English

This report on a project carried out at the Building Research Association of New Zealand describes an investigation on the role of draught sealing  measures in houses and is part of a wider programme concerned with overcoming heat and moisture problems in buildings.

In order to arrive at a simple design procedure to handle heat transfer in a room, certain fairly complicated expressions for radiant exchange are needed, and moreover the further operations that have to be performed on them are involved and lead to non-exact relations.

A generalized multi-chamber ventilation model is developed for air contaminant prediction problems where the parameters of the system, such as airflow rates, are described by Gaussian probability distributions. A numerical solution, utilizing stochastic differential equations (SD E's), is provided to facilitate its application. The model is used to calculate contaminant concentration histories described by means and standard deviations. It is also used to show the sensitivity of concentrations to the variation of such parameters as infiltration flows and contaminant source rates.

A comprehensive theory of multi-chamber air infiltration measurement using a single tracer gas is introduced from the general stand point of system identification. The thermal network model can be applied not only to the temperature transfer and diffusion system but also to the tracer gas transfer system. This model is formulated mathematically in a stale equation. The coefficents in the state equation represent airflow rates of infiltration. Two theories for estimating these coefficients are deduced from the least square.

Flow measurements using rracer gas reclmiques were made on tile exterior doorway of a test house for indoor-outdoor temperature differetJces of 0.5-45 K. The time for door opening and closing was constant at 3.75 s, and fully open hold time varied from 0.5 s to 120 s. Predictions of a variable density sready flow model were in good agreement with the measurements when adjustmenrs were made for the time-varying size of the opening and for the effect of cross-stream mixing between the incoming and outgoing air streams.