YARDPUMP · Technical Article

The Plastic Impeller Hub Looks Sound. Could It Have a Shrinkage Void Inside?

Check a thick impeller hub for hidden shrinkage voids. Link moulding conditions, sections and fracture location before approving a production change.

Suppose an open plastic impeller passes a diameter check and spins freely, but a hub later cracks near the shaft bore during a controlled load test. The outer surface looked good before assembly. An internal shrinkage void is one possibility. Confirm whether the defect was present before installation and where the fracture started. A hidden cavity at a highly loaded part of the hub can matter more than a shallow visible sink mark on a surface that carries little load.

Why the outside can look better than the inside

Plastic near a mould wall cools first. In a thick section the centre remains hot longer and contracts as it cools. Packing can feed more material while the flow path from the gate remains open. Once that path freezes, additional pressure at the machine does not necessarily reach the thick centre. The part may sink at the surface, form a void inside, or show a combination. Gate position, local geometry, resin, process history and reinforcement all affect the outcome. A void is therefore a reason to trace the process, not proof that one setting was wrong.

A void may also result from other mechanisms, including gas or moisture during processing. A crack at an insert or pressed bore can form during assembly. Compare a cut or scanned unassembled part with an installed or fractured part to see which signs were present first. Note the direction of mould filling and the location of thick ribs, the bore, insert seats and the section through which torque reaches the vanes. Avoid assuming that every dark spot in a photograph is a cavity.

Pair the visual check with sections and process history

Take specimens identified by cavity, material batch and run conditions. Begin with a known representative part and a part from the suspect condition. Weigh and measure them using the same conditioning and procedure. Part mass may identify a broad shift, but a single small void can be too small to distinguish from ordinary variation, and mass cannot locate it. Use an approved sectioning plan or capable industrial imaging for the hub. Look at more than one plane where warranted; a section through the wrong angle can miss the defect. Record its location, shape and distance from the torque path or insert.

For an example, two parts both meet bore diameter and outside diameter. One shows a cavity near the thick centre, well away from its mounting wall in the chosen section. Another has a cavity extending toward the loaded edge of the bore. The same total void area does not make them equally acceptable. The designer needs the remaining load-bearing section, the loading direction and repeatable test evidence. Do not set an allowable void percentage from this example.

Fix the cause and check the assembled load path

Compare the process records: fill and pack sequence, mould and melt conditions, cycle time, cavity identity and resin handling. Consider how the gate and hub geometry affect packing. A process adjustment might reduce the defect; a thick transition may need a design change. Either way, confirm the resulting dimensions, balance and connection to the shaft. Validate the approved assembled duty, including applicable transient or obstruction loads under a guarded test method. A change that eliminates a visible sink can still leave internal problems, distort the bore or change the impeller’s running clearance.

Keep fracture evidence before polishing or cutting the returned component, and do not run a visibly damaged impeller. The purchase decision should connect the part location, process evidence and assembled performance. An attractive surface and a passing dimension are useful checks, but neither can see through a thick hub.

Frequently asked questions

Can a small sink mark prove there is a hidden void?

No. Sink and void can share a cause, but the visible mark cannot identify internal position or size.

Will weighing every impeller catch all voids?

No. Mass is a useful screening signal under controlled conditions. It cannot locate a defect or reliably distinguish small cavities.

Does changing to a different resin eliminate the problem?

Not necessarily. Flow, packing, shrinkage, conditioning and load transfer must be validated for the changed design and process.

Related reading