1. Inicio
  2. Atrás
Título: The effect of time-scale truncation and data errors on the linear relaxation spectrum
Autores: Gilbert, Valérie.
Fecha: 1995
Publicador: McGill University - MCGILL
Fuente:
Tipo: Electronic Thesis or Dissertation
Tema: Chemistry, Polymer.
Engineering, Chemical.
Descripción: The knowledge of the viscoelastic properties of molten polymers and the ability to predict their behavior under deformation is of prime interest to polymer scientists and industry. The discrete linear relaxation spectrum $ lbrack G sb{i}, lambda sb{i} rbrack$ is one of the important viscoelastic properties and it is widely used to characterize molten polymers. This spectrum is obtained from experimental values of the dynamic moduli $ lbrack G sp prime( omega), G sp{ prime prime}( omega) rbrack$. A linear regression method with and without regularization and a nonlinear regression method were compared for inferring a discrete spectrum from experimental data. Since experimental values are used, errors in the data will affect the reliability of the spectrum. Because of the limitations of the experimental apparatus, the moduli can only be determined over a certain frequency range. The values of the relaxation times ($ lambda sb{i}$) and the corresponding elastic moduli ($G sb{i}$) depend on the range of frequencies over which the storage and loss moduli ($G sp prime( omega$) and $G sp{ prime prime}( omega$)) are measured and the method used to calculate them. Also, experimental sources of errors such as uncertainty in the gap measurement, variation in the shape of the edge, residual stress due to the preparation of the sample, and temperature uncertainty will induce experimental errors in $G sp prime( omega)$ and $G sp{ prime prime}( omega)$. This work investigates the effect of time-scale truncation and experimental errors on the discrete relaxation spectrum.
Idioma: en