Date of Award
Doctor of Philosophy
James G. Coder
Kivanc Ekici, Ryan Glasby, Devina Sanjaya
Laminar-turbulent boundary-layer transition has a demonstrable impact on the performance of aerospace vehicles. The ability to accurately predict transition is integral to properly capturing relevant flow physics. Traditionally, computational fluid dynamics simulations are performed fully turbulent, meaning that laminar flow is neglected. This, however, can result in errant predictions of vehicle performance as quantities such as skin-friction drag may be overpredicted. Resultingly, development of Reynolds-averaged Navier-Stokes transition models has seen significant attention over the last decades in order to model transition and realize the performance improvements of laminar flow.
In this work, the behaviors of several different transition-prediction methods are analyzed both for their ability to predict transition and vehicle performance. The popular local-correlation transition model is assessed analytically and numerically. It is found that there exists a singularity near the wall after transition to turbulence. This results in singular-like behavior of the destruction term of the turbulent kinetic energy equation and prevents the model from ever achieving asymptotic grid convergence. A turbulence index was developed to more robustly and accurately detect transition relative to other quantities such as turbulent intermittency.
The behavior of the amplification factor transport (AFT) model was examined for a four-bladed helicopter rotor undergoing dynamic pitching conditions. The transition front predicted by the AFT model agreed well with experiments, with exception to that during the upstroke of the pitch cycle. The effect of freestream turbulence intensity on transition was examined by varying the critical $N$-factor, finding that as turbulence intensity increased, the transition front moved increasingly further upstream throughout the pitch cycle.
Additionally, large eddy simulations were performed for a rotorcraft airfoil undergoing dynamic pitching conditions. A laminar separation bubble was found to be the primary mechanism of transition, finding also that the length of the separation bubble decreased as the pitch angle increased. A Kelvin-Helmholtz-like instability was identified near the aft end of the separation bubble which drives transient bursting of the bubble and is partially responsible for the behavior of the transition front. An additional investigation using the AFT model found that predicted transition agreed well, but predicted a shorter separation bubble.
Carnes, Jared Alexander, "Predictive Capabilities of Laminar-Turbulent Transition Models for Aerodynamics Applications. " PhD diss., University of Tennessee, 2022.