Font Size: a A A

Nucleation of gas-supersaturated liquids

Posted on:2002-06-09Degree:Ph.DType:Dissertation
University:Michigan State UniversityCandidate:Liu, XiaobingFull Text:PDF
GTID:1461390011490604Subject:Engineering
Abstract/Summary:
Classical nucleation theory relates nucleation rate to the reversible work required to form a bubble of critical size, regardless of any irreversible effects. It has been successfully applied to predict superheat nucleation thresholds of pure liquids, but fails to predict nucleation thresholds of gas-supersaturated liquids and does not account for the effect of volumetric confinement on nucleation. In this study, the equations for dissipation of energy due to viscous and diffusive effects during bubble growing processes were derived and it was shown that, for many nucleation processes, the dissipation of energy due to diffusion may not be neglected as in the classical nucleation theory. Therefore, the classical nucleation rate equation was modified to include a term of dissipation of energy due to diffusion. This modified theory can predict nucleation thresholds of liquids supersaturated with different gases in good agreement with experimental data. For nucleation of gas-supersaturated liquids in small capillary tubes, a dual dependence model originally proposed by Brereton et al. was refined to account for the effect of volumetric confinement on heterogeneous nucleation thresholds. The new model can predict supersaturation nucleation thresholds of gas-supersaturated water in capillary tubes well. It was also found that if the shape factor for nucleation in small capillary tubes is assumed inversely proportional to the tube diameter, a simpler heterogeneous model can predict accurately the supersaturation nucleation thresholds of water supersaturated with different gas species, as well as the superheat nucleation limits of pure water in small capillary tubes.
Keywords/Search Tags:Nucleation, Small capillary tubes, Gas-supersaturated liquids, Account for the effect, Supersaturated with different, Model can predict
Related items