What is injection molding?

The short answer

A machine is three systems

An injection molding machine looks complicated and is conceptually simple. It has three parts, and knowing which one is which is most of what is needed to hold a useful conversation about a part.

• The injection unit. A hopper feeds pellets into a heated barrel containing a reciprocating screw, usually with a length-to-diameter ratio of 18:1 to 24:1. The screw rotates to melt and homogenise the polymer, then slides forward like a plunger to push a measured volume of melt through the nozzle. Barrel temperature is held in three to five zones from the feed throat to the nozzle.
• The clamping unit. This closes the mould, holds it shut against the pressure of the melt and opens it for ejection. Its capacity is quoted in tonnes of clamp force, and machines range from about 5 tonnes for small precision parts to several thousand tonnes for large panels. Insufficient clamp force lets the halves part slightly and resin escapes as flash.
• The mould. A precision cavity in hardened tool steel or aluminium containing the part geometry, runners, gates, cooling channels and ejector pins. The mould, not the machine, determines the part's shape, where it fills, how it cools and how it releases.

One cycle, six stages

Technical references describe the cycle as four stages and as six, and both are correct: the four-stage version groups packing into cooling, while the six-stage version separates them because they are controlled independently.

1. Clamping. The moving platen brings the two halves together and the clamp builds to full tonnage. The mould must be completely sealed before injection or the melt will leak at the parting line.
2. Injection or filling. The screw moves forward and drives melt through the nozzle, sprue, runner and gate into the cavity. Filling is fast - usually under two seconds - and it determines whether every detail of the cavity is reproduced.
3. Packing and holding. Pressure is maintained for 5 to 20 seconds so that additional melt enters the cavity to compensate for shrinkage as the part cools. This stage is what prevents sink marks, internal voids and dimensional drift.
4. Cooling. Heat transfers from the part into the mould's cooling channels while the screw rotates and retracts to prepare the next shot. Cooling is 50 to 80 percent of the cycle and it is where part dimensions are effectively set.
5. Mould opening. The clamp retracts and separates the halves, exposing the solidified part on the core.
6. Ejection. Ejector pins, sleeves or a stripper plate push the part clear. Draft angles on the walls are what make this possible without marking or distorting the part.

Cooling owns the cycle

The practical consequence of that figure is the single most useful rule in plastic part design: wall thickness drives cost. Solidification time scales roughly with the square of section thickness, so doubling a wall from 1.5 mm to 3 mm multiplies the cooling time by about four. Every additional millimetre of nominal wall, every unnecessary boss and every local thick section adds seconds to every shot for the life of the programme. That is why the standard advice is to keep the wall as thin and as uniform as the function allows and to add ribs where stiffness is needed rather than mass.

The process window

Four conditions determine whether a part comes out right, and they are set independently of the part geometry.

• Melt temperature. Resin dependent, broadly 220 to 300 C for engineering grades and lower for polyolefins. Too cold and the cavity short-shots; too hot and the polymer degrades and loses strength.
• Injection pressure. Typically 60 to 150 MPa, or 600 to 1,500 bar. It has to fill the thinnest wall before the melt freezes without over-packing the area around the gate.
• Mould temperature. 20 to 120 C depending on the resin. Higher mould temperature improves gloss and weld-line strength and reduces residual stress, at the cost of a longer cycle.
• Cooling time. The variable that is genuinely worth optimising, because it is the largest block of cycle time and therefore the largest single driver of unit cost.

Design rules that come from the process

1. Keep the wall uniform and in the 1 to 5 mm range. Thick sections cool more slowly than thin ones, which produces sink marks over ribs and bosses and warping across the part.
2. Add draft to every wall parallel to the opening direction. About 1 to 3 degrees on smooth surfaces and more on textured ones, because the part shrinks onto the core as it cools.
3. Size ribs at 50 to 60 percent of the nominal wall. A rib as thick as the wall behaves like a thick section and sinks on the opposite face.
4. Keep radii generous and avoid sharp internal corners. Sharp corners concentrate stress and restrict flow; rounded transitions fill better and are stronger in service.
5. Decide the gate and parting line at design stage. Both are mould decisions, and both affect where weld lines, cosmetic blemishes and flash appear. Settling them after the tool is cut is expensive.

What injection molding cannot do

• It cannot mould undercuts for free. Features that block ejection need side actions, lifters or collapsible cores, all of which add tooling cost and cycle time. A snap fit that needs a side action should be checked against an alternative geometry first.
• It cannot be justified at very low volume. Tooling runs from a couple of thousand dollars for a simple aluminium cavity to tens of thousands for a hardened multi-cavity tool, and below a few hundred parts that investment does not return.
• It cannot produce a part that does not fit the machine. Shot size sets the maximum part mass and clamp tonnage sets the maximum projected area; both are hard limits, not preferences.
• It cannot hold tighter than the process allows. Around plus or minus 0.025 to 0.05 mm on features cut into the steel is realistic; anything tighter has to be machined afterwards, which is a different operation and a different cost.
• It cannot use every polymer. Thermoplastics dominate; thermosets need specialised machines because they cross-link and cannot be remelted.

Getting a part quoted

The most useful thing a buyer can send is the model together with the function: what the part does, what it carries, what environment it sees, which surfaces are visible and what it has to mate with. From that we can propose a resin and a wall strategy, flag where draft or a side action will be needed, comment on where weld lines and sink marks are likely to land, and give a tooling price with a unit price at two volumes. See plastic injection molding for the process, aluminium die casting where a metal part is the better answer, and surface finishing for the texture, paint and plating options that follow moulding.