness doing most of the work. A fully optimized program fights for single seconds. The difference is whether optimization happened before the tool was cut.

The Cycle Time Question in One Line

Every second of cycle time is machine capacity: the question is not “how fast can this mold run” — it is “which of the five phases is waiting, and which lever removes the wait for the lowest cost?”

Submit your part for a cycle time analysis — or review our high-volume manufacturing economics where 24/7 operation makes every saved second a direct capacity gain.