Compression Waves and Phase Plots: Simulations

Compression wave analysis started nearly 50 years ago with Fowles.[1] Coperthwaite and Williams [2] gave a method that helps identify simple and steady waves. We have been developing a method that gives describes the non-isentropic character of compression waves, in general.[3] One result of that work is a simple analysis tool. Our method helps clearly identify when a compression wave is a simple wave, a steady wave (shock), and when the compression wave is in transition. This affects the analysis of compression wave experiments and the resulting extraction of the high-pressure equation of state.

💡 Research Summary

The paper introduces a novel diagnostic framework for compression waves that distinguishes simple (isentropic) waves, steady shock waves, and intermediate non‑isentropic transitions, thereby improving the extraction of high‑pressure equations of state (EOS). Traditional analysis, dating back to Fowles and Cowperthwaite‑Williams, assumes zero heat conduction and no entropy production, which limits its applicability to modern ramp‑compression experiments that often reach tens of megabars.

The authors derive a key relationship by expressing the Lagrangian sound speed as (C_L(u)=c_0+c_1u+F(u,t)), where the term (F(u,t)) captures non‑linear, non‑isentropic contributions. Introducing (\Gamma^{-1}=(du/dt)^{-1}) leads to the central equation (\Delta\Gamma^{-1}/\Delta h = d(C_L^{-1})/du). When (\partial C_L/\partial t = 0), the wave is a simple (characteristic‑linear) wave; when (\partial C_L/\partial t \neq 0), the wave is non‑isentropic; and as the wave steepens into a shock, (\Delta\Gamma^{-1}/\Delta h) approaches zero, indicating a steady shock.

To validate the method, the authors performed one‑dimensional hydrodynamic simulations with the LLNL CALE code using vanadium EOS (LLNL EOS 9231). Two scenarios were examined: (1) an almost perfectly isentropic compression generated with a very fine mesh (15 000 zones/cm) and minimal artificial viscosity, and (2) a wave that evolves into a shock using a coarser mesh (150 zones/cm) and heightened artificial viscosity (C_Q = 1.9, C_L = 0.75). Particle velocity histories (u(t)) were recorded at 16 Lagrangian points spaced 0.04 cm apart, yielding 136 distinct pairs for analysis.

In the isentropic case, all pairs collapse onto the theoretical curve (d(C_L^{-1})/du), confirming that (C_L(u)) is invariant with respect to the spatial interval (\Delta h) and that (\partial C_L/\partial t = 0). This validates the use of the standard relations (d\sigma = \rho_0 C_L(u) du) and (d\rho = \rho_0 du / C_L(u)) for EOS extraction. In the non‑isentropic case, the authors observed a systematic deviation: (\Delta\Gamma^{-1}/\Delta h) varied with position and time, eventually tending toward zero as the wave steepened into a shock. Consequently, the conventional isentropic formulas become invalid, and the wave can no longer be described by only two families of characteristics; a third family accounting for entropy gradients is required.

The discussion highlights that conventional backward‑integration techniques, which rely on two characteristic families, may produce erroneous (\sigma)–(\rho) relationships when applied to non‑isentropic waves, potentially over‑estimating in‑situ stress. Moreover, even when (\Delta h) is small enough to make the wave appear steady, the underlying non‑isentropic nature can still lead to conflicting EOS results.

In conclusion, the paper provides a practical tool for real‑time identification of wave character based on measured particle velocities, enabling researchers to determine whether standard EOS extraction methods are appropriate or whether more sophisticated, entropy‑aware analyses are required. The authors plan to extend the methodology to quantify entropy production directly, which would further refine EOS determinations for complex, high‑pressure dynamic experiments.