Abstract

Extensive stress-relaxation data were obtained on twelve samples of polyvinyl acetate (including 8 fractionated, 2 roughly fractionated, and 2 unfractionated) over a wide range of temperature from 10° to 130°C. By making use of the time-temperature superposition principle they were combined to form a composite master relaxation curve for each of the samples, where all data were reduced to 75°C. chosen as the reference temperature. The master relaxation curves so obtained all cover the range from glassy consistency to rubbery flow as a function of time. From these, the relaxation distribution functions were calculated using the second approximation method proposed by Ferry, and compared with that derived by Williams and Ferry from dynamic mechanical measurements on an unfractionated sample of polyvinyl acetate. It was found that in the transition region where the comparison could be made, both results agreed quite satisfactorily, except in its upper terminal zone encompassing the glassy state. The real and imaginary parts of the complex dynamic modulus and the mechanical loss tangent were computed as functions of frequency for one sample from its relaxation spectrum using the well-known equations in the theory of linear viscoelasticity. These predicted properties were found to agree fairly well with Williams-Ferry's experimental data in the transition region, again except in its portion near the glassy consistency. From these results, the validity of the time-temperature super-position principle in the high-modulus portion of the transition region was suspected, in line with a similar conclusion drawn by Tobolsky and Catsiff from their stress-relaxation study on polyisobutylene. Steady-flow viscosities were calculated for all the samples studied from their relaxation spectra and found to increase with 3.4 powers of viscosity-average molecular weight, in agreement with the recent results on many other polymers. The shift factor which represents the temperature dependence of viscoelastic behavior was obtained over the temperature range studied and found to be practically independent of molecular weight. The shape of the shift factor curve thus obtained is quite similar to those for many other linear high polymers and agrees closely with that derived by Williams and Ferry for an unfractionated polyvinyl acetate sample from dynamic measurements in the range 50°–90°C. The apparent activation energy for mechanical relaxation calculated from the shift factor data increases sharply with decreasing temperature and passes through a maximum at 29°C. This behavior is also entirely the same as those observed for other polymers. The temperature at which the apparent activation energy for mechanical relaxation appeared may be compared favorably with the second-order transition temperature determined dilatometrically on undiluted polyvinyl acetate. Under the assumptions that the relaxation spectrum for a homogeneous (with respect to molecular weight) polymer in the rubbery region is represented by the so-called box-type function and that interactions between polymer molecules of different chain lengths are negligible, an equation was derived which represents the relaxation spectrum of a given heterogeneous polymer as the sum of relaxation spectra for its component polymers multiplied by its weight distribution of molecular weights. This equation was applied to the relaxation data for one unfractionated sample whose weight distribution of molecular weight had been determined using the successive fractional extraction method. It was found that the molecular weight distribution so derived from mechanical relaxation data is surprisingly close to the one from fractionation data except in the region of low molecular weights.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call