Skip to main content
eScholarship
Open Access Publications from the University of California

Buoyancy and stratification in Boussinesq flow with applications to natural ventilation and intrusive gravity currents

  • Author(s): Flynn, Morris R.
  • et al.
Abstract

Natural ventilation offers an efficient strategy for concurrently disposing of waste heat and improving indoor air quality within the built environment. Most previous studies of naturally-ventilated buildings assume a one- chamber geometry. Here, we relax this assumption and examine by way of theory and experiments the flows that may develop within a multi-chamber domain. A complex internal stratification of buoyancy is typically observed, the details of which depend upon the relative sizes of the adjacent chambers and the size/vertical location of the internal/external openings. In contrast to simple geometries, this stratification is not necessarily eroded by the mechanical action associated with buoyant convection from an isolated internal thermal source. Consequently, the properties of the eventual steady state cannot be determined without investigating the system's transient evolution. Hybrid buildings combine passive summer-time cooling by natural ventilation with active winter-time heating by conventional HVAC systems. A further objective is to explore the inherent challenges associated with this dual design, with particular reference to the hysteretic behavior that may occur when forcing comes from two or more sources, for example internal heat gains and an external wind shear. This thesis also presents a separate investigation of intrusive gravity currents or intrusions, which are associated with density-driven flow along a sharp interface. Theoretical descriptions often stress the similarity between intrusive gravity currents and those that propagate along a solid boundary. Though helpful in certain special cases, this association is inappropriate whenever the intrusion density differs from the depth-weighted mean density of the upper and lower layers, when interfacial waves must be excited. We present herein a more detailed analysis that properly accounts for this upstream influence using two- layer shallow water theory. Model results show good agreement with analogue experimental and numerical data. Finally, we consider intrusions where the initial depth of intermediate density fluid is strictly less than the channel depth such that momentum and energy may be exchanged between the forward- and backward-propagating disturbances. When the upstream interface remains approximately flat, the intrusion speed is accurately predicted using a globally-conservative model

Main Content
Current View