| Specimens of a medium density (approximately 0.94 g/cm;Stress rupture data, accumulated at elevated temperatures, exhibited the expected transition from a ductile to a non-ductile failure mode. A regression analysis was performed to fit the experimental results to an appropriate stress-temperature-failure time equation. Because some of the test specimens never failed, the regression analysis was carried out using a Maximum Likelihood Estimation technique which will accommodate censored (i.e. not failed) data.;Activation energies for both failure modes were calculated. That for non-ductile failure agrees with the results of uniaxial stress rupture tests on similar linear polyethylene materials and with hydrostatic pressure test results obtained on MDPE pipe. The ductile failure mode in the EBT geometry appears to be controlled by a different mechanism than in uniaxial tension tests. |
