Replicating transition with modified Spalart–Allmaras model

Abstract

An algebraic transition model has been developed to preserve the “flow-structure-adaptive” characteristics in “Reynolds-Averaged Navier–Stokes” (RANS) computations for multiple transition mechanisms. The formulation is convenient and plausible in a sense that it relies on the local flow information to trigger transition employing an algebraic intermittency parameter γ rather than a γ-transport equation. The turbulence intensity Tu appearing in γ has been evaluated locally using an empirical relation for the turbulent kinetic energy k, resolving the interaction between local and free-stream turbulence intensities. Splitting γ into low and elevated Tu regimes assists in calibrating the model coefficients as well as minimizing the “trial-and-error” inconsistency, involved in most of the correlation-based transition models for initiating proper simulations. The γ function is directly fed into the production term of a modified Spalart–Allmaras (MSA) turbulence model. The artifact is pivotal to precisely representing the relevant physical aspects of the flow, such as the bypass, natural and separation-induced transitions, and boundary layer (BL) separation and shock BL interactions. Numerical results demonstrate that the MSA transition model maintains decent agreement with other transition and non-transition models available in the literature.