From dynamic modulus via different relaxation spectra to relaxation and creep functions

The main goal of the paper is to compare predictive power of relaxation spe ctra found by different methods of calculations. The experimental data were obtained for a new family of propylene random copolymers with 1-pentene as a comonomer. The results of measurements include flow curves, viscoelastic properties, creep curves and rubbery elasticity of copolymer melts. Differ ent relaxation spectra were calculated using independent methods based on d ifferent ideas. It lead to various distributions of relaxation times and th eir "weights". However, all of them correctly describe the frequency depend encies of dynamic modulus. Besides, calculated spectra were used for findin g integral characteristics of viscoelastic behaviour of a material (Newtoni an viscosity, the normal stress coefficient, steady-state compliance). In t his sense all approaches are equivalent, though it appears impossible to es timate instantaneous modulus. The most crucial arguments in p estimating th e results of different approaches is calculating the other viscoelastic fun ction and predicting behaviour of a material in various deformation modes. It is the relaxation and creep functions. The results of relaxation curve c alculations show that all methods used give rather similar results in the c entral part of the curves, but the relaxation curves begin to diverge when approaching the high-time (low-frequency) boundary of the relaxation curves . The distributions of retardation times calculated through different appro aches also appear very different. Meanwhile, predictions of the creep curve s based on these different retardation spectra are rather close to each oth er and coincide with the experimental points in the wide time range. Relati vely slight divergences are observed close to the upper boundary of the exp erimental window. All these results support the conclusion about a rather f ree choice of the relaxation time spectrum in fitting experimental data and predicting viscoelastic behaviour of a material in different deformation m odes.