The human voice is vital for daily communication, and speech impairment, or dysphonia, can significantly affect a person's career and social interactions. Understanding the phonation process is crucial for distinguishing between healthy and unhealthy voices. Computer-aided simulation serves as a non-invasive tool for studying human phonation, but replicating the complex nature of this process requires advanced computational capabilities and necessitates model simplifications. This thesis explores various simplifications and the errors they introduce in modeling phonation. Utilizing the simulation tool CFS++, the research extends to accurately simulate the interaction between airflow and vocal fold vibrations, as well as determine the sources and propagation of aeroacoustical and vibration-induced sounds. The study first examines the impact of vocal fold geometry through fully coupled fluid-structure simulations comparing different vocal fold models. It also investigates whether fluid-structure coupling can be simplified to a pure flow simulation, using specific boundary conditions to mimic vocal fold vibrations. Additionally, various aeroacoustic analogies are analyzed to pinpoint sound sources during phonation. The model is further enhanced to incorporate the acoustic effects of the vocal tract, enabling precise calculations of sound fields, such as those produced by vowels.
