Development of Numerical Simulator of Ice-accretion on a Body in Oscillation

点击翻译按钮获取中文摘要

Numerical simulator has been developed based on Computational Fluid Dynamics appraoch, which is possible to predict the shape of accreted ice on a body surface in oscillation. In the present simulator, we employ Eulerian approach with 2-fluid description of multi-phase flow model in the basic codes, because this approach has the advantage in simulating flows around an oscillating body. The 2-fluid Saffman equation is used as a model of super-cooled atmosphere Then, we consider potential flow and viscous flow of Naver-Stokes solution as the carrier phase. The present simulator makes the following facts clear.(1) Prediction of local impingement efficiency along an oscillating cylinderBy solving the Saffman equation under the potential flow field of a circular cylinder with circulation in arbitrary translation, local impingement efficiency (distribution of mass flux) along the cylinder surface in pitching or heaving oscillation is obtained in some values of Stokes number. We also find uns … More teady behaviour of density field of water droplets around the cylinder by the visualization of computational results.(2) Prediction of the shape of accreted ice in viscous flowSimulation of hard-rime accretion on a circular cylinder is performed under the condition (Re=10^4, St=10) by coupling the Navier-Stokes equations. In Eulerian approach, there are two kinds of option for the process of hard-rime growth : one is choice of incoming velocity of droplet, the other is the direction of ice-growth. Computational results show each option gives the different shape of hard-rime. Since local impingement efficiency has the most important effect in the process of hard-rime growth, the present results suggest that Eulerian model has advantage of predicting wide variety of ice shapes rather than Lagrangian approach.(3) Density field of Saffman equationDensity field in the stagnation region is obtained analytically as the function of Stokes number based on potential flow model. This expression is a generalization of Michael's result in 1968. Less