Gevelber, Michael A.Jayarathne, Tharanga2023-06-052023-06-052023https://hdl.handle.net/2144/46285The ability to measure air leakage in multi-zone buildings, both through the exterior envelope and between different zones inside the building (interior leakage), is important since they have a direct impact on energy consumption, indoor air quality and moisture accumulation. While blower door testing methods to measure air leaks are well established for over 50 years, current tests lack the ability to directly measure the total interior boundary leakage of a zone. In addition, while power law models have long been used to express the building pressure-flow relationships used for code compliance and understanding building leakage behavior, it has not been clearly established what the relationship is between model parameters to the structure of leak paths, testing conditions, and the relationship of model parameter distributions to leak behavior. This thesis a) compares the existing measurement methods, based on field testing in a wide range of buildings, to identify information provided and workload, b) presents the development of two new methods to directly measure exterior and interior leakage of an attached unit, and c) experimentally evaluates the behavior of leak flow for specific crack structures and measurement conditions which is found to have important implications for measuring and modelling building leaks. A new guarded interior test was developed that directly measures the total interior boundary leakage of an attached unit using common blower door equipment. In addition, a new Zonal Multipoint Pressure Testing (ZMPT) method was developed that simultaneously measures both exterior and interior boundary leaks to identify non-linear leak mode parameters of the tested zone. Testing to evaluate the performance of these two new tests was performed in a wide range of multi-family buildings including newly constructed apartment buildings, old brownstones, and existing dormitories. A new closed unit flow balance test was developed to evaluate the consistency of results for both conventional and new testing methods. In addition, the importance of testing conditions, such as utilizing appropriate flow directions, is also evaluated. Leak flow-pressure relationship for building boundaries is commonly modelled using a power law equation in the form Q=C(ΔP)ⁿ where Q is the leakage flow through the boundary, ΔP is the pressure difference across the boundary, n is the pressure exponent and C is the leakage coefficient. The leakage coefficient can be normalized by the surface area, C ̃=C/A_i in order to compare the leak flow through zones with different areas. Building field test measurements reveal a wide range of values for the parameters C ̃ and n, which raises the question as to what causes these different values. To better understand these differences, an experimental study was conducted where airflow through engineered leak pathways of different geometries was measured for a wide range of pressures. These experiments reveal that both the pressure exponent and normalized leak coefficient scale with the crack geometry and testing flow conditions. Similarly, blower door test results obtained for several buildings were observed to also depend on the tested flow conditions. In these cases, the power law model parameters and leakage flow rates at natural low-pressure building conditions (5-20 Pa) were observed to be significantly different than the high-pressure leakage model parameters and calculated low pressure leakage from high-pressure measurements (50-80 Pa). An analysis of measurement uncertainty verified these multiple pressure regime findings. In addition, a correlation between C ̃ and n is also observed in extensive whole building field measurements in multi-family buildings where higher exponent values (n>0.65) tend to be correlated to tighter building boundaries and lower C ̃. This implies that the distribution of n values for a portfolio of buildings can be an indicator of the relative leakiness of building boundaries.en-USMechanical engineeringThesis/Dissertation2023-05-300000-0001-5763-5339