Investigation of an Improved Acoustical Method for Determining Airtightness of Building Envelopes. Untersuchung eines verbesserten akustischen Verfahrens zur Bestimmung der Luftdichtheit von Gebäudehüllen

Unintended air infiltration in buildings is responsible for 30 to 50 % of the building stock energy demand. The fan pressurization method, also known as the blower-door test, is the most frequently used and standardized measurement method to evaluate a buildings’ airtightness and determine the airflow through a building or a building element. While detection and quantification of individual leaks with smoke tracers or infrared thermography are challenging, time consuming, and depend on the respective operator’s experience, acoustic methods have the potential to localize and quantify leaks in building envelopes without the need for pressure or temperature difference between in and outside of the examined building. In this thesis, two acoustic methods, coherence measurements and beamforming, are introduced to this field of application to estimate the leakage size and location of individual leaks in building elements. This work aims at finding if different leak sizes can be quantified and detected using these acoustic measurement methods. For an estimation of leakage size, acoustic and airflow measurements are compared in a laboratory test apparatus. Test walls representing a single characteristic air leakage path in the building envelope at a model scale and separate two chambers with speaker and microphones. Various types of wall structures with different slit geometry, wall thickness, and insulation materials are tested. The acoustic measurements are performed with a sound source placed in one chamber and ultrasonic microphones located in both chambers. The results of these measurements are compared to the airflow through the test wall measured using a flow nozzle. The results from laboratory measurements indicate a linear trend between acoustic coherence and leak size in the investigated range of several mm2. Although the acoustic measurement uncertainty is still significant (≈ ±50 %), the acoustic method shows the potential to give an order of magnitude of leak sizes. The findings are validated in a real building setup using various reproducible leaks constructed with cable ties wedged in a window gasket. The acoustic and airflow measurements in the building show results similar to the laboratory measurements. Further, the acoustic beamforming method (“acoustic camera”) using a microphone array to detect leak locations and visualize them shows promising results. The same constructed leaks in a window gasket are detected. With decreasing leak size and sound pressure, detection frequency increases. The thesis concludes that acoustic methods have the potential to detect and quantify the leaks in building envelopes without test preparations as for common methods