Some polymer fibres have steel's strength and stiffness thanks to stretching and keep a plastic's density. Stretched polymers could hence excel as fast-spinning parts, especially as impellers of gas compressors and vacuum pumps, as turbines too where the temperature fits.
These parts are uneasily made of fibres in a matrix. Intricate shapes hamper automatic production, thin sections held at thicker ones aren't quite natural to fibres. It's done at individual fan blades of turbofans, not at cheap small integral compressor impellers.
I suggested to deform polymer raw material. Here strength is needed in the radial direction, by squeezing a disk, and in the azimuthal too, by a torsion. Then the part could be machined by usual methods, accurate and automated - if everything works as hoped.
Or just inject the part. For instance LCP is known to harden much from small shear at injection. That would be perfect to harden the blades or thin disk of an impeller: Inject the polymer at the centre so shear is strong at the thin blades and disk periphery. Heat the resin and mould a bit less, compensate with more injection pressure.
The blades and disk periphery of impellers and turbine rotors hold at a massive ring or disk section near the axis, where the material needs azimuthal strength too. This would be achieved by a rotating part of the mould, or maybe a separate construction used as the unmoulded part is still warm. Azimuthal shear could occur in the polymer between concentric tools, an outer one (optionally a mould part) that holds the impeller at its blades and a rotating inner one where the shaft will be. Other arrangement are possible, this one limits the deformations of the impeller.
PA too hardens by deformation. Simple injection and rotation could make very strong and cheap impellers.
==========
When extruding tubes, the kernel could rotate to give azimuthal shear. This can combine with the axial shear given by the extrusion.
Marc Schaefer, aka Enthalpy