Premium
Distribution Function Parameters Determined from Dynamic Mechanical Spectroscopy Data
Author(s) -
Hermida Élida B.,
Povolo F.
Publication year - 1994
Publication title -
physica status solidi (b)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.51
H-Index - 109
eISSN - 1521-3951
pISSN - 0370-1972
DOI - 10.1002/pssb.2221820207
Subject(s) - viscoelasticity , distribution function , mathematics , distribution (mathematics) , function (biology) , statistical physics , perturbation (astronomy) , relaxation (psychology) , mathematical analysis , perturbation theory (quantum mechanics) , normal distribution , thermodynamics , materials science , physics , statistics , quantum mechanics , psychology , social psychology , evolutionary biology , biology
The dynamical properties of linear viscoelastic materials are related to the strength of relaxation as well as to the statistical parameters of the corresponding distribution function. Nowick and Berry established numerically the relationships between the real and imaginary moduli or compliances, the strength of relaxation, and the parameters of a widely used distribution function: the log‐normal one. These relationships are modified, applying a perturbation theory, in order to get the parameters from an internal friction peak measured as a function of temperature. Recently, a modified anelastic element (MAE) has been introduced for describing the mechanical properties, particularly the internal friction, of linear viscoelastic solids. It has been shown that the distribution function associated to the MAE is quite similar to a log‐normal distribution. Therefore, in the present paper the perturbation theory developed by Nowick and Berry is applied to the internal friction peak for the MAE. It is found that the approximate values calculated for the MAE are in excellent agreement with the analytical parameters associated to this element and also with the approximate expressions derived for the log‐normal distribution. However, this treatment is not enough to determine the temperature dependence of the parameters except if very accurate frequency data are measured. Consequently, an alternative procedure is proposed.