Spontaneous cavitation in a Lennard-Jones liquid at negative pressures

Abstract

We report a molecular dynamics (MD) study of homogeneous bubble nucleation in a Lennard-Jones liquid under a negative pressure (cavitation). The rate of bubble nucleation has been determined in the range 2 x 10(-9) < J(*) = Jσ(4)(m/ε)(1/2) < 6 x 10(-6) by the mean lifetime method at temperatures T(*) = kBT/ε = 0.35, 0.4, 0.5, 0.6, 0.7, 0.8, 0.4, 0.5, 0.6, 0.7, 0.8. In molecular dynamics simulation calculations have also been made of the coefficient of bubble size diffusion, the Zeldovich nonequilibrium factor, and the radius of a critical nucleus R*. Different approaches to the determination of the nucleation rate in a stretched liquid have been considered in the framework of classical nucleation theory (CNT). The values of J obtained in MD simulation are by 8-20 orders higher than those predicted by CNT. The work of formation of a critical bubble and the dependence of surface tension γ(R*) at the critical bubble-liquid interface have been determined by data of MD simulation from CNT. The values of γ obtained have been approximated by an extended Tolman formula that takes into account, besides a linear correction, also the quadratic in curvature terms. The Tolman length δ∞ is negative and equals -(0.1-0.2)σ. The coefficient at 1/R*(2) is positive and does not exceed σ(2).

Simple improvements to classical bubble nucleation models

“…Typically, these simulations show large deviations from the CNT predictions. The CNT nucleation rates usually underestimate the bubble nucleation rates by very large factors [15][16][17][18][20][21][22]25,26]. Most of the simulations for bubble nucleation in the literature use around 10 5 or fewer atoms, making it difficult to measure nucleation rates directly.…”

“…Numerical techniques such as molecular dynamics and Monte Carlo simulations are powerful methods which can resolve details of the nucleation process and provide useful test cases for nucleation models [13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28]. Typically, these simulations show large deviations from the CNT predictions.…”

Virial series for inhomogeneous fluids applied to the Lennard-Jones wall-fluid surface tension at planar and curved walls

“…The curvature dependence of surface tension for fluid interfaces is a highly studied issue. [23,25,64,[83][84][85][86][87][88][89] In particular, the amplitude and sign of bending and Gaussian rigidity constants are a matter of discussion. We evaluated analytically both rigidities, k andk, at order z 2 and ρ 2 b .…”

“…Modern studies based on molecular dynamics simulation shed light on its basic properties as viscosity, thermal conductivity, cavitation and melting coexistence. [20][21][22][23][24] Other works have focused on the * iurrutia@cnea.gov.ar curvature dependence of the surface tension. [25][26][27] The virial coefficients of the LJ fluid have been calculated numerically up to sixteenth order [28][29][30][31] and similar studies were done in LJ fluid mixtures up to sixth order.…”

Limiting tensile strength of liquid nitrogen

“…Such a disagreement is also observed between CNT and data of molecular dynamics simulation of cavitation in a stretched Lennard-Jones liquid. 31 The work has been performed with a financial support of the Russian Science Foundation (Project No. 14-19-00567).…”