|dc.description.abstract||Improvements in public health through better sanitary plumbing systems has been mainly due to the prevention afforded by barrier technologies to the ingress of foul air, which can contain toxic gases and pathogens, notwithstanding the nuisance of malodour. The main defence against this ingress is the ‘trap seal’ which comes in two forms; the ‘water trap seal’ and the ‘waterless trap seal’. Whilst these devices form effective barriers, they are vulnerable to, or can produce, transient air pressure fluctuations in the system which can lead to seal loss. Greater understanding of the characteristics of these devices is essential for the development of better protection strategies. The development of novel analytical techniques is central to this research as it increases computer model resolution at these important system extremities.
Current methods employ a laboratory only approach, whereby a single loss co-efficient is developed. These laboratory derived boundary conditions are inherently static and in the case of the waterless trap seal, ignore structure flexibility. This research has produced new methodologies to evaluate performance and generate dynamic boundary conditions suitable for inclusion in an existing 1-D Method of Characteristics based model, AIRNET, which solves for pressure and velocity via the St. Venant equations of continuity and momentum in a finite difference scheme. The first novel technique developed uses photographic image and pressure data, transformed via photogrammetric and Fourier analysis to produce mathematical representations of the opening and closing of a waterless trap under transient pressures. The second novel technique developed focusses on the dynamic response of a water trap seal. Current boundary conditions use a steady state friction factor, ignoring separation losses. Analysis via ANSYS CFX allowed a frequency dependent dynamic representation of velocity change in the water trap seal to be developed, integrating unsteady friction and separation losses for the first time. Incorporation of these new boundary conditions in AIRNET confirms that frequency dependent whole system responses are possible and more realistic, reflecting both laboratory and on-site observations.||en_US