Injection Molding Cost per Part: How the Price Is Really Built

How a piece price is actually built

A moulded part price is not a market rate and it is not a guess. It is the sum of five cost blocks divided by the number of good parts you get per hour. Once you can see the structure, you can see which levers are worth pulling — and, more importantly, which quotation is quietly unrealistic.

Shares are for a mid-size engineering part in the 20 – 300 g range, produced on a 200 – 500 t machine. Small precision parts shift towards machine time; large commodity parts shift towards material.

Notice that material and machine time together are 60 – 90 % of the price. That is why almost every meaningful cost reduction comes from changing the part weight, the wall thickness, the cycle time or the cavitation — not from negotiating the supplier's margin.

The machine rate, and why tonnage is not the main driver

The machine rate is what it costs to have a machine occupied for an hour: depreciation, energy, maintenance, floor space, and the operator and supervision attributable to it. It rises with tonnage, but not proportionally, and it rises steeply with special features.

Indicative European and North American rates for a standard hydraulic or servo machine. Asian rates are typically 40 – 60 % of these; the structure is the same.

Two things surprise buyers here. First, the rate difference between a 200 t and a 400 t machine is much smaller than the tonnage ratio suggests — roughly 1.4× for double the tonnage. So running a part on a machine that is somewhat too large is wasteful but not catastrophic; running it on a machine with the wrong shot capacity is worse.

Second, the shot-size rule matters more than the tonnage rule. A machine's barrel and screw are sized for a shot range, typically 20 – 80 % of the maximum shot capacity. Running a 30 g part on a machine with a 1,000 g maximum shot means the melt residence time in the barrel is very long, which degrades the material, and the process becomes unstable. The correct machine is the smallest one that fits the clamp force and the shot size, not the biggest one that is free.

How much clamp force do you actually need?

Clamp force has to exceed the separating force from injection pressure acting on the projected area of the part plus the runner:

Cavity pressure is the number people get wrong. It is not the injection pressure at the nozzle; it is the pressure inside the cavity after packing. Typical values are 300 – 500 kg/cm² for commodity resins, 500 – 800 for engineering resins, and up to 1,000 for precision work. Overestimating it costs money in machine time; underestimating it costs money in flash.

Cycle time: the four phases and which one you can shorten

Cycle time is the clock that multiplies the machine rate. It has four phases, and they are not equally controllable.

Cooling dominates, and it scales with the square of wall thickness. That single relationship explains more about moulding economics than anything else in this article.

The cooling rule of thumb

Cooling time is roughly proportional to the square of the thickest wall section. Doubling wall thickness quadruples cooling time. Halving it cuts cooling to a quarter. For a part where cooling is 60 % of a 30 s cycle, dropping the wall from 3 mm to 2.4 mm can take the cycle from 30 s to about 23 s — a 23 % reduction in machine time on every part, forever.

Cooling time calculated for ABS at 80 °C tool temperature, 230 °C melt, ejecting at 90 °C. Total cycle adds fill, pack, hold and mould movement.

This is the single most useful table in this article. A designer who thickens a wall from 2 mm to 3 mm to "make it stronger" has just increased the piece price by roughly 80 % — and a ribbed 2 mm wall would have been stiffer than the 3 mm flat wall anyway.

What else moves cycle time

• Tool steel and cooling channel design. Beryllium copper inserts in hot spots, conformal cooling, and baffles in deep cores can cut cooling time 20 – 40 % on difficult parts. It costs more up front and pays back within months at volume.
• Coolant temperature. Colder water cools faster but increases warpage and internal stress. There is an optimum, not a minimum.
• Robot take-out. A robot that removes the part during mould close removes the ejector-return time from the cycle. On a 20 s cycle this can save 1.5 – 3 s — 10 % of the price.
• Hot runner. Eliminates sprue cooling and degating, typically saving 3 – 8 s per cycle and removing the runner from the material cost.

Cavitation: the single biggest lever

Machine time per part is the cycle time divided by the number of cavities. Going from one cavity to four cuts the machine cost per part by 75 %. Nothing else in moulding has that leverage.

Tool cost multipliers are indicative for a simple part with no side actions. The multiplier is lower for complex tools because the base cost dominates.

So why is any tool single-cavity? Four reasons, and they are all real:

• Volume. The extra tool cost has to be paid back by the machine-time saving. At 2,000 parts a year, an 8-cavity tool never pays back.
• Tolerance. Cavities are not identical. On a ±0.05 mm dimension, eight cavities means eight processes that all have to be capable.
• Part size. A large part may not fit more than one cavity inside the clamp force available.
• Risk. One cavity can be replaced; eight cavities means eight times the tool maintenance and a bigger single point of failure.

The payback calculation is simple and worth doing explicitly:

Material cost and the runner you pay for twice

Material is usually the largest single block, and it is the one buyers most often assume is fixed. It is not: it depends on part weight, wall thickness, runner design and how much regrind the supplier can legitimately use.

The shot weight calculation

You pay for the shot, not the part. A shot is the part plus the runner plus the sprue:

Hot runners remove that cost almost entirely. On the example above, a hot runner saves $0.029 per part in material plus a few seconds of cycle time. At 500,000 parts a year, that is $14,500 in material alone — which is why hot runner tools pay back quickly at volume and never pay back at low volume.

Regrind

Most suppliers will blend regrind back into the virgin material, typically 10 – 25 % for commodity resins and 0 % for medical, food-contact and optical parts. Regrind is cheaper than virgin resin, so a supplier who uses it can quote lower — but regrind degrades with each pass, and above about 25 % the mechanical properties and colour consistency start to drift.

The commercial point: if your drawing says "100 % virgin", you are paying for it, and you should know that. If it says nothing, you are probably getting 10 – 20 % regrind and you should confirm the limit in writing.

Indicative 2026 bands for standard moulding grades in tonne quantities. Colour masterbatch, UV stabilisers, flame retardants and glass loading all move the price.

Density is the hidden variable. POM at $3.20/kg looks cheaper than PC at $4.00/kg until you notice POM is 1.41 g/cm³ and PC is 1.20 — per cubic centimetre of part, POM costs $4.51 and PC costs $4.80, which is a much smaller gap. Always compare cost per part volume, not cost per kilogram.

Scrap, start-up and the parts you never see

No process produces 100 % good parts. The yield factor is the ratio of saleable parts to parts moulded, and it hides inside every quotation whether or not the supplier admits it.

On a 10,000-part run, a 30-shot start-up and 1.5 % rejects is about 180 parts of loss — 1.8 % of the order, added to the price.

This is why short runs cost more per part than long runs, and why the same part quoted at 5,000 and at 50,000 comes back with different prices. It is not arbitrary; the start-up loss is spread over ten times fewer parts.

It is also why cosmetic standard matters so much commercially. A part with a strict cosmetic requirement may run at 3 % rejects where the same part at a relaxed standard runs at 0.5 %. That 2.5 % difference is real money, and it is entirely a decision made on the drawing.

Labour, packaging and freight

Labour is smaller than most buyers expect on an automated machine and larger than they expect on a manual one. A single operator can run several machines if parts drop free or are robot-taken; a part that needs manual loading of inserts, manual degating or manual inspection is a different cost structure.

Indicative for European labour rates. Asian rates are substantially lower but the structure and the ranking are the same.

Packaging is small per part but not negligible: an inner bag, a divider layer, an outer carton and a pallet add up to $0.02 – 0.15 per part depending on how fragile the part is and how the customer's warehouse receives goods. Freight is priced per kilogram and per cubic metre, so a part that nests well ships cheaper than one that does not — which is a design decision made years before anyone thinks about freight.

Tool amortisation: when it belongs in the price

There are two legitimate commercial models and one bad one.

Separate tool, separate parts. The customer pays for the tool as a capital item and pays a piece price that reflects only conversion and material. This is the cleanest model: the customer owns an asset, the piece price is comparable across suppliers, and if they move the tool the price does not change.

Amortised tool. The tool cost is spread across a committed volume and included in the piece price. This is legitimate when the customer cannot fund the tool up front, and it is normal for very large programmes. The piece price is higher, and the customer should receive a price reduction once the amortisation is complete — which should be written into the agreement.

The bad one. Tool cost buried in the piece price with no stated amortisation volume and no price reduction later. The customer pays for the tool many times over and never owns it. This is the most common commercial problem in injection moulding, and the defence is simple: ask for the tool cost and the amortisation volume in writing, separately.

Three worked examples

Numbers make this concrete. All three use a $45/h machine rate, ABS at $2.40/kg, and a 2 % yield loss.

Example A: 45 g enclosure, 4 cavities

A typical consumer electronics enclosure. Material is 37 % of the price, machine time 26 %.

Example B: same part, 2.0 mm wall instead of 3.0 mm

Same tool, same material, same machine. A 1 mm wall reduction is worth 26 % of the piece price — and with ribs, the part is stiffer than the original.

Example C: 12 g connector body, 16 cavities vs 4 cavities

The bigger tool looks obviously better and is obviously wrong at this volume. The payback is longer than most programmes last.

These three examples cover the three decisions that set 80 % of a moulded part's price: wall thickness, cavitation and wall thickness again. If you only optimise one thing before cutting steel, optimise the wall.

Twelve ways to cut the piece price

Ranked roughly by impact per unit of effort.

1. Reduce wall thickness to the minimum that meets the structural requirement. Cooling scales with thickness squared, and weight scales linearly. Biggest single lever, no tool cost.
2. Add ribs instead of thickness. Stiffness per gram is far higher, and ribs do not lengthen cooling.
3. Increase cavitation to the volume-justified level. Do the payback calculation explicitly rather than accepting the supplier's default.
4. Convert to a hot runner. Removes runner material and several seconds of cycle. Pays back at volume, never at low volume.
5. Move the gate to shorten the flow path. Less injection pressure needed, smaller machine, shorter fill time.
6. Choose a lower-density resin at equal performance. PP at 0.905 g/cm³ delivers more parts per kilogram than ABS at 1.05.
7. Eliminate secondary operations by design. Mould the snap fit instead of adding a screw; mould the living hinge instead of adding a pin.
8. Relax the cosmetic standard where it does not matter. A hidden surface at SPI B-3 instead of A-2 removes rejects and polishing cost.
9. Consolidate parts. One moulded part replacing three assembled parts usually wins even if the moulded part is more expensive than any one of them.
10. Improve tool cooling. Beryllium copper inserts or conformal cooling in hot spots, 20 – 40 % cooling reduction on difficult parts.
11. Add robot take-out. Removes 1.5 – 3 s from the cycle and stabilises quality.
12. Commit to a blanket order with scheduled releases. Gives the supplier a reason to keep the tool set up and material in stock, typically worth 10 – 20 % on the piece price.

Quotation traps that look like savings

Five patterns that produce a low quote and a high total cost.

A piece price with the tool hidden inside it. You cannot compare it to anything, you never own the tool, and there is no price reduction when the amortisation ends. Ask for the tool cost separately even if you intend to amortise.

A cycle time that is physically impossible. If the quoted cycle time is shorter than the cooling time the wall thickness allows, the supplier is either going to run it hot and ship warped parts, or they made an error they will recover later. Check the cycle time against the wall thickness table.

A cavitation that does not match the volume. An 8-cavity tool for a 3,000-part order means you paid for capacity you will never use. Ask why that cavity count was chosen.

A material grade that is not the one on the drawing. "ABS" covers a wide range; a low-cost high-flow grade may not meet your impact requirement. Specify the grade and ask for the material certificate.

No stated yield or cosmetic standard. A price with no reject allowance is a price that will be renegotiated after the first production run. Agree the cosmetic standard and the acceptable reject rate in writing.

Frequently asked questions