On the parameters identification of three-dimensional aging-temperature-dependent viscoelastic solids through a Bayesian approach.

Temperature and physical aging effects on the three-dimensional behaviors, such as shear and bulk properties, of viscoelastic solids are critical for developing polymer composite systems but are rarely addressed. Furthermore, accurate identification of time-temperature-dependent viscoelastic constitutive theories remains a challenging problem. In this work, we propose a robust and automated method to simultaneously identify the viscoelastic, temperature, and physical aging related three-dimensional model parameters based on the Bayesian framework. The proposed method was applied to three-dimensional experimental data from nonisothermal physical aging tests on polycarbonate at different temperatures. The posterior distribution of the model parameters was obtained through Markov chain Monte Carlo (MCMC) simulations, which allowed us to determine the minimum number of material parameters yielding an optimized accuracy. The experimental results are in excellent agreement with the predictions of the identified parameters. Moreover, we have found through the experimental results that the shear properties of polycarbonate are affected by temperature and physical aging, whereas its bulk properties are affected only by temperature and are relatively insignificant in comparison to the shear properties. The proposed method can be applied to polymers or composites to predict their long-term three-dimensional mechanical properties under complex working conditions. [ABSTRACT FROM AUTHOR]