Investigation of relaxation properties of polymer melts by comparison of relaxation time spectra calculated with different algorithms

In this paper a recently introduced algorithm (Brabec and Schausberger, 1995) for the calculation of relaxation time spectra is compared with two standard methods, i.e., Weese's regularization, and Baumgaertel's and Winter's regression algorithm. A reasonable agreement between those three algorithms is found for the relaxation properties of mono-, polydisperse, bi-, and multimodal polystyrene samples. All three numerical methods reproduce the relaxation properties for long and medium times correctly, but they show some disagreement at short times because of sparse experimental data. The high numerical accuracy opens the possibility to test and improve the physical models which underlie the calculations. The good agreement of the different algorithms suggests that small inconsistencies to physical models are not due to a failure of the numerical methods, but due to an insufficiency of the generalized Maxwell model.