Hi,

The refractive index change with temperature is only 1 part of the problem if you are talking of an optical system, not just a plate of plastic. Temperature will also mechanically distort the lenses and the optomechanics that are holding the lenses. This requires external simulation (STOP). Zemax has recently introduced a module called STAR that eases the integration with external software (e.g. ANSYS, COMSOL).

In addition, there can be a limitation of Zemax in the case where the temperature is not uniform within the system (e.g. you are close to the source or the heat comes for the light itself). IIRC, Zemax only considers a uniform temperature. It should be easy for them to add at least one temperature by surface/object, but gradient of tenmperature would need a little more work (using GRIN equations basically).

Plastics comes with variation (lot to lot, also from processing, e.g. stress birefringence) so it may be less necessary to be extemely accurate in the temperature behaviour. In general, if you need high accuracy with plastic for your system to work, you may already have a problem.