The provision of an adequate supply of outdoor air suitable for the needs of the occupants is an important aspect of building design and construction. Housing in the United Kingdom accounts for 28% of the total energy used with the cost approximated at 9.6 billion . These high costs in energy losses have urged more rigorous thermal insulation standards to be set in new buildings. Well insulated dwellings with low fabric heat losses have isolated ventilation losses as a large amount of total energy loss of a building. Thus implying a reduction in losses associated excessive ventilation as a cost effective method of energy saving.
In practice the elimination of excessive ventilation requires a method of control, however this is sought to be difficult. Many buildings in the U.K are naturally ventilated with air entering and exiting a building to the outdoor surroundings. . Unnecessary ventilation accounted for over 60% of the energy wastage, mainly through the loss of conditioned air. 
The amount of air entering and leaving a building is dependent on the pressure differential across the building and the characteristics and distribution of the leakage paths in the envelope. Pressure differentials can be caused by the dynamic action of the wind. The driving force for infiltration is induced by pressure difference which is due to the wind acting on the dwelling and the temperature differences across the dwelling. This is known as the stack effect. In a heated building air will rise within the structure due to changes in density. Alternatively it may move in the opposite direction if the air in the building is cooler than the outside. In reality the combines effect of wind and temperature produces complex and variable air flow patterns throughout the building.
It is necessary the reduction of air exchange through ventilation does not drop to such a level where exposure to pollutants in the air may result in a health risk as a result in bad indoor air quality. Fresh air is essential to eliminate contaminants (body odour, tobacco smoke, carbon dioxide) to a level where the indoor concentration does not cause harm to its occupants in the building. Ventilation is also important to get rid of water vapour and allow it to escape to the surrounding outdoor air, to avoid humidity which may lead to problems involving condensation.
The aim of this study was to carry out laboratory testing to measure air leakage flow through adventitious openings from a sash window. The major section of this paper consists of testing through self made timber cracks. The dimensions where changed to calculate the airflow and determine the behaviour of air of straight through cracks where the experimental results and theoretical calculations were compared using both the 'power law' and the 'quadratic formula'. The results in both cases were analysed and discussed to determine the best method for the experiment.
In the infiltration process the amount of air entering a building is primarily governed by the wind speed, wind direction, indoor/outdoor temperature difference and air leakage characteristics of the building. As the meteorological conditions are unpredictable, the specific air flow due to infiltration is a variable parameter which is beyond the control of the occupants. In order to harness the climate parameters which influence infiltration, buildings can be purposely provided with natural ventilation.
To obtain more control over ventilation it is necessary introduce mechanical systems into the building. Air can be removed from a building by a mechanical extract fan or it can be driven into space using a supply system. Extract ventilation driven into a space using a supply system. Extract ventilation necessitates the provision of sufficient openings in the building envelope, to ensure that the incoming air may easily replace that which is extracted. Similarly, with a supply system the displaced air has to leave the building through any adventitious or purpose provided openings in the building fabric. Whilst to some extent negating the influence of climatic parameters, the correct functioning of mechanical extract or supply systems will depend upon the air leakage characteristics of the building envelope.
Air Leakage Characteristics of Buildings
For many buildings the only means of ventilation is air infiltration. This is an entirely passive process and relies upon the fortuitous leakage of air through various cracks and gaps in the building envelope. Typical leakage paths are shown in Figure 1 .
The air leakage characteristics of building enclosures are dependent on the flow characteristics of the leakage path, the pressure difference across the leakage pats and the size and distribution of the leakage path. To assess the airflow through the buildings envelope it is important that these three factors are known. The most important concept of estimating air leakage is that it relies on the mass balance of air movement throughout the whole building. Finding the airflow rate of a building requires some degree of complexity, therefore an estimate can be made using basic equations which is dependent on the size of the gaps.
For relatively large openings such as a vent, the flow is assumed to be turbulent. BS 5925 recommends using Equation 1 for gaps and openings above 10mm. It can be seen that the volume flow rate is directly proportional to the square root pressure difference.
For small narrow cracks the flow is assumed to be laminar. In these cases the flow is rate is given by the Couette equation (Equation 2).
For cracks that have high aspect ratio (wider in length) the flow is the flow is neither in the laminar region nor the turbulent region. This is known as transition. The two above equations can be put together and be expressed as a single power law equation (Equation 3)
Methods of Ventilation and Infiltration
Residential buildings today mostly rely on infiltration for meeting their ventilation requirements. This ventilation rate is measurement is renowned as air changes per unit time, the number of times the interior volume of air is replaced. In theory, air flow entering a building can be calculated if the locations were known by using anemometric methods. However in reality this is not possible, therefore tracer substances must be used. There are three types of tracer methods commonly used today are constant injection, constant decay and constant concentration.
The tracer gas is injected continuously at a constant rate into an enclosed space. To obtain accurate results for mixing, it normally injected at many locations as possible. After some time the tracer gas concentration in the system will reach equilibrium (steady state). The value attained is used to calculate the air change rate. However, using this method can cause problems and irregularities when calculating the ventilation rate. It is unlikely the concentration will remain constant during the test period due to the ever changing weather conditions surrounding the buildings. The most common solution to remove these fluctuations is to take regular readings throughout testing and averaging the value of the ventilation rate. However, the drawback of using this is that the accuracy of this method is dependant on the mixing of the tracer with the surrounding air and even with the use of fans this is difficult overcome.
Constant decay involves the mixing of a tracer gas in an enclosed space. The mixture causes a high value of concentration in which the decay is registered as a function of time in one or numerous points. To calculate the air change rate the logarithm of the tracer gas concentration is plotted against the elapsed time. As this method is not a steady state measurement problems can arise due to the tracer's absorption and adsorption characteristics.
This method uses an automatic dosing used to inject the tracer gas into an enclosed space. The dose is varied to keep the concentration constant throughout the testing period. This method uses complicated equipment employed by feedback devices to measure the tracer gas concentration and the injection rate. The equipment used for this technique is cumbersome to use in dwellings as it is bulky.
An alternative approach to the tracer techniques is fan pressurisation. This method does require a great deal of preparation before tests can be carried out. To the results obtained are dependant on the meteorological conditions, therefore a relatively calm day must be chosen for the testing. Depending on the size of the building several fans may be used to avoid the build up of air towards one section. The accuracy of the results also relies on the structure of the building. For a typical flat, the leakage will not only consist of the amount of air escaping outside but also any cross leakage to adjacent spaces through ceilings, walls and floors. In order to avoid cross leakage reading an additional fan must be used to pressurise/depressurise the adjacent space with the same pressure.
As previously mentioned, air infiltration is very much dependant on the wind pressure and velocity. It is very difficult to have good meteorological conditions throughout the length of the experiments which may take hours to complete. The wind direction and density are likely to change which in effect will give inaccurate results. To reduce to errors due to weather conditions, the experiments will be conducted in a laboratory where the pressure and ambient temperatures are expected to remain constant.
The apparatus consisted of a fan used for pressurising and depressurising the chamber and an orifice plate to measure the rate of fluid flow. Pressure tappings and pressure gauge was used were used across the orifice plate in order to obtain the volume flow rate. A manometer was with two pressure tappings was also used to measure the pressure across the window and cracks, one positioned centrally in the pressurisation box and the other to the ambient surroundings. A shield box was not used to protect against the external pressure fluctuations as a simple sheltered pressure tapping was adequate to obtain accurate results.
As a vacuum cleaner was used as a fan, only depressurisation of the chamber was only carried out. Ideally both pressurisation and depressurisation of the chamber should be carried out and the mean value should be calculated. However, with limited resources this was not possible. The test chamber consisted of a 0.9m X 0.5m X 1.15m cuboid plywood box. One side of the box was omitted so the window could be inserted. The chamber was made of an exact size so that it could accommodate the window without any gaps. Clamps were used to tighten the hold between the chamber and the test specimen. When measuring the individual cracks using timber wood, the missing side was boarded up and cut out accommodate the cracks. Figure 2 shows a schematic diagram of the test rig.
Once the apparatus was set up the experiments were carried out. The pressure drop across the test specimen was set up by adjusting the bleed valve of the fan system. The bleed valve was initially set to fully close and the vacuum was switched on. The corresponding pressure difference between the chamber and surrounding was noted along with the volume flow rate. The bleed valve was adjusted to various pressure levels and the results were taken.
To reduce errors in the experiment background leakage reading were taken. The chamber was covered by a sheet of polythene to make it air proof. The readings from the background leakage were deducted from the specimen tests to gain accurate results.
Draft proofing was also carried out. This was done by duct taping each side of the window where air leakage was found. Initially the whole window was duct taped and gradually each side was stripped and the respective leakage of the cracks was noted. With the use of feeler gauges the crack lengths were defined to obtain the data needed to calculate the theoretical air leakage which was compared to the experimental values.
The pressure differentials chosen for the experiment were in accordance to the B.S standards. In the UK it is normal practice to measure the volume flow rate for a range of steady pressure differences up to 50Pa. Although the apparatus allowed a maximum pressure difference of 100Pa it was deemed unnecessary to carry out experiments at such high pressures. The guidelines also stated that readings must be taken at minimum pressure difference of 10Pa therefore to obtain accurate results an interval of 5Pa was chosen up to a maximum pressure difference reading of 50Pa. This was then repeated all the way down to unity. After the testing was complete the mean values for the volume flow rate were calculated for each pressure reading by using the orifice plate formula.