Design of the novel tensile creep experimental setup, characterisation and description of the long-term creep performance of polycarbonate

Experimentally investigating the complex long-term behaviour of thermoplastic materials is crucial for reliable lifetime prediction models. Due to the material's pronounced stress- and temperature-dependence, the measurement results are extremely sensitive to inaccuracies in the loading conditions. For this reason a testing machine was designed that allows long-term creep measurements under exact temperature conditions. In the following, we will first describe the sample preparation and afterwards explain the construction of the testing machine designed in-house in detail, especially the optimisation of the test conditions. A total of five high-precision measuring cells were constructed in order to obtain reliable long-term creep data for polycarbonate (PC) at different stress levels under constant temperatures of up to 170 degrees C. The measurement results were then compared with those of the commercially used testing machine Q800 from TA Instruments, thus validating the measurement precision. The tests used for the comparison were conducted in the linear viscoelastic range, where the requirements for temperature control and measuring accuracy are particularly high. The deviations between the measurement curves turned out to be negligibly small. In the next step, the stress-dependent creep behaviour of polycarbonate was investigated under different load conditions and constant temperatures of 40 degrees C, 60 degrees C, 80 degrees C and 100 degrees C. The measurement results were evaluated with respect to time and strain and show a significant influence of crazing. The stress- and temperature-dependence of the three fundamental creep processes could be observed in polycarbonate. A yield model based on Eyring's flow theory with a single activation volume is proposed and validated. The accelerated creep progression in the tertiary creep stage exhibits no ductile-brittle transition and can therefore be described with an empirical isotropic damage variable concept that fits the experimental data exactly, including ductile fracture. The proposed concept, which is applied to spring element of the material model, can accurately reproduce the tertiary creep behaviour of PC.