Numerical Investigation of Stationary Crossflow Instability Evolution in a Rapidly Deformed Base Flow

Abstract

The flow of air over a swept wing initially starts in a smooth laminar state (referred to as base flow) and entrains disturbances which subsequently grow and transition the flow to a chaotic, turbulent state. Active efforts have been made to study and control these disturbances, which manifest as stationary crossflow modes. Excrescences in the form of a forward-facing step (FFS) impose a base flow deformation in the form of a rapid near-wall pressure change, flow separation, and strong upwash. This modifies the behaviour of stationary crossflow instability and subsequently leads to an upstream or downstream shift of transition location depending on FFS height. The mechanisms responsible for the above behavioural modification are unknown, motivating the current thesis. The evolution of primary stationary crossflow instability, in its linear growth phase, is studied through a spanwise invariant, synthetic, and idealized rapid base flow deformation imposed by changing the near-wall pressure distribution of a clean swept flat plate (possessing a favourable pressure gradient) via Gaussian-like pressure variations. An energy balance framework is developed that identifies the production term's behaviour as the differentiator between regions of perturbation growth and decay. The behaviour of the production term is described by two competing mechanisms, the first controlled by wall-tangential base flow shear and the second controlled by wall-tangential base flow acceleration or deceleration. The balance of these mechanisms shows that perturbations grow faster than the clean case in regions of wall-tangential base flow deceleration and slower than the clean case in regions of wall-tangential base flow acceleration. Perturbations are even found to attenuate in some cases when a region of wall-tangential base flow acceleration follows a region of wall-tangential base flow deceleration. The modification of energy transfer mechanisms brings into question whether the initial modal stationary crossflow mode deviates from modal character on interacting with the base flow deformation. The Orr mechanism is shown to identify differences in perturbation behaviour from local modal character. However, criteria from the literature hint towards an absence of any non-modal effects. Finally, the extent to which a synthetic idealized base flow deformation mimics the effects of an FFS on deforming the base flow and changing trends of stationary crossflow instability evolution is tested to show the applicability of methods developed in the thesis to instances of natural base flow deformation. The progress in understanding mechanisms by which a deformed base flow affects the linear phase of primary stationary crossflow instability growth leads to suggestions on devices that can be tested to delay this phase of instability growth. These devices could potentially also delay subsequent stages of instability growth and hopefully lead to the development of novel transition delay techniques.