An Experimental Study of Roughness-Induced Instabilities in a Supersonic Boundary Layer

Laminar–turbulent transition induced by a discrete roughness element in a supersonic boundary layer

“…On average (for 80 x 116) the most unstable varicose mode grows ∼17 % faster than the most unstable sinuous mode (σ K = 0.18 for F = 0.14). These results are in agreement with the BiGlobal stability and experimental results of Choudhari et al (2010) and Kegerise et al (2012), which suggest that the varicose mode dominates over the sinuous mode for high Re h (arguably Re h 426). However, contrary to what was found by Choudhari et al (2010), here the difference in growthrates between varicose and sinuous modes increases for increasing x-position, at least within the streamwise extent of the computational domain used, which could be due to a number of reasons including the higher Re h considered here, increased compressibility effects at the higher Mach number or a dependence on the shape of the roughness element.…”

“…The criteria mentioned above represent a useful tool for predicting roughnessinduced transition, without attempting to provide a physical explanation of the flow instability and transition to turbulence. In order to gain a better understanding of the physical mechanisms driving the roughness-induced transition process, Choudhari and co-workers (Choudhari et al 2010;Kegerise et al 2012) analysed the growth of instabilities in the wake of diamond-shaped roughness elements at M = 3.5, both experimentally and through BiGlobal linear stability calculations. The results show that both sinuous and varicose modes can grow substantially in the linear stages of the transition process.…”

Effect of isolated roughness element height on high-speed laminar–turbulent transition

“…For a k/δ k =0.44 element (Re kk =791), the roughness wake was found to be dominated by varicose mode instability driven by wall-normal shear, while sinuous mode instabilities (associated with lateral shear) were found to be relatively weaker. Results were in general agreement with previous numerical and experimental studies by Choudhari et al (2010Choudhari et al ( , 2013 and Kegerise et al (2012) on isolated diamond elements at Mach 3.5, where sinuous modes were found to experience greater amplification for shorter elements and varicose instability to become more dominant and drive the breakdown to turbulence for higher Re kk . In subsequent DNS studies on square elements by De Tullio & Sandham (2015) in a Mach 6 boundary layer, two main mechanisms responsible for the excitation of wake modes were described: a first one due to the interaction of the local reversed flow with external disturbances and a second one due to the interaction between the roughness wake and the different boundary layer modes.…”

“…Below k/δ k =0.15, x tr is fixed at the corner due to the local destabilising effects and it is then shortened with increasing height, eventually asymptoting towards x tr ≈21±1mm (∼12δ k ). The effective transition length is of a similar order as that found under 'noisy' (highdisturbance) environmental conditions in Choudhari et al (2013) and Kegerise et al (2012). In their studies, the effective transition length induced by diamond elements was shown to be over 2.5 times longer under 'quiet' (low-noise) conditions, where the stability analysis of the element wake was based on N-factor calculations; the effect of environmental disturbance levels on x tr was further found to lead to even greater differences between 'quiet' and 'noisy' conditions for shorter elements, with the breakdown of the dominant modes associated to transition onset exhibiting a strongly nonlinear dependence on roughness height.…”

Investigation on the wake flow instability behind isolated roughness elements on the forebody of a blunt generic reentry capsule

“…Groskopf and coauthors [12] performed temporal 2D eigenvalue analyses on the instability of an isolated 3D roughness element in a Mach 4.8 boundary layer and reported a very good agreement to results from a DNS. Recent measurements via hot-wire anemometry by Kegerise et al [13] behind a diamond element in a Mach 3.5 §at plate boundary layer showed a close agreement of spatial distribution of measured disturbance amplitude with the spatial distribution of the computed eigenfunction for the most dominant wake mode, further proving the applicability of 2D instability theory to wake §ows. So far, 2D instability theory has only been used to calculate the instability characteristics behind discrete roughness elements on a §at plate, but not yet for an isolated roughness element located on the heat shield of a reentry capsule.…”