Even a small stenosis asymmetry can trigger post-stenotic transition to turbulence: Implications for modeling and simulation

Sonu S Varghese, Purdue University

Abstract

Direct numerical simulations (DNS) were employed to illustrate that even a small stenosis asymmetry can trigger post-stenotic transition to turbulence. DNS predicted a laminar flow field downstream of an axisymmetric stenosis but the introduction of a geometric perturbation at the stenosis throat, in the form of an eccentricity that was only 5% of the main vessel diameter, resulted in stenotic jet breakdown and transition to turbulence in the post-stenotic region. In the case of steady inflow, transition was accomplished by the breaking up of streamwise, hairpin like vortices into a localized turbulent spot. Under pulsatile conditions, transition was achieved as a starting vortex structure, that formed during early acceleration, broke up into elongated streamwise structures and subsequently underwent turbulent breakdown during peak inlet flow, confirmed by turbulent statistics and broadband velocity spectra. Beyond mid-deceleration, through to minimum flow, the inlet flow lost its momentum and the flowfield began to relaminarize. The start of acceleration in the following cycle saw a recurrence of the entire process of localized, periodic transition to turbulence. Wall shear stress (WSS) at the throat exceeded upstream levels by a factor of thirty but levels were significantly lower in the flow separation zones that formed immediately downstream. Transition to turbulence manifested itself in large temporal and spatial gradients of WSS, with the turbulent region witnessing a sharp amplification in instantaneous magnitudes. A number of popular two-equation turbulence models and the Reynolds stress model were tested against the DNS results but none were able to predict the post-stenotic flow field within any reasonable level of accuracy. Preliminary large eddy simulations indicate that, as of now, this may provide the most computationally viable alternative to DNS to accurately predict post-stenotic transition to turbulence.

Degree

Ph.D.

Advisors

Frankel, Purdue University.

Subject Area

Mechanical engineering|Biomedical research

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS