168 / 2023-04-15 01:20:54

Validating the ramp loading experiment scheme using molecular dynamics simulation

高压物理与材料科学 > 动高压-海报

Validating the ramp loading experiment scheme using molecular dynamics simulation



Jingxiang Shen¹ and Wei Kang<sup>1

1</sup>Center for Applied Physics and Technologies, Peking University, Beijing, China



As a milder dynamic loading method compared to shock compression, ramp loading is capable of exploring high-pressure states at much lower temperatures. Generally, in laser driven ramp loading experiments, the samples will be compressed to about a quarter of its initial volume and TPa pressure within tens of nanoseconds while still remain solid.



Since strong shocks are avoided, ramp compression is usually considered to be quasi-isentropic. Therefore, reverse hydrodynamic calculations based on isentropic assumptions are usually used for data analysis, which gives an estimated “quasi-isentropic” P(ρ) relation inside the sample using information provided by the free-surface velocity trajectories.



However, there exists effects that may deviate the real flow field away from the isentropic hydrodynamic picture, like anisotropic pressure owning to uniaxial loading, irreversible plastic heating, phase transitions, possible rate-dependent stress-strain relation, etc. Although these effects are generally believed not to affect the validity of the experiment pipeline, a direct validation of the ramp loading experimental method with the presence of these non-hydro effects is still lacking.



Here, we provide such a validation using molecular dynamics simulation. That is, the entire ramp loading experiment process is simulated with non-equilibrium molecular dynamics simulation, and free-surface velocity data is extracted and processed just as in real ramp experiments. This simulated "experimental” results are then compared with isentrope derived from equilibrium molecular dynamics, which is the ground truth here. We found that if the initial shock (the shock prior to the ramp) is properly treated, P(ρ) result obtained by the ramp loading experiment would be consistent with the group truth.



Obviously, due to limitations on the simulation scale, the strain rate in our molecular dynamic simulations exceeds the realistic experimental values by about 3 orders of magnitude. However, we reason that the influence of the non-isentropic effects should actually be more significant for small-scale systems. We have also changed the simulation scale within the range allowed by our computing power and the conclusion has not been altered. Therefore, our results provide a straightforward validation for the ramp loading experimental scheme that is wildly used over the past years.