The Results of Stress-Temperature Measurements

The Results of Stress-Temperature Measurements.—Hysteresis in the stress-strain behavior of rubber and rubberlike materials has presented the most serious problem encountered in the execution of otherwise simple experiments on the change of stress in stretched rubber with temperature at constant length (L) or at constant elonga- 

The intercepts obtained by linearly extrapolating data such as are shown in Fig. 85 to 0°K are given by the lowest curve in Fig. 86. These represent, of course, E/dL)T,p. The force / is given by the uppermost curve in this figure, and the quantity —T dS/dL)T,p = T(df/dT)p,L obtained by difference according to Eq. (13 ) is given by the dashed curve. 

No thermoelastic inversion should appear in the force-temperature coefficient at constant elongation a, inasmuch as the effect of ordinary thermal expansion is eliminated by fixing a instead of the length L as the temperature is varied. As the elongation approaches unity, both the force and its temperature coeffi.cient df/dT)p,a must van- 

Anthony, Caston, and Guth applied similar corrections to their measurements in order to express the force as a function of the temperature at various constant elongations. The results are shown in Figs. 87 and 88. The force-temperature plots are linear within experimental error. Their intercepts at 0°K are near zero, except at the higher elongations. These intercepts, which according to Eq. (22) must represent (dE/dL)T,Vi are plotted in the lowest curve of Fig. 88. It is doubtful 

These conclusions have been confirmed by Wood and Roth, who carried out measurements at both constant lengths and at constant elongations using natural rubber vulcanized with sulfur and an accelerator. Their results at constant elongation, to be considered later in connection with the thermodynamics of rubber elasticity at higher elongations, are summarized in Fig. 89. 

---