Don't forget that polymers are always tackled by their chemical or physical weak points. Consequently, the study of degradation depends on our capability to see and quantify:
1. the degradation that is an evolution of aspect, colour, mechanical or physical or electrical or chemical properties. Lifetimes of polymer parts are often in the order of about ten years, sometimes several decades and more rarely a century. Obviously, commercially and economically, it is not possible to make tests during such long periods.
2. The weak points where the degradation preferably starts.
To speed up the service life estimation it is usual to intensify the surrounding constraints: heat, load, UV... and to treat the results with predictive models. But there are some boundary lines. Too high a test temperature can activate chemical mechanisms differing from those reached during the service life. Artificial light irradiation can provoke discoloration differing from natural exposition. Too high a loading can change the cracking kinetics etc.