This is not as straight forward as it sounds. For instance, yield in some polymers can be better modeled with the Extended Drucker Prager model. This takes into account the dependency of yield strength on the hydrostatic stress state. The idea is that an increase in the dilatational stress component increases free volume, thereby increasing chain mobility and reducing the yield strength. All very nice, but you need special material properties to use this model and these are not commonly available. Unless you want to invest money in testing, you are stuck with common data sheet allowables and common failure criteria.
Yes as Chris pointed out, this can get hairy. Strain rate, ambient temp, environmental factors are all a concern. Creep of course is an issue with continuous loads, and fatigue analysis in plastics is like herding cats. Plus, the materials you listed are hugely dissimilar, so what works for one may not work for the other. If you want to do some reading, I suggest the following:
Failure of Plastics by Browstow and Corneliussen
Structural Analysis of Thermoplastic Components by Trantina and Nimmer
Fracture Behaviour OF Polymers by Kinloch and Young
Deformation and Fracture Mechanics of Engineering Materials, Chapter 6, by Hertzberg
Rick Fischer
Principal Engineer
Argonne National Laboratory