I would guess that the cause of radial surface waviness between round plastic plates with one rotating (presumably without interface lubrication) is thermal deformation perhaps related to different surface roughness conditions and/or differences in subsurface porosity. Hot spots are known to occur on the surface of rotating disks which sometimes move around to different locations. Is your surface distortion just an arc across the radius or is it more like a sinewave? We've had experience with rotating annular rings of resin impregnated , but still porous, carbon graphite material showing radial waviness after some period of operation with water-lubricated fluid film bearing average pressures  near 50 psi (peak pressure up to 150 to 175 psi). High contact pressure effects may be involved with your plastic disks.

The surface distortion is more like a sine wave. I learned that the part was molded cold, which for this semi-crystalline material would result in less crystallinity and lower wear resistance. I think this will be retested after being molded in a hot mold. There is another similar application that wears evenly, but I think the force or contact pressure is less. So it sounds like there may be a critical level of contact pressure which could cause this phenonmenon. I also posted this in the Plastics Engineering section.

PV values for various polymers are bar-charted in Fig. 6.2 of Myshkin,N.K., Kim,C.K. & Petrokovets,M.I. (1997), "Introduction to Tribology", Cheon Moon Gak (Korea)ISBN 89-7088-465-3. Five materials with PV above 1.0 MPascal-meters/sec. are carbon -filled nylon and PEEK-PTFE, Phenol fabric laminate, #8 (unidentified) and polyimide. Others of the 12 types listed have PVs of less than 0.5 some as low as 0.1.

Both DSM and LNP publish interesting data on the frictional properties of plastics.
Nylon 4.6 (Stanyl by DSM) is an interesting material if you are looking for excellent PV.
You might not see it listed outside DSM's data as it is a relativly new polymer.
It is very highly crystalline, and as such is not all that mould temperature sensitive (re crystalisation) as it reaches it's maximum crystalinity almost instantly.
It has good physicals up to it's melting point of about 300 deg C. A major application because of this is Automotive Timeing Chain Tensioners and Guides, where it typically tests at over 3 times the life of MoS2 filled nylon 6.6, and up to 10 times in one case

Regards
pat

I think the fact that the center of the disk is not fixed uniformly may be part of the problem. This leads to a nonuniform stiffness, thus nonuniform deflection and force distribution around the circumference, which could cause uneven wear. The part is gated on the side, which results in nonuniform fiber flow across the part. This additional variation in stiffness is also a possible contributor. Thanks for the info on this problem.

Consider calculating the temperature rise based on energy absorption in the braking mode. (Q = m.c.delta T) If you find this too high, consider a material change: metal to metal; metal to fiber; metal to other clutch type material. Most plastics have low specific heat and conductivity, and they cannot take much temp rise. A well-placed thermocouple will tell you the temp at braking and repeated applications.