Exploration into the use of numerical modeling to assist the two-microphone transfer function free field test method
The two-microphone transfer function technique of measuring absorption coefficient in a free field has remained unchanged since its development in the 1980s. The technique was initially proposed as a way of overcoming the limitations of the impedance tube method due to the sound field within the tube. The free field technique has remained scarcely used. This is due to usage restrictions caused by sound field contributions from diffraction from the test sample edge. Currently, the technique is only valid for instances where the edge diffraction is sufficiently minimized. We use acoustic numerical modeling to study the effects of error sources on the free field technique. Numerical models have been developed and used to quantify theeffects of “image source deviation” and edge diffraction on the implementation of the free field technique. Each error source is quantified independently. Updated guidance on the usage restrictions of the free field technique is provided to the reader. This guidance includes accuracy band analysis for each independent error source. Finally, an improvement to the free field technique using a correction method is proposed. This correction method was informed by results of the numerical models. The geometry of the experimental field is numerically modeled to provide an estimate of the error to the free field two-microphone technique. The numerical model, including the error, is used to define several variables that capture the impactof the error sources on the technique. These variables are subsequently used by the correction method to improve two-microphone free field experimental data. An experimental validation of the correction method was performed for 1" Owens Corning type 705 pressed fiberglass board. The correction method showed improvement to the current two-microphone free field technique for higher frequencies ( 800 Hz) for samples larger than 12" as long as the nearest microphone location is no more than 16.7% of the sample width.
In collections
Stats
Viewed 35 timesDownloaded 7 times