Plastic injection moulds: which types exist and when to use each

Before an injection-moulded part can be made, you have to decide what its mould will be like. That decision sets much of what the mould will cost, how long each cycle will take and how the part will look. Here we explain what types exist, how they work inside and when each one makes sense.

We write from our factory in Tibi (Alicante), where we mould with machines from 100 to 500 t of clamping force. If you would like to see first how an injection moulding machine works, the technology section of our home page takes you through the cycle of a part step by step.

An injection mould is an assembly of machined metal plates with one or more cavities, which the injection moulding machine fills with molten plastic at high pressure. The mould also acts as a heat exchanger, and inside it the plastic solidifies in the shape defined by the cavity^{Source 1}. These are the parts we will refer to.

• Cavity and core. The cavity is the hollow that gives the part its shape, and the core forms its interior^{Source 1}.
• Parting line. This is the surface along which the two halves of the mould separate when it opens^{Source 1}.
• Sprue, runners and gate. The plastic arrives from the machine nozzle through the sprue, is distributed through the runners and enters the cavity through the gate, a passage that is generally small and the first thing to solidify^{Source 1}.
• Runner. This is the plastic that solidifies in the runner system and leaves the mould every cycle^{Source 2,Source 3}.
• Ejectors. These are pins that push the part out of the mould when it opens^{Source 1}.
• Cooling and vents. Water or another fluid circulates through channels to control the temperature of the surfaces in contact with the plastic, and air escapes from the cavity through very fine vents^{Source 1}.

This is the most widely used type of mould. It has a single parting line, so the runners and gates have to lie on it in order to come out when the mould opens. Compared with a three-plate mould, it is cheaper, simpler to make and has a shorter cycle^{Source 5}.

The part and the runner come out together. With an edge gate, which fills the part through its edge from the parting line^{Source 1}, they have to be separated afterwards in a secondary operation^{Source 6}. With a submarine gate, on the other hand, the gate is cut from the part during ejection^{Source 6}. And if, with several cavities, the runners cannot be balanced on that single parting line, you may need a three-plate mould or a two-plate mould with a hot runner^{Source 5}.

A runner stripper plate is added between the clamping plate on the fixed side and the A plate, which holds the cavity, and the mould then has at least three parting lines^{Source 7,Source 8}. The plastic enters through a pin-point gate, an opening from a few tenths of a millimetre to one and a half millimetres across, which is torn off the part when the mould opens^{Source 6,Source 1}.

Its advantage is that the gate can go at the centre of the part or at several points, the runner separates on its own, and a hot runner, which is more expensive, can often be avoided^{Source 7}. In return, the mould costs more and needs more maintenance, the cycle is longer and the machine has to open considerably further^{Source 7,Source 8}.

Removing the parts and the runner by robot is also harder^{Source 7}. This is worth bearing in mind when removal is automated, as it is in our case, where we use a Cartesian robot.

This mould works thanks to the order in which its plates separate. A technical series on three-plate mould design recommends this sequence^{Source 9,Source 8}.

1. First, the line between the stripper plate and the A plate opens a few millimetres, generally no more than 3 to 6 mm. The gate breaks while the part is still held between cavity and core, and puller pins keep the runner on the stripper plate^{Source 9}.
2. The main parting line, the one that separates the cavity from the core, opens^{Source 8}.
3. Stop bolts or limit bars, or a sequencing latch, pull the A plate along and open the runner parting line fully^{Source 8}.
4. Once that line is fully open, the stripper plate separates from the clamping plate and the runner drops^{Source 8,Source 10}.
5. The ejectors push the parts out of the mould^{Source 1}.

Cutaway model of a three-plate injection mould mounted in a machine on a bench, fully open. The stripper plate has moved a few millimetres away from the fixed clamping plate. The runner parting line has opened between the stripper plate and the cavity plate, and the runner has dropped through it and lies in its tray below the mould. The main parting line has opened between the cavity plate and the core plate, and the two parts have dropped through it into another tray. Limit bars on the plates set the order of the openings, and the scale on the bench shows the total stroke.

According to that series, if the main parting line opened first, the gate could distort the part, leave a taller vestige or even pull the part off the core^{Source 9}. Even so, opening the main parting line first is also common, for example with delay latches, and the pin-point gate is then torn off as the cavity separates^{Source 6,Source 11}. When the runner is left to drop, it needs room to fall freely^{Source 8}. If a robot removes it, a moulded peg can be added that holds it on the stripper plate until the robot picks it up^{Source 7}.

This choice affects the investment, the material left over each cycle and how easily you can change colour.

In a cold runner mould the runner channels are not heated and the runner comes out with the parts every cycle^{Source 2,Source 3}. The investment is lower, there is less thermal variation because no heaters or controllers are needed, the previous colour does not stay in the runners and the gate can go wherever suits the part. On the other hand, the runner has to be removed and reground for reuse, and material costs are higher^{Source 2}.

A hot runner is an assembly of heated components that carries the molten plastic to the cavities. Its manifold distributes the plastic entering the mould among several nozzles^{Source 3}. The part comes out already separated and cycles are usually faster^{Source 12}.

Its costs are less visible. The plastic waiting in the manifold adds residence time, the time the material spends molten^{Source 13}. Colour changes take time, material and purging compound, maintenance has to be allowed for, leaks or stoppages can generate waste, and the gate locations are fixed by the manifold^{Source 2}.

A thermal gate leaves a protruding vestige of one third to one half of the gate diameter, and is reserved for parts where gate quality is not critical^{Source 4,Source 12}. A valve gate closes the gate with a moving pin and leaves a small mark the size of the gate, with no drooling or stringing^{Source 12,Source 4}. It is more expensive, but it lets you open the gates in sequence and decide where the weld lines fall, the marks that form where two plastic flow fronts meet^{Source 4}.

This is a hot runner without heating, with runners so thick that the plastic stays molten as long as the mould runs continuously. It only works with commodity plastics such as polypropylene (PP) and polyethylene (PE), and if the mould stops, the runner freezes and the mould has to be opened to remove it^{Source 14}.

With a single cavity, there are no differences between cavities to keep an eye on. With several, each shot produces more parts, but the runners have to distribute the plastic evenly^{Source 5}.

Symmetrical runners are not enough. The plastic heats up through friction as it flows, and not uniformly across the runner cross-section. From four cavities upwards this causes filling differences, and in moulds with eight or more cavities the worst cases can even be seen with the naked eye. The parts come out with different density, shrinkage and warpage, and a well-designed runner layout improves uniformity^{Source 15}.

A family mould produces different parts in the same mould, such as a box and its lid, and helps reduce the total investment in moulds^{Source 16}. The parts usually need to be similar in size, weight, wall thickness and volume, because they are all moulded with the same parameters^{Source 16}. They also have to be made of the same material and produced in the same quantity^{Source 17}. If they are very different, one solution is to control each cavity separately with valve gates^{Source 16}.

A stack mould has two parting lines instead of one^{Source 18}. It can almost double the output of the machine with a similar clamping force, although up to 15% more is sometimes needed, and only the opening and closing time is added to the cycle^{Source 18,Source 19}. It suits flat parts better than parts with deep cores^{Source 18}. In return, the mould has a greater shut height and needs a longer opening stroke, so you have to check that it fits the machine^{Source 18}.

An undercut is a feature of the part that is trapped in the mould and prevents it from being ejected in a straight line^{Source 1}, such as a side hole or a thread. It is dealt with by changing the geometry or by adding mechanisms to the mould^{Source 20}.

• Slides. These are blocks that move in from the side when the mould closes and retract when it opens^{Source 20}. They can be driven by an angle pin, which moves them sideways, away from the part, as the mould opens^{Source 21}, and a retainer holds them back while the mould is open^{Source 22}.
• Angled lifters. They release internal undercuts by converting the forward movement of the ejector plate into a sideways movement^{Source 20,Source 23}.
• Hand-loaded inserts. These are metal pieces that an operator places in the mould every cycle^{Source 20}. They simplify the mould at the cost of manual work.
• Unscrewing. A threaded part may have to be unscrewed from the core before it can come out, and the gears that rotate the core can take up a lot of space in the mould^{Source 24}.
• Collapsible cores. Spring-loaded segments form the inside of a cap and, when the centre pin is withdrawn, collapse towards the centre to release it^{Source 25}.

Yes. Insert moulding encapsulates a metal component in plastic, such as a threaded insert that gives a stronger assembly. The bond is mechanical, so the insert is designed for the plastic to grip it^{Source 26}.

Overmoulding adds a layer of plastic over a part that has already been moulded, for example a soft, elastic layer over a rigid body to give a non-slip grip. The bond can be chemical or mechanical, and it is best not to leave edges of the overmoulded material exposed, so that it does not peel off^{Source 26}.

Two-component moulding injects two materials in the same cycle, with a single mould and a machine with two injection units^{Source 27}. To make room for the second material, a rotary platen mounted on the machine moves the first shot to the second-shot cavity, or an index plate built into the mould transfers it to a different core, a task that a robot performs in some cases^{Source 27}. Alternatively, the mould itself, usually without opening, can move part of the core or cavity^{Source 27}. Moulding a part on one machine and overmoulding it in another mould on a different machine is a related process that falls outside this definition^{Source 27}.

Moulds can be made of aluminium, soft or pre-hardened steels, or hardened steels, and their life also depends on how they are built and on the cleaning, lubrication and maintenance they receive^{Source 28}. The SPI classes, from the former Society of the Plastics Industry, now the Plastics Industry Association, provide a common language for discussing that life^{Source 29}.

The sources agree on the order of magnitude, but not always on the exact limits of each class^{Source 30,Source 29,Source 28}. One of them also uses a 400 series for moulds used on machines of 400 tons or more^{Source 30}. As it is a US source, these are most likely short tons, about 360 t.

The table summarises the above. These are general trends, and every part needs its own assessment.

We also manufacture by blow moulding, so this is worth clarifying. In blow moulding, a tube of molten plastic or a heated preform is inflated with compressed air inside the mould, producing hollow, lightweight parts such as bottles^{Source 31}. Because the pressures are lower than in injection moulding, the mould can be made from cheaper, less durable materials, usually aluminium, stainless steel or beryllium-copper alloys^{Source 31}.

The plates and runners described in this article are specific to injection moulding. In the frequently asked questions on our home page, we summarise when each process is used.

It depends on how many parts you need per year and for how many years. That figure guides the mould class, the number of cavities and whether a hot runner pays off^{Source 28,Source 13}. It is worth being clear about it before you request a quote.

Not necessarily. It avoids the runner, but it adds investment, maintenance and time at every colour change, and at low annual volumes it may not pay off^{Source 2,Source 13}.

It affects colour changes. With a cold runner, the runners are emptied every cycle, whereas with a hot runner a colour change takes time, material and purging compound^{Source 2}. If the colour of your part is critical, you may be interested in our article on colour measurement in plastics.

At Bersil we manufacture by plastic injection and blow moulding in Tibi (Alicante), with injection moulding machines from 100 to 500 t of clamping force and semi-automated processes. From your very first call, you deal with a member of our team.

Tell us what part you need and in what quantity, and we will study it with you. You can talk to our team or first see what backs up our quality.