Experimental analysis of the effect of suction and step height on boundary-layer transition

Kurzfassung

Laminar flow on transport aircraft is challenging to maintain in the presence of surface imperfections such as steps and/or gaps, which can occur at structural joints. A substantial delay of laminar-turbulent transition, however, can be achieved by suction. This work focuses on the influence of suction on transition in the presence of forward- and backward-facing steps. Systematic experimental investigations were performed in the low-turbulence Cryogenic Ludwieg-Tube Goettingen for large Reynolds numbers (chord Reynolds numbers up to 16 Million), Mach numbers from 0.35 to 0.77 and various stream-wise pressure gradients. The examined wind-tunnel model was a two-dimensional flat plate, designed to have a large area of uniform pressure gradient on the surface of interest (upper side of the model). The model consisted of two main parts that could be assembled with shims of different thickness to obtain sharp steps at 35% of the model chord length. Alignment pins enabled a streamwise displacement to achieve a gap directly upstream of the steps. Suction was achieved passively by exploiting the pressure difference between upper and lower side of the model, induced by the model’s geometry. Depending on the free stream conditions, suction rates of q = vD/d*U of up to around 1 were obtained with average suction velocity v, gap width D, computed displacement thickness d* and the freestream velocity U. Transition was determined non-intrusively by means of Temperature-Sensitive Paint (TSP). Furthermore, the model was equipped with pressure taps to obtain a streamwise pressure distribution. The average suction velocity v was calculated with the help of additional pressure taps installed along the gap inside of the model. The investigation comprised of a reference configuration (without installed step and gap), a gap configuration without step, two forward-facing steps with step Reynolds numbers up to 4800 and a backward-facing step with step Reynolds numbers up to 2400. Compared to the reference configuration, the implemented surface imperfections caused transition to occur at a more upstream location. With additional suction however, transition was significantly shifted downstream and even overcompensated the adverse effect of step and gap, as transition was detected further downstream than for the reference configuration. Detailed results including additional numerical considerations on linear stability analysis will be provided and discussed in the final version of this contribution.