SOLID-LIQUID PHASE CHANGE HEAT TRANSFER IN ENCLOSURES
Abstract
The objective of this study was to gain basic understanding and obtain heat transfer data during solid-liquid phase change (melting and solidification) of n-octadecane in enclosures (cavities) of the type which are relevant to practical applications. The enclosures considered in the present study include rectangular cavities with different thermal boundary conditions and a horizontal cylindrical capsule. The shadowgraph technique was used to measure local heat transfer coefficients at the heat source surface. The solid-liquid interface motion during phase change was recorded photographically. The convective motion in the liquid during melting was visualized using aluminum powder as a flow tracer. Independent of the thermal boundary conditions imposed in the rectangular cavities, the results clearly established that the melting rate, the solid-liquid interface motion, and the heat transfer are significantly affected by the fluid motion in the melt induced by the buoyancy force and by the volumetric expansion accompanying the phase change from the solid to the liquid. The latter are significant particularly at the early times while heat transfer is still dominated by conduction. For the solidification starting with a superheated liquid, natural convection did not affect solid-liquid interface motion and was confined to the early stage of the process. During melting in the rectangular, open cavities with isothermal and conducting vertical walls, a strong time-dependent vortex motion has been observed in the bottom melt region. Melting in a horizontal cylindrical capsule was studied experimentally and analytically. The experimental results reaffirmed the dominant role played by the natural convection in the melt during the inward melting in the capsule. In addition to the major natural convective recirculation flow patterns in the liquid, a secondary vortex circulation occurred at the bottom part of the melt annulus. The experimental data are compared with the numerical predictions.
Degree
Ph.D.
Subject Area
Mechanical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.