Answers are grounded in this paper and cite their sources.
Investigates vocal fold vibrations using a novel fluid solver, implying reduced computational costs.
Continuum-based computational 3-D larynx models for simulating the fluid–structure interaction of the vocal fold vibrations during phonation are highly computational expensive when using combinations of Finite Volume, Finite Difference, and Finite Elements methods. The high computational costs are based on cost-intensive solving of the Navier–Stokes equations for modeling the turbulent supraglottal flow field and the communication between flow and structural deformation solver of the tissue model. Hence, in this study, we propose a CFD solver that is based on the Lattice Boltzmann method (LBM) combined with a lumped-mass model involving six masses for each vocal fold reducing the computational time down to approx. 8 min for one oscillation cycle using a workstation equipped with a medium-performance NVIDIA GeForce RTX 2080 Ti GPU. A homogenized LBM based on the D3Q19 velocity discretization is employed to simulate the fluid flow. The vocal folds within the fluid domain are treated as a moving porous medium whose geometry is based on the M5 model. Turbulent effects in the supraglottal region are modeled using a large eddy simulation with the Smagorinsky subgrid-scale model. The simulations show fundamental frequencies and oscillation amplitudes in physiological range. Further, complete glottal closure and clear mucosal wave-like dynamics are achieved.
No takes yet. Share an insight, caveat, or question.
Tur et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: