Viscoelastic-Viscoplastic Model to Predict Creep in a Random Chopped Mat Thermoplastic Composite

Abstract

Random glass-mat thermoplastic (GMT) composites are used in automotive applications due to their mechanical properties and low processing cost. However, there is an inherent issue in that the thermoplastic matrices exhibit viscoelastic behaviour. In order for manufacturers to have confidence in their products, it is important to be able to predict the long-term behaviour of these materials.

In this work, chopped glass fibre mat reinforced polypropylene was studied over a stress range at room temperature. Through short-term creep experiments, a material variability of ±18% was determined.

One day creep experiments indicated the presence of viscoplastic damage in the composite, which was verified using microscopy. The creep deformation has been modeled using a viscoelastic-viscoplastic model. To verify the model, two creep tests of 33 day duration were conducted. The test results agreed well with the model.

The temperature effects and applicability of time-temperature superposition (TTS) have been investigated. Parametric studies conducted suggest that the failure modes for chopped fibre composite become matrix dominated at temperatures higher than the secondary glass transition of 60°C. Through the development of a master curve based on 20MPa data, it was shown that TTS is applicable to the composite.

Short-term tests indicated that the material response of chopped fibre mat composites is far too random to be meaningfully quantified and this is further exacerbated at higher temperatures.

From 1 day creep tests at various temperatures, it was seen that temperature appears to increase plastic strain in the material exponentially. By comparing the results from these tests to micrographs of the material, it showed that above the secondary glass transition, bulk deformation of the matrix phase in the composite is dominant due to matrix softening. Deformation of the matrix phase accelerates fibre-matrix debonding and therefore the progressive failure process.

Overall, there is evidence that even for materials with inherently high property scatter, it is possible to identify the effects of nonlinearities arising from external factors using short-term creep tests on single specimens. It is, however, more difficult to develop an accurate long-term model that can account for stress and temperature conditions because of the wild experimental scatter.