Dear Kadhim K. Resan, are you take into account the peculiarities of fatigue tests of thermoplastic materials such as polymers? In general, the tests of such materials are analogous to metals, but thermoplastic polymer materials demonstrates more expressive sensivity to temperature. So, we must prevent the temperature increasing due to energy dissipation in specimens under their cyclic loading, expecially, of high frequency. And we must take into account the creep of polymer materials when experimental set up realizes the changing of specimen deflection in non-symmetrical type of cyclic loading and accumulation of viscoelastic displacements takes place.
The heat generated as a result of subjected to force variable.
The problem is: that the devices used to measure fatigue in metals depend on the deformation, which we control it, the question is how long can we say that the polymeric material has failed if it does not result in cracking?
If I understand you right, the question is concluded in the choice of failure criteria for the thermoplastic materials under cyclic loading. Really, the fatique failure of such materials do not effected in visual (surface) cracks generation as for most metals. But, as widely known, the numerous volume microdefects are generated in material due to failure process under repeating loading. In this case I may recommend you to involve the method of registration of hysteresys losses (or heat generation) in specimen under testing. The hysteresys losses increases markedly in course of critical accumulation of defects mentioned above and became precursor of failure.
If you are testing transparent polymer materials, the dimness of specimen due to accumulation of defects under fatique may be registered by photometer. Good luck!