Key takeaways
- Injection molding melts plastic, injects it into a closed steel mold, cools it under pressure and ejects the part. The cycle takes seconds; the mold is the real commitment.
- Parts must be designed to leave a mold: draft angles and uniform wall thickness are the two rules that matter most.
- Cooling is roughly 80 to 90 percent of cycle time and scales with the square of wall thickness, so thick walls drive unit cost up fast.
- A molder needs 3D CAD, 2D drawings with tolerances, material, finish and volumes to give a real quote. A description alone gets a wrong number.
Injection molding works by melting plastic pellets, injecting the melt under pressure into a closed steel mold, holding it under pressure while it cools into shape, then opening the mold and ejecting the finished part. The process itself takes seconds once a mold exists. What makes it a serious commitment is the mold: a precision steel or aluminum tool, custom-built for one part, paid for before a single unit comes out of it. It is one of several ways to turn a design into a physical plastic part, and the one best suited to higher volumes once a design is locked.
What happens in one molding cycle
A molding cycle repeats the same sequence every time, whether it runs once for a sample or hundreds of thousands of times for a production order.
- Clamp: the two mold halves close and lock under high pressure.
- Injection: a screw pushes molten plastic through a nozzle into the mold, filling the cavity in one to two seconds.
- Pack and hold: pressure is maintained while the gate freezes, which compensates for the plastic shrinking as it cools.
- Cooling: the part solidifies inside the closed mold. This step takes the largest share of the cycle by far.
- Ejection: the mold opens and ejector pins push the finished part free.
Injection and filling take only a second or two. Cooling is what actually sets how fast, or how slow, a molding job runs, which matters directly for the section on cycle time and unit cost below.
The parts of a mold, in plain language
A mold is an assembly of parts that each do a specific job, and knowing the vocabulary helps when reading a quote or a tooling drawing.
| Part | What it does |
|---|---|
| Cavity | The half of the mold that forms the outside surface of the part. |
| Core | The half that forms the inside surface and any hollow features. |
| Gate | The narrow opening where molten plastic enters the cavity, sized to fill the part properly and then freeze quickly so it does not leak backward. |
| Runner | The channel that carries plastic from the machine’s nozzle to the gate. Oversized runners waste material and add cooling time; undersized ones cause incomplete fills. |
| Ejector pins | Pins that push the finished part out of the core once the mold opens. They typically leave small, visible marks on the part where they made contact. |
Cavity, core, gates, and ejector pins
Several gate styles exist, and the choice depends on the part’s geometry and where a visible gate mark can be tolerated.
- Edge gate: simple, placed on the parting line, suited to flat parts.
- Submarine gate: tucked beneath the parting line so it separates from the part automatically on ejection.
- Pin gate: small and round, common on multi-cavity molds feeding several parts at once.
Cooling channels are drilled directly into the cavity and core plates so water can circulate through them, which controls how evenly and how fast a part cools. Uneven cooling is a direct cause of warping.
Design rules the process imposes
Injection molding is not a manufacturing method you design a part for afterward. The part has to be designed to come out of a mold at all, and two rules matter more than any others. The right point in the schedule to lock these decisions is its own topic, covered in our guide on when manufacturing decisions have to be locked.
Draft angles and why vertical walls fail
A wall with zero draft, meaning perfectly vertical relative to the direction the mold opens, tends to drag, scuff, or stick during ejection because the part grips the steel instead of releasing from it. A design-for-manufacturing engineering reference recommends 1 to 2 degrees of draft per side on vertical walls as a general starting point, rising toward 5 degrees or more as surface texture gets heavier.
- No draft: risks scuffing, warping, or a stuck mold.
- 1 to 2 degrees: the standard starting point for most enclosures and housings.
- 3 to 5+ degrees: needed as surface texture gets heavier, since texture adds friction the part must release against.
Uniform wall thickness and sink marks
Plastic shrinks as it cools, and it does not shrink evenly if wall thickness is inconsistent. A thick section next to a thin one cools at a different rate, and the thick section can pull the surface inward as it solidifies last, creating a sink mark. The same reference recommends keeping wall thickness consistent, commonly 1 to 3 millimeters depending on material, sizing ribs and bosses at roughly 50 to 60 percent of the wall they attach to, and adding a minimum internal fillet radius of around 0.5 millimeters at corners to reduce stress concentration and help plastic flow.
| Design choice | Effect |
|---|---|
| Uniform wall thickness | Even cooling, minimal sink, predictable shrinkage |
| Thick section next to thin section | Sink marks, internal stress, longer cooling time driven by the thickest point |
| Sharp internal corner | Stress concentration and disrupted plastic flow |
| Rounded internal fillet | Reduced stress concentration, smoother fill |
Cycle time and why it drives unit cost
Cooling is not a minor step. Independent process-training sources that study molding cycles put cooling at roughly 80 to 90 percent of total cycle time, driven mainly by the square of the part’s maximum wall thickness: doubling wall thickness roughly quadruples cooling time, not doubles it. That is one of the clearest financial reasons uniform, thin walls matter.
Cycle time converts directly into cost because a molding machine is billed by the hour, and every second added to a cycle multiplies across the full run.
- A 30-second cycle running 500,000 units takes roughly 4,167 machine hours.
- The same run at 32 seconds, just two seconds slower, takes roughly 4,444 machine hours: an extra 277 hours that were never priced into the job.
Common defects and what causes them
Most defects in a molded part trace back to a small number of root causes: too little material, uneven cooling, too much pressure, or a flawed mold.
| Defect | Typical cause |
|---|---|
| Short shot | Incorrect shot size, material too viscous to fill the cavity, or inadequate venting |
| Sink marks | Uneven wall thickness, insufficient cooling time, or inadequate holding pressure |
| Warping | Uneven cooling or inconsistent wall thickness across the part |
| Flash | Insufficient clamping force, excess injection pressure, or worn mold parting lines |
| Weld lines | Two flow fronts meeting weakly, from low pressure, low temperature, or gate placement |
| Voids | Trapped gas, moisture, or insufficient material reaching the cavity |
A first article inspection and ongoing sampling exist to catch these before a full production run scales up the problem, covered in our guide to quality control in manufacturing.
What you need to hand a molder to get a real quote
A molder cannot give a meaningful number from a description. Skip any of the following and the quote usually comes back wrong.
- 3D CAD in a neutral format (STEP or IGES), used to calculate part weight, projected area, and clamp-force sizing
- 2D drawings with dimensions and tolerances called out explicitly, not just implied by the 3D model
- A specific material callout by resin type and, ideally, grade, since “ABS” alone leaves too much unspecified
- A surface finish specification
- Expected annual volume or total lifetime volume, since this drives how many cavities the mold needs and what tooling material makes sense
In our experience at Inventornest, we develop the PCB before finalizing the enclosure so the board’s real dimensions inform the enclosure design, rather than forcing a board to fit an enclosure finalized first, a sequencing choice that avoids a common cause of late redesign.
The honest downsides nobody mentions early
The first downside is that a mold is a large, fixed cost committed before a design is completely final, which is why most programs use 3D printing during development and move to injection molding once the design is validated and volume justifies the spend, the same question covered in our guide on when to switch from low volume to mass production. In our experience at Inventornest, we do not use a fixed unit-quantity threshold for that call: the right point depends on enclosure complexity, material, finish, tolerances, tooling cost, and the actual per-unit savings molding offers over printing.
A simple way to estimate that point is to divide the tooling investment by the savings per unit versus the alternative process. As an illustration, not a supplier quote, a $5,000 tool that saves $10 per part breaks even around 500 units. Sampling, revisions, and freight still have to be added on top before that math reflects a real program.
The second downside is a design change after the mold is cut. The third is mold lead time itself, commonly weeks rather than days, sitting on the critical path of the whole schedule, which is why the decisions behind it need to be genuinely final before it starts.
- Before the mold is cut: a design change means editing a file and re-running a simulation.
- After the mold is cut: the same change means re-machining or rebuilding steel, at a cost and delay that scales with how much of the tool has to change.
Frequently asked questions
How long does an injection molding cycle actually take?
It varies by part size and wall thickness, but cooling alone typically accounts for 80 to 90 percent of the cycle, and it scales with the square of wall thickness rather than linearly.
What draft angle should I design with?
A common starting point is 1 to 2 degrees per side on vertical walls, with more needed as surface texture gets heavier. The exact figure depends on material and mold geometry.
Why do sink marks appear on molded parts?
They usually form where a thick section sits next to a thin one, or where a rib or boss is too thick relative to the wall it attaches to, causing uneven cooling and shrinkage that pulls the surface inward.
What do I need to give a molder to get an accurate quote?
At minimum: 3D CAD in STEP or IGES format, dimensioned 2D drawings, a specific material and grade, a surface finish callout, and an expected production volume.
When should I switch from 3D printing to injection molding?
Once the design is validated through testing and expected volume justifies tooling cost. A breakeven calculation, tooling cost divided by savings per part, gives a starting estimate, though sampling, revisions, and freight still add to it.
Can I fix a defect like warping after the mold is already built?
Sometimes, through cooling, pressure, or temperature adjustments, but a defect rooted in inconsistent wall thickness usually requires re-cutting steel, which is slow and expensive compared with catching it before the mold is built.
Where Inventornest fits
Inventornest designs parts for the manufacturing method they will actually use, developing the electronics before finalizing an enclosure so the two never fight each other later. If your product is approaching the point where tooling decisions need to be made, our product development and manufacturing service can help you get there with a design that actually molds cleanly. Get a quote once you have a target volume and material in mind.
