Long-term stiffness prediction of particle filled polymers by dynamic mechanical analysis: Frequency sweep versus creep method

Abstract The long-term service life of polymers can be estimated with much shorter experiments by applying the time temperature superposition principle (TTS). In this approach, data is obtained at different temperatures, usually through a stepped isothermal method (SIM) on the same sample. Dynamic mechanical analysis (DMA) instruments offer two different measurement methods to obtain SIM data: (i) static creep tests and (ii) dynamic frequency sweeps. This paper compares both methods on highly graphite filled polypropylene. Our studies on reproducibility of each method show that the uncertainty for 20 year prediction can be lower than 6% for both methods. While creep-based tests require a shorter experimental time, frequency sweep based tests show a lower scatter on the final result. The two main factors introducing uncertainty on the end results are related to (i) the reproducibility of the experimental raw data and (ii) the TTS optimisation using shift factors. The optimisation of the shift factors by a numerical method improves the accuracy of the master curve. By comparing creep and frequency sweep SIM, it shows that for predictions of one decade, the methods deliver very comparable results (less than 10% difference). For longer predictions, the methods differ and are not interchangeable. Furthermore, DMA was also effectively used as a three-point bending setup, providing information about strain rate sensitivity and the linear visco-elastic region using the same test setup and same sample dimensions as for TTS.