Extensional stress-strain measurements on a polyurea (PU) were carried out at strain rates up to 830 s -1 , in combination with ex post facto small-angle X-ray scattering (SAXS) measurements and temperaturedependent SAXS. The elastomer is of interest because of its application as an impact-resistant coating. The highest strain rates used herein fall within the softening, or transition, zone of the viscoelastic spectrum and are thus relevant to the working hypothesis that the performance of a polyurea impact coating is related to its transition to the glassy state when strained very rapidly. While quasi-static and slow deformation of the PU gives rise to irrecoverable strain and anisotropic SAXS patterns, when stretched at high rates the PU recovers completely and the scattering is isotropic. Thus, the deformation of the hard domains observed at low rates is absent at high strain rates. Linear dynamic mechanical measurements were also carried out, with the obtained segmental relaxation times in good agreement with dielectric relaxation measurements on this material. The PU exhibits the usual breakdown of time-temperature superposition in the transition zone. This thermorheological complexity underlies the fact that published time-temperature shift factors for this material are unrelated to the segmental dynamics, and therefore use of these shift factors to predict the onset of glassy dynamics during impact loading of the PU will be in error.

Master curves of the small strain and dynamic shear modulus are compared with the transient mechanical response of rubbers stretched at ambient temperature over a seven-decade range of strain rates (10 À4 to 10 3 s À1 ). The experiments were carried out on 1,4-and 1,2-polybutadienes and a styrene-butadiene copolymer. These rubbers have respective glass transition temperatures, T g , equal to À93.0, 0.5, and 4.1 C, so that the room temperature measurements probed the rubbery plateau, the glass transition zone, and the onset of the glassy state. For the 1,4-polybutadiene, in accord with previous results, strain and strain rate effects were decoupled (additive). For the other two materials, encroachment of the segmental dynamics precluded separation of the effects of strain and rate. These results show that for rubbery polymers near T g the use of linear dynamic data to predict stresses, strain energies, and other mechanical properties at higher strain rates entails large error. For example, the strain rate associated with an upturn in the modulus due to onset of the glass transition was three orders of magnitude higher for large tensile strains than for linear oscillatory shear strains.