The impact of bulk atmospheric motion upon condensation heat transfer in the presence of noncondensable gases

Passive removal of energy to structural surfaces is a determining factor in the response of nuclear containments to reactor accidents. Owing to the high latent heat of the phase change process, condensation is the predominant mechanism of heat transfer. A severe resistance to energy transfer is imposed by the mass diffusion of water vapor to the condensing interface through a boundary layer in which the concentration of noncondensable gases is greater than the bulk conditions. Energy transport is enhanced by bulk motion of the containment atmosphere which serves to mix the gas-vapor boundary layer.;In containment analysis difficulty arises from the specification of a bulk velocity. Forced flow velocities cannot be accurately known a priori to modeling. Natural convection analysis specifies a zero bulk velocity yet this is seldom, if ever, realistic. Although bulk atmospheric motion effects are significant, other independent parameters such as noncondensable concentration, surface to atmosphere temperature difference and atmospheric pressure also play a determining role. The modeling of complex containment energy transport rates requires that all the important variables be considered and that their relative importance be realistically reflected. No model is known to exist which accomplishes such a goal and only the heat transfer package of the CONTAIN accident code has attempted it. Furthermore, an analysis of the heat transfer models in the CONTAIN code exhibit shortcomings in the manner in which bulk atmospheric motion is treated.;A re-investigation is justified due to the recent completion of a considerable containment heat transfer database at the HDR facility. Two statistical approaches are developed. The first method utilizes the HDR database to modify the heat and mass transfer analogy method of the CONTAIN code. An improved method for determining inter-cellular velocities is coupled with a mixed convection formulation of energy transport. In the second method, the database provides empirical normalization to the boundary layer model of Almenas (3). Values are then related to a code-determined parameter indicative of the level of bulk atmospheric motion within a containment volume. Improved predictions for multi-cell models of the V-44 and E11.4 HDR tests are realized.