| Popis: |
An extensive reevaluation has been made of several major aspects of the simulation of the structural response of salt to improve the ability to represent actual deformations. As a result of the effort, it is now evident that the Tresca flow potential represents the behavior of salt. The previously used von Mises flow potential does not appear to be appropriate. The multimechanism steady state, workhardening/recovery transient (M-D) model now has been modified to be more representative of salt creep. Reevaluation of material parameters, based on new creep data and a better conceptual understanding of construction of the virgin state parameters has resulted in significant improvement in the material parameters available for calculation. In addition, reevaluation of the coefficient of friction on clay seams was found necessary. Actual underground observations also suggested a revised stratigraphic sequence at the WIPP facility horizon. The combination of the above improvements, when incorporated into a numerical simulation of the closure response of an underground test room (Room D), produced excellent agreement between calculated and measured closures. The agreement between the calculated and measured vertical closures is within 2% for the nearly 4.4 years represented by the experimental data. Agreement between the calculated and measured horizontalmore » closures is not quite as good, with the calculation overpredicting the observed closure by 18%. However, based on these calculations, the approach presented here appears to have resolved much of the earlier discrepancy between calculated and measured closures. 56 refs., 22 figs., 2 tabs.« less |
