<p>Multiphase equations of state (EoS) must remain numerically robust during simulations of dynamic loading when large variations of density and internal energy can occur. In the liquid regime, removing van der Waals loops via a Maxwell construction captures the liquid gas transition and prohibits negative pressures (<i>liquid-boiling</i>), while leaving the loops in place allows tension (<i>liquid-tension</i>) and enables approximate spall-like behaviour in the absence of a cavitation model. This paper documents the pragmatic approximations introduced for robustness in a SESAME-style tabular multiphase EoS: namely, <i>extended</i>, <i>bridging</i>, and <i>buffer</i> zones; a monotonic energy-temperature constraint along isochores; and a mass-weighted sound-speed for use with artificial viscosity. Comparisons of the two liquid treatments in 1D hydrocodes are made against (i) shock-vapourisation experiments in lead, and (ii) the tin shallow-bubble-collapse (SBC) experiment. Across shock-vapourisation cases, both liquid treatments reproduce the witness-plate velocity history with only minor differences, although the liquid-boiling option occasionally yields a more ramp-like rise representative of certain lead shots. In the SBC experiment, the computed temperatures with liquid-boiling agree closely with the measurements, whereas liquid-tension under-predicts the temperature rise. The results support using liquid-boiling where temperature predictions inside cavitated zones matter, while documenting the approximations and safeguards that make such EoS practical for production hydrocodes.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Pragmatic Treatments of the Liquid-Vapour Transition for Hydrocode Simulations

  • G. A. Cox

摘要

Multiphase equations of state (EoS) must remain numerically robust during simulations of dynamic loading when large variations of density and internal energy can occur. In the liquid regime, removing van der Waals loops via a Maxwell construction captures the liquid gas transition and prohibits negative pressures (liquid-boiling), while leaving the loops in place allows tension (liquid-tension) and enables approximate spall-like behaviour in the absence of a cavitation model. This paper documents the pragmatic approximations introduced for robustness in a SESAME-style tabular multiphase EoS: namely, extended, bridging, and buffer zones; a monotonic energy-temperature constraint along isochores; and a mass-weighted sound-speed for use with artificial viscosity. Comparisons of the two liquid treatments in 1D hydrocodes are made against (i) shock-vapourisation experiments in lead, and (ii) the tin shallow-bubble-collapse (SBC) experiment. Across shock-vapourisation cases, both liquid treatments reproduce the witness-plate velocity history with only minor differences, although the liquid-boiling option occasionally yields a more ramp-like rise representative of certain lead shots. In the SBC experiment, the computed temperatures with liquid-boiling agree closely with the measurements, whereas liquid-tension under-predicts the temperature rise. The results support using liquid-boiling where temperature predictions inside cavitated zones matter, while documenting the approximations and safeguards that make such EoS practical for production hydrocodes.