A bio-mechanical model is derived which describes the fundamental principles of laryngectomee substitute voice production. Within the model the substitute voice generator (PE segment) is modelled as an elastic tube which is set into vibrations by streaming air. The model bases on the well known two-mass-model by Ishizaka and Flanagan (1972) wich has been successfully used to describe regular phonation. The morphology of the PE segment is considered by several two-mass-models which are orbitally coupled with spring and damping elements. The main parameters which affect oscillation are vibrating masses, muscle tensions and lung pressure. Within the model, the time dependent minimum aperture serves as measure of PE segment deformations. The performance of the PE-Model is demonstrated by adapting the PE-Model to experimental PE segment vibrations which arc extracted from high-speed sequences.

Index Terms. substitute voice, high-speed-recording, two-mass-model, PE-model