The motor shaft turns, but a plastic impeller does not follow it reliably. No obvious crack appears after dismantling. The weakness may be at the hub-to-shaft torque interface rather than in the vanes. First identify whether torque travels through flats, a key, an insert, threads or interference friction. Then assess whether that path remains effective under temperature, liquid exposure and starting loads.
A fitting hole is not a complete torque connection
A correct bore diameter establishes only part of the geometry. A round friction fit depends on contact pressure and engaged length. A flat or keyed interface depends on effective contact, local bearing stress and geometry. A metal insert must also transmit torque from its outside surface into the plastic. The load path does not end when torque reaches the metal.
Draw the complete route from shaft to vane. Identify possible slip, local crushing and components that provide only axial retention. A fastener preventing the impeller from falling off is not necessarily the primary torque-transmitting element. A threaded connection likewise requires its intended rotation and locking arrangement to be understood.
Why an initially tight fit can change
Plastic dimensions and contact pressure can change with temperature, moisture, sustained loading and the liquid. A tight assembly at room temperature does not establish permanent pressure at a friction interface. Where interference is important, assess stress relaxation and dimensional behaviour in the relevant material condition, rather than recording only the initial assembly feel.
Starts and occasional obstructions can impose loads different from steady running. Repeated impacts against a hub flat may produce local wear, debris or rounding. Locate the debris before attributing every plastic particle to vane rubbing. Reducing impeller diameter will not necessarily repair an impaired hub connection.
| Interface | Main checks |
|---|---|
| Round friction fit | Dimensions, engagement, assembly and long-term retention |
| Flat or key | Actual contact, local stress and angular clearance |
| Metal insert | Torque transfer on both sides and stress in the plastic |
| Threaded retention | Rotation, axial location and validated locking |
Understand torque without inventing it from watts
Mechanical power equals torque multiplied by angular speed. In a purely illustrative calculation, 200 watts of shaft output at 2850 rpm corresponds to approximately 0.67 newton-metres of steady torque. This is neither a product specification nor a hub acceptance value. Starting, stalling and debris impacts may involve different torque histories, so the steady figure cannot define the complete design margin.
Electrical input cannot replace shaft output when efficiency is unknown. A test system calculating output from a default efficiency must disclose that assumption. Interface validation needs evidence related to the design load, rather than an apparently precise electrical number converted into an unmeasured mechanical capacity.
Preserve evidence before replacing parts
Isolate power and retain the shaft, hub, fasteners and their pairing. Photograph polished contact patches, debris, deformation and relative-slip marks. Record liquid, temperature, starting conditions and any obstruction event. If measuring a removed bore, identify whether the sample is dry, recently soaked or cooled. Dimensions without a material-state boundary can mislead the comparison.
Competent personnel can investigate suspected relative movement in a controlled fixture. Users should not mark or restrain accessible rotating parts during operation. A worn pair does not become demonstrably reliable merely because one member is replaced. Temporary adhesive is also not a substitute for an engineering review of the interface.
Validate the complete revised assembly
The design owner should select an interface that can be manufactured and inspected, with defined dimensions, material condition and assembly requirements. Where relevant, compare retention before and after liquid and temperature exposure, and include the intended starting loads. Changes can also affect concentricity, axial location, balance, sealing and hydraulic operation, so recheck those consequences.
The purchasing question is not whether an added metal part looks stronger. It is which torque path improves, how a sample demonstrates that improvement and how production will preserve it. More complicated construction does not automatically produce a better load distribution.
Questions about hub slip
Does a metal insert always solve the problem?
No. The insert-to-plastic interface still needs adequate torque transfer and long-term stability. Otherwise the weak interface has only moved.
Can tightening the axial nut improve every torque connection?
No. The effect depends on the intended load path. Excessive tightening can damage plastic or change axial clearance.
Can electrical watts divided by speed specify the hub?
No. Establish actual shaft output, correct units, transient loads and interface behaviour. Electrical input is not a direct torque measurement.
